The impact on gasoline consumption from worksite child care facilities

Material Information

The impact on gasoline consumption from worksite child care facilities metro Denver area
Peeples, Re'Jean Anne
Publication Date:
Physical Description:
vi, 118 leaves : forms ; 29 cm

Thesis/Dissertation Information

Master's ( Master of Arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Economics, CU Denver
Degree Disciplines:


Subjects / Keywords:
Employer-supported day care -- Colorado ( lcsh )
Parents -- Employment -- Colorado ( lcsh )
Commuting ( lcsh )
Commuting ( fast )
Employer-supported day care ( fast )
Parents -- Employment ( fast )
Colorado ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Economics.
General Note:
Department of Economics
Statement of Responsibility:
by Re'Jean Anne Peeples.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
28863450 ( OCLC )
LD1190.L53 1993m .P43 ( lcc )

Full Text
Re1 Jean Anne Peeples
B.A., Colorado Women's College, 1981
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
.2 .*


This thesis for the Master of Arts
degree by
Re'Jean Anne Peeples
has been approved for the
Department of
Steven Beckman

Peeples, Re'Jean Anne (M.A., Economics)
The Impact on Gasoline Consumption from Worksite Child
Care Facilities: Denver Area
Thesis directed by Professor Suzanne Helburn
The purpose of this study is to examine potential
gasoline savings of parents who commute to work and have
primary responsibility for transporting their child(ren)
to a child care center. Secondly the study identifies
the basis for the difference between worksite and non-
worksite parents who perceive they have limited employ-
ment opportunties because of child care responsibilities.
The results of this study summarize a survey of parents
from seven Denver area child care centers which provide
on or near worksite care, and parents employed by three
metro Denver companies not having worksite child care.
The survey asked parents to give information on their
transportation arrangements from home to child care to
work for themselves and any secondary transporter.
The most significant result of this research is that
a true worksite center has some potential for saving
parents mileage. Primary transporters using worksite
centers save 1.5 miles per day compared to nonworksite
primary transporters. The excess mileage is considerably
greater for the secondary transporter commuting to child
care located close to the primary transporter's work-

place. The low gasoline savings is attibuted to the
percentage of primary transporters who are not employees
of the workplace and high excess mileage travelled by the
secondary transporter. The gasoline savings is greatest
when the parent can park near the center and walk to
work. However, the finding of this study was the sec-
ondary transporter travels much farther and would save
gasoline with a child care provider located closer to
Parents are also asked if they limited their job
location due to convenience to their child care provider.
While 75% of worksite primary transporters answered, yes,
only 15% of nonworksite primary transporters gave the
same response, indicating that availablity of worksite
child care was important to parents using worksite cen-
ters. A logit regression model was used to examine
parents who did limit their job opportunties to look for
any similarities between the worksite and nonworksite
parents. This analysis revealed the parents in the two
samples are differently motivated when limiting employ-
ment opportunities.
This abstract accurately represents the content of the
candidate's thesis. I recommend its publication.
Suzanne Helburn

1. INTRODUCTION ......................................... 1
Purpose of the Study.................................. 1
Scope of the Study.................................... 4
Arrangement of the Thesis ............................ 9
Family Transportation Patterns and Preferences . 12
Guiliano's Literature Review on
Transportation Needs of Women .................. 13
Florida State University Study of Household
Expenditures on Gasoline .................. 15
Santa Clara County Solo Driver Study .... 17
Hanson and Johnston's Study of Gender
Differences in Work-Trip Length .... 18
Spatial Constraints and Women's Employment in
Toronto.................................... 20
Studies Related to Child Care and Energy............. 24
Toronto Study of Women1s Transportation
Needs................................... . 25
"Big Six" Survey................................ 31
National Child Care Survey, 1990 34
Blacksburg, Virginia Journey to Daycare
Study..................................... 35
Journey to Day Care Centres in Metro Toronto 37
Conclusions.......................................... 39

