”was
I LIBRARY
My Michigan State
2'") University

This is to certify that the
thesis entitled

WHEN MOMMY AND DADDY GET HOME: THE IMPACT OF
INCOME ON PARENT-CHILD INTERACTIONS FOR SHIFT
WORK AND NONSHIFT WORK PARENTS.

presented by
JILL KATHLEEN ARNOLD

has been accepted towards fulfillment
of the requirements for the

MA. degree in Sociolog

 

 

(dag YW L4H With/0m

Major Professor’s Signature
77' 14% //, 2m?

Date

 

 

MSU is an affinnative—acfion, equal-opportunity employer

 

.— -.—u—---v-.—.-

a—nnuno-a-c-u--.-n—-----u-—o----.-.—. .

..-..—....—.-.- o‘c-nv-

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:/ProllAcc&PrelelRC/DaleDue.indd

 

WHEN MOMMY AND DADDY GET HOME: THE IMPACT OF INCOME ON
PARENT-CHILD INTERACTIONS FOR SHIFT WORK AND NONSHIFT
WORK PARENTS.

By

Jill Kathleen Arnold

A THESIS

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

MASTER OF ARTS
Department of Sociology

2008

ABSTRACT
WHEN MOMMY AND DADDY GET: THE IMPACT OF INCOME ON
PARENT-CHILD INTERACTIONS FOR SHIFT WORK AND NONSHIFT
WORK PARENTS.
By

Jill Kathleen Arnold

This paper examines the impact of income and shift work on parent-child
interaction so as to understand how income and a challenging work arrangement
affect familial time poverty. Specifically, I study the parent-child interactions of
shift work and nonshifi work parents in married dual-earner households across the
income spectrum. I analyze data from the first wave (1987-88) of the National
Survey of Families and Households to determine effects of shift work and
income, independent of and interacting with each other, on the odds that parents
will spend time with their children during meals and other activities. My findings
suggest that Shift work parents in married dual-earner households are as likely to
spend time with their children as nonshifi work parents in married dual—earner

households across all income levels.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................. iv
INTRODUCTION ................................................................................................... 1
LITERATURE REVIEW ........................................................................................ 2
Parents’ Time: Something Has Got to Give ...................................................... 2
The Challenge of Shift Work ............................................................................. 5
Time Poverty: Does Income Play a Role? ......................................................... 7
Shift Work and Income in Conjuncture: Its Impacts ......................................... 8
Research Questions ............................................................................................ 9
DATA AND METHODS ...................................................................................... 10
Measures of Model Predictors and Covariates ................................................ 12
Measures of Model Outcomes ......................................................................... 15
Sample Descriptives ......................................................................................... 17
FINDINGS ............................................................................................................. 18
DISCUSSION ........................................................................................................ 23
FUTURE DIRECTIONS FOR RESEARCH ......................................................... 25
APPENDIX ............................................................................................................ 28
REFERENCES ...................................................................................................... 35

iii

LIST OF TABLES

Table 1. Description of Independent Variables Used in the Analysis ................... 13
Table 2. Description of Dependent Variables Used in the Analysis ...................... 16
Table 3. Bivariate Percentages of Dependent Variables ........................................ 19

Table 4. Logistic Regression Predicting the Likelihood that Parents Eat Dinner
with at Least One Child over Four Every Night .................................................... 29

Table 5. Logistic Regression Predicting the Likelihood that Parents Eat Breakfast
with at Least One Child over Four Two or More Days a Week ............................ 30

Table 6. Logistic Regression Predicting the Likelihood that Parents Engage in
Leisure Activities Outside the Home with at Least One Child over Four Several
Times a Week or More .......................................................................................... 31

Table 7. Logistic Regression Predicting the Likelihood that Parents Engage in
Private Talks with at Least One Child over Four Almost Everyday ..................... 32

Table 8. Logistic Regression Predicting the Likelihood that Parents Engage at
Least One Child over Four in Play or Projects Almost Everyday ......................... 33

Table 9. Logistic Regression Predicting the Likelihood that Parents Help at Least
One Child over Four with Reading or Homework Almost Everyday ................... 34

iv

Introduction

A debate rages on in public, political, and academic spheres over whether
changes in American families are strengthening or weakening the family as a
social institution. At the core of the debate, and often the measure that journalists,
pundits, and academics cite as support to this argument, is how much time parents
spend with their children.

In academia, and in particular Sociology, the question of how much time
parents spend with their children is a well-covered topic, yet sociological research
on the subject reveals contradictory answers. Demographic trends such as
increases in employment hours for both men and women, along with greater
tensions in family relations, are all cited as sources for decreases in parental time
with children (Preston, 1984; Presser, 1989; Schor, 1992; Hocschild, 1997).
However, while these sources result in more familial “time poverty,” more
parental hours at work do not necessarily translate into less parental time with
children. Parents have various avenues to pursue in order to make time for their
children, and some scholars argue that parents most often sacrifice leisure time
and sleep to do just that (Sayer et al, 2004; Sandberg et al., 2001; Bianchi et al.,
2006)

The debate over parental time with children occurs in a number of
different approaches such as analysis of historical trends within the family,
changes in family formations, work patterns, challenges particular to low-income
families, etc. The terrain of this topic is expertly plotted, yet how income might

influence parent-child interactions in a particularly challenging work schedule,

such as shift work, is overlooked. Income is an important factor to include in the
analysis of parent-child interactions between shifi workers and nonshifi workers
so to understand how income and time poverty intermesh and play out in the daily

lives of families.

Literature Review
Parents’ Time: Something Has Got to Give

Societal concerns regarding the amount of time parents spend with their
children, and in turn debates of whether parents spend enough time with children,
appear as natural concerns. Yet, why do we assume parental time with children is
important? Why Should we care whether little John or Jane sees his or her parents
more than a couple of hours a week? Do parent-child interactions really make a
difference?

In fact, social research demonstrates that parental interactions with
children do matter. Parent-child interactions have been shown to play a major
role in a child’s development both intellectually and emotionally (Hofferth, 1998).
Moreover, research by Nock and Kingston (1988) and Pleck (1997) illustrate the
connection between parental time and the well being of children, solidifying our
natural assumptions into sociological knowledge. As sociologists confirm the
importance of parent-child interactions for the well being of children, attention
turns to how much time parents are spending with their children.

Apprehensions of familial time poverty due to increases in parental paid

work, specifically those children might be getting the short-end-of-the-stick

spread throughout 19803. Demographers who examined parental paid work in the
United States during the 19805 concluded that parents who were facing time
poverty did spend less time with children (Preston, 1984; Presser, 1989). In the
early 905, Julie Schor (1992) published The Overworked American, again
demonstrating large increases in American paid work, arguing that Americans are
time poor. She further argued that such time poverty wreaks havoc on our lives,
including our families. While Schor’s research started a hot debate surrounding
work trends, her research also supported those who argue that increases in
parental paid work decrease parental time with children. In continuation of this
research, Hochschild (I999) qualitatively explored parental paid work and its
relationship to family life in her book The Time Bind. Hochschild argues that
parents spend more time at work rather than at home because family life is
increasingly stressful and emotionally strained whereas work life is increasingly
supportive and affirming. This switch between the worlds of work and home
reduces parental time with children. While parents report feelings of guilt
regarding less time with their children, they maintain high time commitments to
paid work. This thread of research centered on increases in parental paid work,
concludes that parents Spend less time with their children due to demographic
increases in paid work. However, this conclusion has been hotly debated among
social scientists with alternative conclusions drawn regarding parental time with
children.

Whether Americans work more paid hours is the first proposition that is

debated in the argument that increases in parental paid work result in decreases in

parental time with children. A closer inspection of paid work reveals a more
complicated story for American work trends. Although some Americans are
indeed working more hours, others are not (Robison and Godbey, 1,996; Jacobs
and Gerson, 1997). Moreover, in examining work hours based on family
formations, Jacobs and Gerson (1997) report that dual-earner and single-parent
households are particularly susceptible to time poverty. Differences in paid work
hours across family formations are an important consideration when considering
time poverty among families. In addition and contrary to Hochschild’s results,
Jacobs and Gerson (1997) find that parents want to spend less time at work in
order to spend more time with their children. This research on paid work hours
indicates that the presumed inverse relationship between time in paid work and
time with children cannot be generalized to all workers who have dependent
children.

That parental time with children is decreasing because of increases in
parental paid work is a second proposition debated among social scientists. Many
scholars contest zero-sum measures that equate increases in parental paid work
with decreases in parental time with children (Bianchi et al., 2006). Analysis of
historical trends finds that parents and children spend the same amount of time, if
not more, with each other. Sandberg and Hofferth (2001) find that the time
American children spend with parents did not decrease between 1981 and 1998
and actually increased for dual-earner families. Similarly, Sayer, Bianchi, and
Robinson (2004) find that it increased between 1965 and 1995. In addition to

historical trends of parental time with children, scholars advocate for “time-

deepening” studies that holistically examine various and multiple ways in which
people spend their time rather than focusing on one dimension of time (Bianchi et
al., 2006). Bianchi, Robinson, and Milkie (2006) again find overall parental time
with children has not decreased over the past few decades. Rather than cutting
down on time with children, these researchers find that parents are spending a lot
less time on household labor, civic pursuits, and together, household labor being
by far the largest decrease. This thread of research illustrates the complex nature
of time poverty and its impact on families.

But what about particularly challenging work arrangements? How might
parents working night shifts make time for their children when nights are largely
when children are at home? Additionally, how might income impact familiar
time poverty in challenging work arrangements, such as shift work? I review the '

literature on these questions in the following section.

The Challenge of Shift Work

The amount of parental paid work is at the center of debates on familial
time poverty. Yet the hours of the day when parents are working are equally
important for understanding familial time poverty, particularly parental time with
children. “Shift work,” defined as the majority of weekly paid work hours outside
of 8 AM and 4 PM (Presser, 2003), is increasing popular among parents, as well
as the population at large. V It means that parents often work during evening and
night hours. Increases in the prevalence of shift work are accounted for by a

number of factors. Presser (2003) argues that the rapid increase in shift work is a

result of changes in the economy, demography, and technology; specifically, to
booming service sector, demand for entertainment and recreation during evening
and night hours, and new technologies, such as email, all of which contribute to
workers being “on call” all hours of the day. Together these structural changes
result in an extension of the workday, thus more shift work. In turn, we see
changes in the segment of the day that parents work.

More parental shift work and the resultant increase in evening and night
hours result in parental work arrangements that run counter to children’s school
schedules. This conflict between the schedules of shift work parents and school-
aged children impacts parental time with children. In-depth qualitative studies
illustrate how shift—work parents construct opportunities to spend time with their
children. Specifically, Garey (1999) finds that female nurses in shift work
schedules are able to maximize time for traditional maternal tasks with children
by manipulating their schedules. In addition, Tubbs et al (2004) finds that low-
income shift-work mothers construct time with their children within the time and
financial parameters available to them. This research suggests that shift-work
mothers are often successful in setting aside time to spend with their children.

While these findings present a positive picture of shift-work parents’
ability to balance work and children, quantitative studies comparing shift-work
and nonshift-work parents indicate the former spend less time with children.
Using time-diary data, Nock & Kingston (1988) find that parental shift work
results in less time spent with children when compared to parental nonshift work.

Phillips (2002) discovers mixed effects in parental involvement for shift work

parents with children Six to eleven. Shift work parents are more involved in
children’s schoolwork, but less engaged in extracurricular activities than nonshift
work parents. Presser (2003) finds that shift work parents spend less time with
children than nonshift work parents in a number of different activities, such as
eating breakfast and dinner, having private talks, working on projects, etc. This
research suggests that while shift work parents may be successful in making time
for children, they are less successful than are nonshift work parents. In order to
tease out the relationship of shift work and parental time with children, though, it

is important to examining other factors that may play a role in familial time

poverty.

Time Poverty: Does Income Play a Role?

Prior research demonstrates the significance of when parents work for the
time families spend together. However, scheduling incompatibility of work with
family time is not the only factor that likely influences familial time poverty. The
role income plays in time poverty is a complex relationship that begins with
parent-child interaction. In Unequal Childhoods: Class, Race, and Family Life,
Lareau (2003) illustrates the sharp class differences in the time parents spend with
the children, both in the type of activities and the intensity of the interaction.
Lareau’s examination of class differences in parental time with children has
implications for various dimensions of class. While class is an abstract concept

drawing from cultural, material, and ideological differences, income is a concrete

material difference that leads to class distinctions. Therefore, income is related to
differences in parental time with children.