Procedures........................................... 43
Survey Results ...................................... 49
Characteristics of Respondents ................ 49
Transportation Modes and Preferences .... 50
Distances to Child Care........................ 55
Extra Miles Travelled to Transport Children
to Child Care............................. 58
Does Child Care Create A Perceived Job
Limitation?.................................... 80
Procedures..................................... 81
Results for the nonworksite sample ............ 85
The results for the worksite sample............ 90
Results of the cross-pooling tests ............ 93
2. CONCLUSIONS AND REFLECTIONS ........................... 96
Potential Gasoline Savings .......................... 96
Does Child Care Create Perceived Job Limitations? 100
Recommendations for Further Research ................ 102
A. Logit Regression Results
(Nonworksite variables) ....................... 105
B. Logit Regression Results
(Worksite variables) .......................... 107
C. Cover letter and Worksite Survey
Questionaire .................................. 109
D. Nonworksite Survey Questionaire ............... 115
BIBILIOGRAPHY ............................................ 119

Purpose of the Study
The purpose of this study is to examine the
potential gasoline savings of parents who commute to work
and have primary responsibility for transporting their
child(ren) to a child care center. The fundamental
question asked is whether worksite or near worksite child
care arrangements can save commuting miles and time for
parents. The focus for answering this question is on the
parents routine commute from home to child care to work.
Previous studies have examined the location of child care
centers in communities and the behavior patterns of
commuting parents, but few have looked at the impact of
the parents daily commute as it relates to gasoline
consumption. Parents, who are primarily responsible for
dropping off and picking up their child(ren), have been

shown to use their automobile with more frequency to get
to work.
Over the past twenty years the percentage of adult
females working in the labor force has steadily risen to
57 percent (U.S. Bureau of the Census, 1992). Of this
group employed wives account for thirty one million (U.S.
Bureau of the Census, 1992) This influx of working women
has significantly changed the workplace because many of
these women are under 35 years old and have preschool
children. The U.S. Bureau of the Census (1992) reports
the percentage of women in the labor force between 16 and
34 years old is 45%.
The rewards for women working outside of the home
are obvious, but increasing labor force participation of
women has created many new issues. One related
phenomenon is the single parent. In 1979 one out of
every seven families was headed by a women who was a
single parent (Wallace, 1982). Analysis of more recent
trends reveals that women workers will continue to be an
important factor in the work force. In 1990, 8% of all
single women and 24% of married women in the labor force
have children under age 6 (U.S. Bureau of the Census,

1992). The U.S. Bureau of the Census (1992) has
forecasted a 5% increase in the female labor force
participation by 2000.
The participation rate of men- in the labor force
from 1970 to 1990 declined by 3.6%, while the
participation rate of women increased by 14% (U.S. Bureau
of the Census, 1992). While 25 million additional women
have entered the civilian labor force in the last 20
years (1970 1990) the number of men has risen by only
17 million (U.S. Bureau of the Census, 1992). During
this period, labor force participation rate for married
women between the ages 25 and 34 increased 31% (U.S.
Bureau of the Census, 1992).
The idea that a large portion of working women are
young and in their child bearing years still have primary
responsibility for making child care arrangements, even
in the 1990s. Child care for married, middle-class
mothers implies that the family's daily routine includes
dropping off and picking up children at a child care
provider as an add-on trip to the workplace. Therefore,
the issue of greater gasoline consumption, is an

important issue today given the significant impact of
women in the work force.
Scope of the Study
The results of this child care study summarize a
survey of parents from seven Denver area child care
centers which provide on or near worksite care, and par-
ents employed by three metro Denver companies not having
worksite child care. The survey asked parents to give
information on their transportation arrangements from
home to child care to work for themselves and any
secondary transporter.
The following definitions are' used in this study.
WORKSITE CHILD CARE is any center located within
.2 miles from work (within two city blocks or a
large parking lot and an easy walk). Three centers
were clearly within .2 miles of the associated
employer's workplace. The term worksite care in
this paper also refers to all seven centers in the
study representing near or worksite care.
NEAR WORKSITE CHILD CARE is any center located
within one mile from one or more large employers,
but not meeting the worksite definition. One mile
was considered a reasonable distance for some of the
centers in this study because they were not spon-
sored by one employer. Four centers meet this
NONWORKSITE CHILD CARE refers to centers used by
employees of three Denver area employers who do not
have access to worksite or near worksite child care.