How income is related to familial time poverty is under debate within
Sociology. There are three possible connections. First, those higher up in the
occupational hierarchy can face multiple complex responsibilities requiring long
work hours beyond the traditional eight-hour day shifi. The structure of the job
creates both a high income and a constraint in making time for children (Jacobs
and Gerson, 1999). In other words the relationship between high income and
familial time poverty is held to be spurious. Second, higher income can reduce
familial time poverty if it is used to hire a housekeeper or an au pair or to
purchase Iaborsaving devices and services (e.g., microwave ovens, take-out
meals). Third, familial time poverty can reduce income if highly paid
professionals (usually married mothers) ease the time squeeze by quitting their
jobs to become full-time homemakers (Moen, 2008, Stone, 2007). Since most
studies of the connections between labor force participation and time spent with
children have not included both high-income and low-income families, the

relationship between income and familial time poverty is currently unknown.

Shift Work and Income In Conjuncture: Its Impacts
Given the associations of shift work and income with familial time
poverty, examining the interactions among these factors is a reasonable next step.
Most research on shift work approaches income as either a characteristic of shift

work or a sample of interest. Scholars examine the income bracket in which shift

workers are more likely to be, or they focus solely on low —income families
because shift work is likely to correspond to low—wage jobs. For example, Presser
(2003) establishes a relationship between shift work and low-wage jobs through
an analysis of the top ten occupations in shift work and earnings by shift workers
versus nonshift workers. She concludes that shift Workers do not earn more than
nonshift workers. In the second vein of research, scholars often choose to study
family outcomes of parental shift work in low-income families (Perry-Jenkins,
2004; Heyman, 2000; Tubbs et at, 2004; Roy et at, 2004; Rubin, 1994).
Therefore, prior research on shift work and income has significantly contributed
to our understandings of how these two factors interact but remains incomplete.
What is lacking are studies that include the broad spectrum of parental wages.
Therefore, the purpose of the current study is to re-examine the interrelationships
of shift work and the frequency of parent-child interactions by type of activity for

parents irrespective of occupational status.

Research Questions
This study extends current research on shift work and parent-child
interactions by including the possible interaction between shift work and income.
While shift work is disproportionately represented in low-wage jobs, it is not
confined to these jobs. Shift work occurs in a wide spectrum of income levels and
may have a differential impact on parent-child interactions depending on income
level. Specifically, this paper analyzes how income affects parent-child

interactions among shift workers and. nonshift workers.

I anticipate that parents in shift work Spend less time with children
compared to nonshift work parents (Hypothesis 1). In addition, I anticipate that
income is associated with the amount of time parents spend with their children in
a number of activities (Hypothesis 2). Finally, I hypothesize that a low-income
shift-work parent will spend less time with children than either a higher-income

shift-work parent or a low-income nonshift work parent (Hypothesis 3).

Data and Methods

I use sequential logistic regression models to explore the impact of shift
work and the interaction of shift work and income on parent-child interactions:
particularly the likelihood that parents in shift work versus nonshift work
participate in various activities with at least one of his/her children including
eating meals, private talks, etc. These analyses utilize data from a nationally
representative sample, the National Survey of Families and Households (N SF H).
This study examines data from the first wave in 1987-88 (Sweet et al., 1988)
rather than more recent second and third waves because the first wave contains
the largest number of children, ages 5 to 18. As the National Survey of Families
and Households is a longitudinal study, most children in the first wave were over
age eighteen by the later waves. The NSF H is well suited to test the three
hypotheses, as it contains questions regarding work schedules, couple wages and
salary, and, frequency and type of parent-child interactions. Given that this data

set contains information collected from household members ages 19 and older,

10

information on parent-child interactions is reported from the perspective of the
parent rather than the child.

I limit my analytic sample to parents who are married with at least one
child between the ages of five through eighteen living in the home. In addition, 1
only include parents who worked at least fifteen hours in the last week if their
spouse also worked fifteen hours or more in the last week. I focus on married
dual-earner couples for two reasons. First, these families are the most common
formation in the United States today accounting for 54% of all couples in the US.
in 2001 (US. Census Bureau, 2000). Secondly, and perhaps more importantly,
dual-earner couples are arguably the most time starved family formation (Jacobs
and Gerson, 1997). In order to understand the effect that Shift work may have on
parent-child interactions, it makes sense to start with the most time starved family
form when looking at how income affects time with children. The sample is also
limited to those individuals who filled out the parent-child interactions questions
in the at-home questionnaire. Sensitivity tests suggest that parents who did not
complete the questionnaire are not markedly different from those. who did. The
resultant dataset contains information on 1,032 individuals, of which 7% worked
nonstandard shifts in the last week.

Logistic regression models are used to predict the likelihood of each
outcome controlling for relevant covariates. These models are run in SPSS using
NORREG. Each model estimates the odds of scoring a “1” on the dependent

variable, such as breakfast (i.e. eating breakfast with at least one child two or

11

more days per week) relative to scoring a “0” (i.e. eating breakfast with at least

one child less than two days per week).

Measures ofModel Predictors and Covariates

Table 1 defines the measures of predictors and covariates. This study
utilizes Harriet Presser’s definition of shift work, which is modeled from the
Bureau of Labor Statistics (US. Department of Labor, 1981). According to the
Bureau of Labor Statistics, nonshift work or a fixed-day schedule is one in which
at least half the hours worked most days in the prior week fall between 8 AM.
and 4 P.M. Shift work schedules are those in which the majority of weekly work
hours fall outside of 8 AM. and 4 P.M.

Presser and the Bureau of Labor Statistics differentiate various shift work
schedules including fixed evening, fixed night, and rotating schedules. Although
Presser argues that these different shifts affect parent-child interactions in various
ways, the number of individuals within these data who work each of these shifts is
very small, fewer than six percent in each category. In order to gain more
statistical power, I collapsed evening, night, and rotating shifts to create one
measure of shift work. AS a result, 15% of the sample were engaged in a shift
work arrangement.

As the analytic sample consists of married, dual-earner parents, I measure
household income using the couple’s wages and salary. This measure is
appropriate as both spouses are contributing income to the family budget and

presumably have access to one another’s monetary contributions. AS is

12

commonly a problem with measures of income, this variable is missing in seven

percent of cases.