PRIMARY TRANSPORTER is the individual who most
often drops off and/or picks up the child at the
child care center, self-classified by the question-
aire respondents.
SECONDARY TRANSPORTER is the second individual who
drops off and/or picks up the child at the child
care center. The secondary transporter of these
families travels to the same center as the prima-
ry transporter.
ALTERNATE CHILD CARE relates to parents using
worksite or near worksite child care who identified
another provider which they would use if the current
provider were not available to them.
DOWNTOWN DENVER includes the central business
district and the immediately adjacent Auraria higher
education campus.
SUBURBAN DENVER includes portions of the city and
county of Denver that are essentially suburban in
their neighborhood characteristics as well as the
counties adjacent to Denver.
EXCESS MILEAGE (sometimes labelled as "differ-
ence") is the difference in distance in miles from
home to child care to work minus home to work. This
mileage was estimated using gross streets reported
by the respondent and computing the mileage for each
trip using a large scale map and map measuring
This study explains the demographic make up of
parents using worksite or near worksite child care
arrangements, and parents without worksite child care
available to them. The survey includes information
regarding: the number of days per week the primary and
secondary transporters dropped off and picked up the

child(ren) at the provider; the ages of the children in
the family; the number of providers used by the family;
modes of transportation for commuting to the child care
center and work; location preference of the child care
provider relative to home or work;- limitations of job
opportunities because of the burden of child care
responsibility; number of errands performed each week
during the commute from home to work, and willingness to
use public transportation.
The worksite surveys were distributed to parents as
they dropped off their child(ren) at the child care
center in the morning. They were asked to complete the
survey and return it to a box placed in the lobby of each
center. Extra surveys were left in the lobby area for
parents not handed a copy during the time of
distribution. The centers surveyed were: Auraria,
Denver Bulk Mail Facility, Denver Place Montessori,
Fitzsimmons Army Hospital, Porter Hospital, St. Mary's
Academy, and Swedish Hospital. Overall, the response
rate of worksite parents was 40%. Of those responding in
the worksite sample, primary transporters represented a
fairly homogeneous group. However, there were

differences between worksite and nonworksite parents; the
latter had less education, lower incomes, and were
slightly older.
Comparison of practices of parents with worksite
care and those waiting to enroll their children was not
possible because the directors of these centers were
uncomfortable with allowing us to contact parents on
their waiting lists. Had we been able to survey the
parents on the waiting list a better comparison could
have been made of the worksite and nonworksite groups.
Instead, a second survey was conducted with employees of
three metro Denver area employers who were willing to
assist us in identifying employees with young children
and distributing of the questionnaire. The nonworksite
employers were: Colorado National Bankshares, Manville
Corporation, and US WEST. The survey was randomly made
of the personnel given the quickest method of
distribution used by that firm. These nonworksite
parents were used as the comparison group or control
group for this study.
The most significant result of this research is that
a true worksite center has some potential for saving

parents mileage. Primary transporters using worksite
centers save 1.5 miles per day compared to nonworksite
primary transporters. However, the excess mileage is
considerably greater for the secondary transporter
commuting to child care located close to the primary
transporter's workplace. Also, single parents had
greater excess mileage than married families. The low
gasoline savings is attibuted to the percentage of
primary transporters who are not employees of the
workplace. One center, defined as a worksite, had 20% of
the primary transporters not working on the site.
Parents definitely used their car commuting to child
care. (92% of the worksite parents and 85% of the
nonworksite parents.) Both single and married parents
used their automobile for convenience and saving time.
Married parents in the higher income levels also
indicated they would consider public transportation and a
child care center located at a park-n-ride. Most park-n-
ride facilities are located near large suburban housing
areas and this may explain married, parents response to
public transportation.

One of the questions asked on the survey was "Have
you limited your job location due to convenience to your
child care provider?" The purpose of this question was
to determine whether parents do or do not limit job
opportunties when worksite child care is made available.
A logit regression model was used to examine parents who
did limit their job opportunties and find any
similarities between the worksite and nonworksite
parents. This analysis revealed the two samples were
differently motivated when limiting employment
Arrangement of the Thesis
The following chapters are organized as follows.
Chapter 2 is a discussion of studies related to child
care and the commuting behavior of parents and families.
The first part of the chapter reviews articles concerning
urban transportation, urban planning, and gender
differences in commuting and serves as a precursor the
next section. "Studies related to child care and energy"
is the second section and examines recent studies on the
subject of mileage to child care. Chapter 3 covers the
findings from this study by first identifying the

procedures. The survey results are separated into four
areas: characteristics of respondents, transportation
modes and preferences, distances to child care, and extra
miles travelled to transport children to child care. The
last section in chapter 3 discusses of test results from
a question on the survey regarding parents perceptions
about the effect of child care location on their choice
of jobs. The final chapter summarizes the results of the

Very few studies have been conducted on child care
and its impact on energy consumption. However family
transportation patterns and needs have been thoroughly
studied. The research dates from the 1970's when the
federal government expanded its role in developing
effective public transit systems following the passage of
the Urban Mass Transportation Act of 1964. These studies
on family (particularly women's) commuting patterns
provide the broader context for understanding decisions
about transporting children to child care. Therefore,
the first section in this chapter titled "Family
Transportation Patterns and Preferences", briefly reviews
some of the earlier studies which address general issues
of women's transportation needs.
The second section, "Studies Related to Child Care
and Energy", describes five previous studies which

collected data on the transportation of children to child
care. Four of the studies specifically estimate extra
commuting mileage of working parents using child care.
As precursors to the present study, both the methodology
used and the results of all these studies are of
Family Transportation Patterns and Preferences
This part of the literature review looks at several
studies which address transportation patterns and gender
differences. The Guiliano (1979) study is a literature
review written in the late 1970's concerning women's
transportation needs. The Florida State University study
(Rhodes, 1982) looks at the proportion of household
expenditure on gasoline for two different geographical
regions. The Santa Clara (Crain, 1984) study was
commissioned by the local government to identify common
characteristics of drivers commuting alone to work. The
last two studies (Hanson and Johnston, 1985; Brooker-
Gross and Maraffa, 1985) discuss gender differences in
distance to work.

Guiliano's Literature Review on
Transportation Needs of Women
Guiliano reviewed literature on women's use of
public transportation (Guiliano, 1979), focusing on the
travel behavior of women and their demand for public
transportation resulting from their projected increased
participation in the labor force. Based on this
literature review, Guiliano concluded that it is critical
for public transit systems to become more demand
responsive to their ridership. She discussed the
alternatives and identified single women with children as
particularly dependent on public transportation due to
their lower incomes and greater financial
responsibilities, arguing that limited transportation
options reduced their ability to work and find suitable
child care arrangements. The following four studies
reviewed by Guiliano support her arguments.
The most dramatic demonstration of public transit
needs of women is provided in a 1974 study of New York
State, (excluding New York City) by the N.Y. Department
of Transportation. This study revealed that women
outnumbered men by 4 to 1 in their use of public

transportation and comprised 94% of the ridership on
nonwork related bus trips.
A 1973 study in Seattle by the Metro Transit agency,
surveying riders' origins and destinations, reported that
67% of the trips involving public transportation, within
the city, were made by women. Of the trips made outside
the city limits 61% were made by women. Ridership was a
function of income, household size, age, automobile
availability, and the number of issued driver's licenses.
The study indicated that public transit systems would be
used the most by women (Guiliano, 1979).
Next a study performed in Davenport, Iowa and
Hicksville, New York concerning shared taxi service is
reviewed. Both cities offer regular bus routes as well
as shared taxi service. This study revealed some similar
characteristics to the others in her review such as,
family income level, number of issued drivers licenses,
and the individual's employment status as being the most
significant determinants of demand for public
transportation (Guiliano, 1979).