Table 1. Description of Independent Variables Used in the Analysis

 

 

Mean/
Variable Name Definition Prmrtion

Work Status (N=985)

Majority of work hours are 0.15
Shift Work (D) outside of 8 AM and 4 PM

Majority of work hours are 0.85
Non-Shift Work (D) (R) between 8 AM and 4 PM
Household Income Level (N=103l)
High Income (D) Greater than or Equal to $48,755.56 0.26

Less than $48,755.56 and 0.22
Middle-High Income (D) Greater than or Equal to $36,493.35

Less than $36.493.35 and 0.29
Middle-Low Income (D) Greater than or Equal to $24,000.00
LOW Income (D) (R) LCSS than $24,000.00 O.I3
Demographic Covariates
Sex (N=1032)
Male (D) Self-identified as male 049
Female (D) (R) Self-identified as female 051
Race (N= 1030)
White (D) Self-identified as White 0.76

Self-identified as Black,
Non-White (D) (R) Hispanic, Asian, or Other 0.24
Education (1031)
Less than High School Diploma/GED (D) Eleven years Of education or less 0.12
High School Diploma/GED (D) (R) “lgh school diploma 0' GED 0.44
Education beyond a
More than High School Diploma/GED (D) high school diploma or GED 0.44
Young Children (1032)
Children under the age of four
Children under Four at Home (D) in the home 0.3
NO children under the age of four
No Children under Four at Home (D) (R) in the home 0.7
Number of Children (1032)
Three of more children
Three or More Children at Home (D) in the home 0.32
Two or less children

Two or Fewer Children at Home (D) (R) in the home 0.68
Number of Work Hours in Past Week Number of hours worked
Qi=1032) in the past week 40.98
Notes: (D) Dichotomous;
(R) Reference Category

13

In order to reduce the number of cases with missing data, I used individuals with
valid data on education and household income to impute data on the latter
variable, when the data on household income were missing. The imputation
resulted in only missing one person, consistent with the number of people missing
within the educational variable. Thereafter, I collapsed household income into
categories of high-income (equal or above $48,755.56) middle-high income
(between $48,755.55 and $36,493.35), middle-low income (between $36,493.34
and $24,000), and low income (below $24,000) based on couples’ wages and
salary quartiles.

In addition, the models include controls for demographic characteristics
that are known to affect the frequency and type of parent-child interactions, as
well as the likelihood of Shift work. First, respondent’s sex is included, coded as
a binary variable (I=men, 0=women). Prior research finds that family activities
are gendered, as men participate in active care at higher rates than passive care
(N oonan, 2001). Secondly, respondent’s race is included coded as a binary
variable (I=White, 0=Non—White). This variable is collapsed because there are
not enough cases in each racial/ethnic group to make more precise distinctions.
Whites account for the largest percentage (75.9%) followed by Blacks at 14.6%.
Other racial groups were much smaller including American Indians at 0.3%,
Asians at 1.2% Mexican/Chicano/Mexican-Americans at 6%, Puerto Ricans at
0.6%, Cubans at 0.4%, and other Hispanics at 0.9%. In order to compare racial
backgrounds in any meaningful way it is necessary to combine minority racial

groups into the broad group of non-White. Thirdly, completed education of the

14

respondent is included coded dummy variables including, high education
(I=More education than a high school diploma/GED) and low education
(1=Eleven years of education or less). High school diploma/GED is the excluded
and thus comparative category for these variables.

In addition, measures of household composition are included. I assess
whether there are any children under the age of four in the household (1=at least
one child four and under in the home, 0=no children under the age of four in the
home). The number of children in the home is also included as a binary variable
(1=three or more children in the home, 0=two or fewer children in the home).
Note that the number of children includes children from the ages of 0 to 18.
Aside from these family composition measures, I include a measure of the
number of hours the respondent worked in the past week. This is included as a
continuous measure ranging from fifteen to ninety-five hours worked in the last
week. Given that there are only twenty-four hours in a day, the number of hours
parents spend on the job is going to impact the amount of time he/she has for

children, regardless of the type of work he/she does.

Measures of Model Outcomes
Measures of parent-child interactions within this study focus on parent-
child meals and activities where a parent spends time with at least one child over
the age of five in these various interactions. A full coding description of the
measures of model outcomes is presented in table 2. These measures make use of

Six questions, two regarding meals, four regarding other activities. Questions

15

regarding parent-child meals are coded on an 8-point scale, indicating the number
of days that a meal was shared between the responding parent and at least one

child over the age of four.

Table 2. Description of Dependent Variables Used in the Analysis

 

 

Mean/
Variable Name Question/Definition Proportion

Parent-Child Breakfast (N=987)
Two or More "How many days last week 0.44
Mornings at Week (D) did you eat breakfast
Less than Two with at least one of your children 0.56
Momings a Week (D) (R) over the age of five?"
Parent-Child Dinner (N=993)
Seven Evenings "How many days last week 0.5
a Week (D) did you eat dinner
Less than Seven Evenings with at least one of your children 0.5
a Week (D) (R) over the age of five?"
Parent-Child Leisure Activities
Outside the Home (N =I027)
Several Times "How often do you
a Week or More (D) spend time with the children 0.5
Less than Several Times in leisure activities away from home
a Week (D) (R) (picnics, movies, sports, etc)?" 0.5
Parent-Child Play Time (N=1029)
Almost "How often do you
Everyday (D) spend time with the children 0.51
Several Times at home working on a project
a Week or Less (D) (R) or playing together?" 0.49
Parent-Child Private Talks (N=1010)

"How often do you 0.42
Almost Everyday (D) spend time with the children
Several Times having private talks?" 0.58
a Week or Less (D) (R)
Parent-Child Reading
or Homework (N=1026)

"How often do you 0.59
Almost Everyday (D) spend time with the children
Several Times helping with reading 0.41
a Week or Less (D) (R) or homework?"
Notes: (D) Dichotomous;
(R) Reference Category

Possible responses range from zero to seven days in the last week. Responses to
questions regarding the frequency of other parent-child activities are based on a 6-

point scale, indicating the responding parent and at least one child over the age of

16

99 ‘6

four engaged in these activities “never or rarely,” “once a month or less, several
times a month,” “about once a week,” “several times a week,” and “to almost
everyday.

Responses to all of these questions are heavily skewed. For example,
most parents eat breakfast with at least one of his/her children less than two days
a week. Conversely, most parents eat dinner everyday with at least one of his/her

children. Therefore, I collapsed these variables into binary variables with

approximately half of the respondents fall'into each category.

Sample Descriptives

Among respondents in the data set, shift work occurs throughout the
income Spectrum. Shift work is equally likely across the income spectrum for
parent in low income households: 21.4% of shift work parents are in low income
households, compared to 31.7% of shift work parents in middle-low income
households, 20.7% of Shift work parents in middle-high income households and
26.2% of shift work parents in high income households. While this runs counter
to previous research that shows shift work disproportionate in low-wage jobs, it
may be the result of using couple’s income rather than individual income
measures. Perhaps the addition of spousal income is responsible for the increased
diversity in income for shift workers.

Shift work is slightly more common among men compared to women—
15.1% of men compared to 14.4% of women working in nonstandard

arrangements- but this difference is not significantly different at the .05 level.

17

Shift work is slightly more common among Non-Whites compared to Whites —
13.5% of Whites compared to 18.2% of non-Whites engage in shift work
arrangements — although not significantly different at the .05 level. Shift work is
not significantly related to either the presence of young children in the home or
with the number of children in the home. Shift work is not more common among
parents with young children at home: 29.7% of Shift work parents with young
children in the home compared to 29.9% of nonshift work parents with young
children in the home. Shift work is also not more common among parents with
three or more children in the home compared to two or fewer children: 37.9% of
shift work parents have three or more children in the home compared to 31.2% of
nonshift work parents with three or more children in the home.

In these data there are significant associations between education and
parent’s weekly hours. Nonshift work parents are considerably more likely to
have education beyond high school and work more hours than shift work parents.
Nearly 90% of nonshift work parents have education beyond a high school
diploma or GED compared to 82.3% of shift work parents. Moreover, nonshift
work parents work more hours during the week compared to shift work parents.
Nonshift work parents work an average of 42.88 hours a week compared to Shift

work parents who work an average of 40.03 hours a week.
Findings

Employed respondents were significantly more likely to eat dinner seven

nights per week with at least one child over age five if they were not doing shift

18

work than if they were in Shift work (52% v. 41%; Table 3). On the other hand,
the majority of both groups of parents ate breakfast fewer than two mornings per
week with at least one child over age five, and this tendency was somewhat

stronger for nonshift work parent than shift work parents (58% v. 51%; Table 3).

Table 3. Bivariate Percentages of Dependent Variables
All Shift Work N onshift Work
Parent-Child Meals Parents Parents Parents

Parent-Child Breakfast (N=985)
Two or More

 

 

 

Mornings at Week (D) 43% 49% 42%
Less than Two

Mornings a Week (D) (R) 57% 51% 58%
Parent-Child Dinner (N=985)

Seven Evenings a Week (D) 51% 41% 52%
Less than Seven Evenings a Week (D) (R) 49% 59% 48%
Notes: (D) Dichotomous;

(R) Reference Category

A closer inspection of how shift work and the covariates predict the odds
of eating dinner with a child over age five is found in a logistic regression (Table
4, see Appendix). Being a man and working long hours significantly reduce the
odds of nightly dinner with child. But even after these covariates are controlled, a
shift worker is only 58% as likely as a nonshift worker to have that nightly dinner.
A long number of working hours reduce the odds that a parent will eat breakfast
with a child at least two mornings weekly (OR=.98; Table 5, see Appendix). But
shift workers and nonshift workers have even odds of doing so (OR=.76, not
significant; Table 5). Therefore Hypothesis 1 is rejected for the meal of breakfast,
but not for dinner. This finding reflects the temporal ridgedness of dinner.

Dinner can only occur during a small window within the day, a window that shift

work parents miss more often due to evening and night hours.

19

Income did not prove to have a significant relationship with the likelihood
of eating either breakfast or dinner with children (Table 4 and 5). Thus,
Hypothesis 2 is rejected for eating breakfast and dinner with children. Parents
with low household income are just as likely to eat dinner every night with
children than parents with high household incomes. In addition, Hypothesis 3,
that states an interaction between shift work and income will prove to be a
significant predictor of a parent’s time with children, is also rejected for parental
time eating breakfast and dinner with children (see bottom of Tables 4 and 5).
This is not surprising as one of the variables, income, has no independent impact
on breakfast and dinner.

Among the covariates, sex of the parent proves to be significant for
breakfast until weekly work hours is introduced into the model. Men are 30%
less likely than women to eat breakfast with children at least two mornings
weekly (Table 5). However, when weekly work hours are introduced into Model
3 to form Model 4 (Table 5), the effect of sex disappears. This indicates that
While men are less likely than women to eat breakfast at least two days a week
with their children, this sex difference is actually a function of men working more
hours than women.

In addition to sex, parental education is significantly related to the
likelihood of eating breakfast with at least one child over age five two or more
times a week. Parents with education beyond a high school diploma or GED are
1.41 times as likely to eat breakfast two or more times a week with at least one

child over four than are parents with a high school diploma/GED, regardless of

20

other variables added into the model (see last model, Table 5). This finding
contradicts the 'time poverty' idea that more highly educated respondents would
have heavier job commitments that would conflict with time to spend with
children. Perhaps more educated parents have job schedules that allow them to
report to work later in the morning (10 AM instead of 8 AM) which would
facilitate breakfast with at least one child at least two mornings per week.

There is no relationship between shift /nonshift work and leisure activities
outside the home with children, playing with children, private talks with children,
and helping children with homework or reading (Tables 6, 7, 8 and 9, see
Appendix). Thereby, Hypothesis 1, which states that a shift work parent will
spend less time with children than a nonshift work parent, is rejected for all four
parent-child activities.

Contrary to Hypothesis 2, income is unrelated to these four activities,
except for help with homework and/or reading, where a parent from a middle-low
income household is 1.56 times as likely as one from a low-income household to
engage in this activity with a child (Model 8, Table 9). This finding is surprising
as income is not significant on any other parent—child activities. Perhaps parents
in middle-low income households put greater priorities on helping children with
reading and homework than in lower-income households as a way of encouraging
upward mobility. Hypothesis 3, which states that an interaction between shift
work and income will prove to be a significant predictor of a parent’s time with

children, is rejected for all four parent-child activities.

21

Among the covariates, parental sex proved to be significantly related to
parent-child activities. Men have lower odds than women of engaging in play
(OR=.76) in table 8, private talks (OR=.35) in table 7; and help with reading or
homework (OR=.37) in table 9 with at least one child over the age of five on a
daily basis. Similar to breakfast and dinner, parental sex differences in play with
children is a function of parental weekly work hours rather than parental sex.
When parental weekly work hours are introduced into the model of parental play
with children, the effect of sex disappears (compare Models 3 and 4; Table 8).
This indicates that while men are less likely than women to play with their
children, this sex difference is actually a function of men working more hours
than women. In addition to sex, parental race is significantly related to the
likelihood of helping at least one child over five with reading or homework.
White parents are 58% less likely to help at least one child over five with reading
or homework on a daily basis than Non-White parents. This finding indicates that
Non-White parents are taking more responsibility for their children’s success in
school upon themselves compared to White parents, perhaps as a strategy for
intergenerational upward mobility. Not surprisingly, parents with more than three
children are 28% less likely than parents with two or less children to engage at
least one child over four in daily private talks. The more children parents have
the less time they have for any one child and it seems the less time they have for

private talks.

22

Discussion

This paper seeks to contribute to the debate of familial time poverty,
specifically parental time with children. Given increases inwork demands of
parents, and especially shift work parents, which often translates into increases in
paid work hours, are parents spending less time with their children? This paper
participates in the larger discussion of familial time poverty by examining how
income might influence parental time with children in the particularly challenging
work schedules of shift work. Including shift work and income, two factors that
play a role in parental time with children, enriches the research on familial time
poverty.

Findings from multivariate logistic regression models reveal little
evidence of relationships between shift work, income, or interactions between
shift work and income and the frequency that parents will engage in various
parent-child interactions. What did prove to be significant were relationships
between parental shift work and eating dinner nightly and income and helping
children with homework and/or reading. While these findings demonstrate
impacts of Shift work and income, their influence is hardly ubiquitous.

Moreover, interactions between shift work and income proved unrelated to
all parent-child interactions. When considering that shift work and/or income did
not independently impact most parent-child activities, it is not surprising when an
interaction between shift work and income does not occur. However, even when
shift work or income is Significantly related to two parent-child activities, there

are no interactions between these two variables. Although parental shift work

23

significantly impacts the likelihood of eating dinner with at least one child over
five every night, parental household income is not significantly related. Shift
work decreases the likelihood of parents eating dinner with at least one child
every night equally across the income spectrum. However, parental likelihood of
engaging children in play is more complex because income is only significantly
related to parental likelihood of engaging children in play when Shift work is also
in the model, yet the interaction of shift work and income is not significant.

Overall, Shift work parents are as likely as nonshift work parents to spend
time with their children. In addition, parents in hi gh-income households are no
more likely to spend time with their children than parents in low-income
households. These findings suggest that shift work parents do not appear more
vulnerable to time poverty as they Spend as much time with children as do
nonshift work parents. Similarly, parents with lower household incomes do not
appear more vulnerable to time poverty as they spend as much time with children
as do parents with lower household incomes.

From these results, shift work married parents in dual-earner households
are as likely as nonshift work married parents in dual-earner households to Spend
time with their children. Shift work, a challenging work arrangement, and income
do not largely influence parental time with children. These findings support the
work of Bianchi, Robinson, and Milkie (2006) in that parents make time for their
children regardless of extraneous pressures; even ones as daunting as work

schedules and money.

24

Future Directions for Research

Who is included and at the center of research samples regarding familial
time poverty is an important consideration for future research. This study I
examined only one family formation to avoid emphasis on comparisons between
family formations rather than on work and income. However, research with
emphasis on diverse family formations is needed to explore theses issues in a
broader context. Dual-earner married couples are important to these research
questions but so are dual-earner cohabitating couples, single parents, and so on.
In addition to diverse family formations, research on current data sets is crucial to
furthering our understanding of familial time poverty. This paper used an older
data set for methodological reasons. However, shift work is ever increasing in its
prevalence so more research using more current data is needed.

Moreover, future research may benefit from a dataset that is child-
centered, assessing how much time parents spend with children from the child’s
perspective. This would be especially important when expanding research on
parent-child interactions to family time in general or time spent with various
family members. Given data from a child’s perspective, examining the dyadic
relationship between partners, married and otherwise, would be easier‘to
accomplish. If one partner is working out-of-sync with a child’s schedule, the
other partner may be adapting in ways we cannot understand using this data in
order to accommodate the family. A child’s perspective would also be important
to connect parent-child interactions with child outcomes. Perhaps the quality of

parent-child interactions is different when parents Spend time with one child

25

versus two. Asking individual children how much time they spend in the
company of parents and siblings would give us a better understanding of these
interactions and implications for child outcomes. A child-centered survey would
also be helpful in understanding how the child’s status as a biological, adopted, or
stepchild might affect parent-child interactions.

As in the work by Bianchi, Robinson, and Milkie (2006), examining a
number of activities of parents and children in order to gain a more complete
understanding of time usage, or “time deepening,” is important when teasing out
the many factors in familial time poverty. “Time deepening” is an extensive
project to take on in a research project, but scholars who do get an enriched
picture of family time. This paper hopes to recommend that work schedules and
considerations of income should also be incorporated into meaningful dimensions
of family time and thus into project of “time deepening.”

In addition to suggestions as to who and what is included in research on
familial time poverty, I also suggest the incorporation of mixed methods. The
time families spend together is hard to measure in surveys as people multitask and
attach different meanings to various activities. Daly (2001) finds the meaning of
family time reveals a diverse range of activities that fall under this concept
grounded in person’s experiences of this time. Therefore, the addition of
qualitative methods will give a deeper understanding of how parents and children
negotiate time together.

These suggestions for future research came out of the research in this

paper and in new ideas on the horizon. Familial time poverty is a concern for

26

academics, joumaliSts, politicians, and more importantly families. We need to
understand what is contributing to and alleviating familial time poverty in
addition to how families navigate their day-to-day lives in order to carve out time
for one another. This research is one step further in understanding familial time
poverty. Even in challenging work arrangements and low-incomes, parents figure
out how to spend time with their children. What seems to hold true is parents’
persistence and determination in making time for their children, which is good

news for our children.

27

APPENDIX

28

 

 

Table 4. Logistic Regression Predicting the Likelihood that Parents Eat Dinner with at Least One Child over Four Every N10, ht

 

 

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9
Shift Work Income Shift Work, Shift Work Shift Work
Weekly Income, & & Income & Hours
Covariates Shift Work Income Hours & Hours & Hours Hours Interaction Interaction
Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio
Demographic Covariates
Sex
Male 0.61 *** 0.62*** 0.61*** 074* 075* 073* 075* 075* 075*
Female ---------------------------------------------
Race
White 0.96 0.92 0.97 0.99 0.95 l 0.96 0.96 0.94
Non-White —————————————————————————————————————————————
Education
Less than High School Diploma/GED 1.32 1.22 1.3 1.35 1.26 1.33 1.24 1.24 1.24
High School Diploma/GED —————————————————————————————————————————————
More than High School Diploma/GED 0.87 0.8 0.89 0.88 0.81 0.9 0.82 0.82 0.82
Young Children
Children Four and under at Home 1 03 1 05 1 04 1 04 1.06 1 04 1 O6 1 06 1 05
No Children Four and under at Home ----------------------------------------------
Number of Children
Three or More Children at Home 1.07 1.06 1.07 1.07 1.06 1.07 l 06 1.07 l 07
Two or Less Children at Home ---------------------------------------------
Independent Variables
Work Status
Shift Work 0 6? * 0 58 ** 0.58 ** 0 76 1 07
Non-Shift Work ------------------
Household Income Level
High IIICOIIIC 0 87 0 89 091 0.97 0 92
Middle-High Income 0 96 0.96 0.94 0 97 0 97
Middle—LOW Income 0 96 0-96 0 98 1~03 1
Low Income """"""""""""""""
Number Of Weekly Hours 0.98*** 0.98*** 0.98*** 0.98*** 0.98*** 0.98***
Interaction
Interaction between
Shift Work & Income
Shift Work & High Income 0-61 0-61
Shift Work & Middle-High Income 0-83 083
0.67 0.67

Shift Work & Middle-Low Income

Interaction between 0 98
Shift Work and Weekly Hours .
Constant 1.32 1.55* 1.38 2.79*** 3.68*** 2.89*** 3.8*** 3.62*** 3.62***
N:948; *P < .05. **P < .01. ***P < .001

29

 

 

 

  

Table 7. L00l istic ReUression Predictinn the Likelihood that Parents Engage in Private Talks with at Least One Child over Four Almost Evervdav
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7

 

         
  
 
   

Model 8 Model 9

 

 

 

Shift Work Income Shift Work, Shift Work Shift Work
Weekly Income, & Income & Hours
Covariates Shift Work Income Hours & Hours & Hours & Hours Interaction Interaction
Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio
Demographic Covariates
Sex
Male 0.35*** 0.34*** 0.35*** 0.36*** 0.35 *** 0.36*** 0.35*** 0.35 *** 0.35***
Female —————————————————————————————————————————————
Race
White 0.89 0.88 0.88 0.89 0.89 0.88
Non-White ————————————————————————————————————————
Education
Less than High School Diploma/GED 1.01 1.02 1.03 1.01 1.03 1.04
High School Diploma/GED ---------------------------------------------
More than High School Diploma/GED 1.31 1.31 1.29 1.31 1.31 1.29
Young Children
Children Four and under at Home 0.91 0.89 0.9 0.91 0.89 0.9

NO Children Four and under at Home ——————————————————————————
Number of Children

Three or More Children at Home 0.68 * 0.72 0.68 * 0.68 * 0.72 0.68 *
Two or Less Children at Home ———————————————————————————————————

Independent Variables

Work Status

Shift Work 1.1 1.08
Non—Shift Work ————————————————————————————————————————
Household Income Level

High Income 1.16 1.17

Middle-High Income 1.09 1.09

Middle—Low Income 1.14 1.14

Low Income ------------------------------------------
Number of Weekly Hours 1 1 1

Interaction

Interaction between

Shift Work & Income

Shift Work & High Income

Shift Work & Middle-High Income
Shift Work & Middle—Low Income
Interaction between

Shift Work and Weekly Hours

Constant 1.32 1.28 1.21 1.55 1.48 1.43
N=962; *P < .05. **P < .01. ***P < .001

32

 

Table 8. L00, istic Regression Predicting the Likelihood that Parents Engage at Least One Child over Four in Play or Projects Almost Everyday

Model 1 Model 2 Model 3

Income
Odds Ratio

Shift Work
Odds Ratio

Covariates
Odds Ratio
Demographic Covariates
Sex
Male 0.76 *
Female ———————————————
Race
White 0.89 0.86 0.89
Non—White ---------------
Education
Less than High School Diploma/GED
High School Diploma/GED ———————————————
More than High School Diploma/GED 1.18 1.27 1.23
Young Children
Children Four and under at Home 1.11 1.1 1.12
No Children Four and under Four at Home ----------------
Number of Children
Three or More Children at Home 0.8 0.85 0.79
Two or Less Children at Home ---------------

Independent Variables

Work Status

Shift Work 1.29

Non—Shift Work ---------------
Household Income Level

High Income 0.9
Middle-High Income 1.19
Middle-LOW Income 1.23
Low Income ————————————————
Number of Weekly Hours

Interaction

Interaction between

Shift Work & Income

Shift Work & High Income

Shift Work & Middle—High Income
Shift Work & Middle—Low Income
Interaction between

Shift Work and Weekly Hours

Ix)
t9
I‘d

Constant 1.3 l.
N2979; *P < .05. **P < .01. ***P < .001

33

Model 4

Weekly
Hours
Odds Ratio

0.98 ***

2.86***

Model 5
Shift Work

& Hours
Odds Ratio

255 ***

Odds Ratio

   
  
   
 

 

Model 9
Shift Work
& Hours
Interaction
Odds Ratio

Model 8
Shift Work
& Income
Interaction
Odds Ratio

Model 6
Income

Model 7
Shift Work,
Income,
& Hours
Odds Ratio

& Hours

0.98***

 

2.59***

 

Table 9. Logistic Regression Predicting the Likelihood that Parents Help at Least One Child over Four with Reading or Homework Almost Everyday

 

 

      

 

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9
Shift Work Income Shift Work, Shift Work Shift Work
Weekly Income, & Income & Hours
Covariates Shift Work Income Hours & Hours & Hours & Hours Interaction Interaction
Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio Odds Ratio
Demographic Covariates
Sex
MEIIC 0.37*** 0.37*** 3.7*** 4.05 *** 0.4*** 0.41 *** 0.4*** O.4*** 0.4***
Female _____________________________________________
Race _ *4:
White 0.58** 0.58** 0.57*** 0.59** 0.59** 0.58** 0.57** 0.56** 0.58
Non—White _____________________________________________
Education 0 96 0 97
Less than High School Diploma/GED 1.02 0.98 1.01 1.03 0.99 1.01 0.97 - __'___
High School Diploma/GED """""""""""""""""""""""" 1-1—9— 1 2
More than High School Diploma/GED 1.12 1.14 1.16 1.12 1.14 1.16 1.19 . .
Young Children
0.87
Children Four and under at Home 8.45 0.85 8.45 0.85 0.86 0.86 0.87 0.86 _____
NO Children Four and under at Home ---------------------------------
Number of Children 1 l6 1 14
Three or More Children at Home 1.1 l 1.15 1-1 1-11 1'15 1'1 1‘15 _.___ __'___
Two or Less Children at Home --------------------------
Independent Variables
Work Status
0.96
Shift Work 143 1'41 1'4 1:5. _____
Non-Shift Work ___________________________________
Household Income Level 0 91
High Income 9'29 0'?4 (1)2: (iii 1:26
Middle-High Income 133 ”3 ' 1,. 156 * 1,49
Middle-Low Income L45 1'46 1'5“ __' ________
Low Income """""""""""""""""""""" 0 99
0.99 0.99 -
Number of Weekly Hours 0.99 0.99 0.99
Interaction
Interaction between
Shift Work & Income 1.41
Shift Work & High Income 2.12
Shift Work & Middle—High Income 1.32
Shift Work & Middle—Low Income
Interaction between 1.01
Shift Work and Weekly Hours
4.12 ***
Constant 3.43 *** 3.16 *** 2.95 *** 5.03 *** 4.37 *** 4.29 *** 374*” 0‘89

 

N2840; *P < .05. **P < .01. ***P < .001

34

 

   

 

REFERENCES

Bianchi, Suzanne M., John P. Robinson, and Melissa A. Milkie. 2006. Changing
Rhythms of A merican Family Life. New York: Russell Sage Foundation.

Casper, Lynne M. and Suzanne M. Bianchi. 2002. Continuity and Change in the
American Family. Thousand Oaks, Calif: Sage Publications.

Daly, Kerry J. 2001. “Deconstructing Family Tie: From Ideology to Lived
Experience.” Journal of Marriage and the Family 632283-94.

Garey, A. I. 1999. Weaving and Work and Motherhood. Philadelphia: Temple
University Press.

Gershuny, J. 2000. Changing Times. Work and Leisure in Postindustrial Society.
Oxford: Oxford University Press.

Heyman, .l. 2000. The Widening Gap: Why American ’5 Working Families are in
Jeopardy-and What Can Be Done About It. New York: Basic Books.

Hill Collins, P. (2000). Moving Beyond Gender: Intersectional and Scientific
Knowledge. In Myra Marx Ferre, Judith Lorber, and Beth B. Hess (eds),
Revisioning Gender, Walnut Creek, CA: Altamira Press.

Hochshild, Arlie. 1997. The Time Bind: When Work Becomes Home and Home
Becomes Work. New York: Viking.

Hofferth, Sandra. 1998. Healthy Environments, Healthy Children: Children in
Families. Ann Arbor, MI: Institute for Social Research.

Jacobs, Jerry A. and Kathleen Gerson. 1999. The Time Divide: Work, Family, and
Gender Inequality. Massachusetts: Harvard University Press.

Lareau, Annette. 2003. Unequal Childhoods: Class, Race, and Family Life.
Berkley: University of California Press.

Mestdag, Inge and Jessie Vandeweyer. 2005. “Where Has Family Time Gone? In
Search of Joint Family Activities and the Role of the Family Meal in 1966
and 1999.” Journal of Family History 30:304-23.

Moen, Phyllis. 2008. “It’s Constraints, Not Choices.” Science 319:903-904.

Nock, Steven L. and Paul William Kingston. 1988. “Time with children: The
impact of couples’ work-time commitments.” Social Forces 67:59-85.

35

Noonan, Mary. 2001. “The Impact of Domestic Work on Men's and Women's
Wages.” Journal of Marriage and Family 63:1 134-1 145.

Perry-Jenkins, Maureen. 2004. “The Time and Timing of Work: Unique
Challenges Facing Low- Income Families” In Work-Family Challenges
for Low-Income Parents and Their Children, edited by Ann C. Crouter
and Alan Booth. London: Lawemce Erlbaum Associates.

Phillips, Katherine Ross. 2002. “Parent Work and Child Well-being in Low-
Income Families.” Occasional paper 56. Washington, DC: Urban
Institute.

Pleck, Joseph H. 1997. “Parental Involvement: Levels, Sources, and
Consequences.” In The Role of the Father in Child Development, 3rd ed.,
edited by Michael E. Lamb. New York: John Wiley & Sons.

Presser, Harriet B. 2003. Working in a 24/7 Economy: Challenges for American
Families. New York: Russell Sage Foundation.

Presser, Harriet B. 1989. “Can We Make Time for Children? The Economy, Work
Schedules, and Child Care.” Demography, 26:523-43.

Preston, Samuel H. 1984. “Children and the Elderly: Divergent Paths for
America’s Deendents?” Demography 21:435-57.

Robinson, John P. and Geoffrey Godbey. 1997. Time for Life: The Surprising
Ways Americans Use Their Time. University Park: Pennsylvania State
University Press.

Roy, Kevin M., Carolyn Y. Tubbs, and Linda M. Burton. 2004. “Don’t have no
time: Daily rhythms and the organization of time for low-income
families.” Family relations 53: 168-78.

Rubin, Lillian B. 1994. Families on the Fault Line: America ’3 Working Class
Speaks About the Family, the Economy, Race, and Ethnicity. New York:
Harper Collins Publishers.

Sandberg, John F. and Sandra L. Hofferth. 2001. Changes in children’s time with
parents: United States, 1981-1999. Demography 382423-36.

Sayer, Liana C., Suzanne M. Bianchi, John P. Robinson. 2004. Are Parents
Investing Less in Children? Trends in Mothers’ and Fathers’ Time with

Children. American Journal of Sociology 1 10:1-43.

Schor, Juliet B. 1992. The Overworked American: The Unexpected Decline of
Leisure. New York: Basic Books.

36

Smith, T.E. & Graham, PB. (1995). Socioeconomic stratification in family
research. Journal of Marriage and the Family, 57, 930-940.

Stone, Pamela. 2007. Opting Out? Why Women Really Quit Careers and Head
Home. Berkeley: University of California Press.

Sweet, James, Larry Bumpass, and Vaughn Call. 1988. The Design and Content
of the National Survey of Families and Households. Madison: University
of Wisconsin, Center for Demography and Ecology.

Tubbs, Carolyn Y., Kevin M. Roy, and Linda M. Burton. 2004. “Family ties:

Constructing family time in low-income families.” Family Process 44:77-
91.

U.S. Department of Labor. Bureau of Labor Statistics.l998. “Workers on Flexible
and Shift Schedules.” Labor Force Statistics from the CPS, USDL98-119.
Washington: US. Government Printing Office.

37

1irrigrtttjiititiI