In conclusion Guiliano demonstrates the need for
further research in the area of women's issues related to
travel behavior.
Florida State University Study of
Household Expenditures on Gasoline
In 1980 1981 Rhodes studied the relationships
between household expenditures of gasoline and the
characteristics of households in Florida (Rhodes, 1982).
Expenditure on gasoline was divided into gas consumed in
the journey to work, and for other purposes. Household
characteristics included the presence of children,
occupation of the head of the house, and price of the
home. Rhodes first examined the relationship between
household structure and gasoline expenditure for a sample
of 1,082 families. Then he resurveyed the same sample
group a year later and tested for the stability of each
household's characteristics and travel expenditures.
Finally, he compared his findings to an identical survey
performed in Rhode Island. The most significant results,
for purposes required in this study, were:
Household travel was directly related to the stage
in life of the family. Households with young

parents and young children tend to consume more
Income was not a good predictor of gasoline usage,
but household participation in the labor force did
explain gasoline usage.
The relationship between gasoline usage and
household structure in Florida and Rhode Island were
Households in both surveyed samples averaged
about 9% expenditure of annual income on
1/4 of the commuters in Florida and 1/5 of the
commuters in Rhode Island shared their ride to
work with a member of their own family.
Half of the workers in Florida lived within one
mile of a bus stop and 7 out of 10 in Rhode
Island were within a 15 minute walk, but only
2% of the labor force in Florida and 4% in
Rhode Island use public transportation.
Occupation of family members had only a modest
relationship to gasoline consumption.
In summary, this study demonstrated that people

basically prefer to drive, and do not change their travel
patterns despite changes in the price of gasoline or
their spatial proximity to public transportation.
Santa Clara County Solo Driver Study
This 1983 study grew out of Santa Clara County,
California government's desire to develop a ridesharing
program (Crain and Assoc., 1984) because predicted
employment growth (50% over the next ten years) would not
be matched by highway expansion or bus service. Small
groups of people representative of the county communities
were interviewed to identify commuter perceptions of
ridesharing. Of the twelve types of solo drivers
identified, the three most common types held the
following attitudes: 1) time is everything; 2) they
might consider taking the bus; 3) they did not think that
carpooling could work. The most common perceptions of
both males and females were: traffic congestion is
getting worse; employees prefer ride sharing over public
transportation when a firm offered some kind of benefits
to its employees; people's schedules are incompatible for
ridesharing. The study describes the attitudes and
perceptions of solo drivers in categories according to

age, sex, occupation, income, employment, and ridesharing
Hanson and Johnston's Study of
Gender Differences in Work-Trip Length
Hanson and Johnston's (1985) sought to explain why
women in metropolitan areas work closer to home than men,
and identify any patterns between the home to work
commute. In this study they found that men on average
travelled 2 miles further each day than women and an
average of 3 minutes longer. Findings were consistent
with (Michelson, 1985; McGinnis, 1978) women's lower
incomes, shorter job tenure, shorter working hours, and
greater household responsibility (Hanson and Johnston,
1985). The authors identified the following influences
on women's commuting distance:
Income. A higher percentage of women (73%) earned
less than $10,000 per year and averaged 5 miles from
home to work, while 65% of the men earning higher
incomes, commuted over 8 miles to work. Wages were
a significant predictor because someone with low
wages cannot afford a long, costly journey to work.
Labor Force Characteristics. Women and men make up
different proportions of major occupational

categories. Part-time work for women was not a
factor determining distance to work, but job type
was more important. Females with clerical, sales,
and service-oriented jobs had the closest commutes
from home to work.
Household responsibilities. Women were much more
responsible than men for the division of household
duties. Women with the greatest demand on their
time travelled the shortest distance. The number of
children in the family was an important factor in
determining the wife's demand at home.
Mobility Constraints. Fewer transportation options
constrains mobility, and further limits job oppor-
tunities. Public transportation requires more of
the user's travel time, and is an important
consideration for a parent facing time constraints.
Public transportation was used 20% by women as
compared to 13% of the men.
Spatial factors. The convenience of job
opportunities affects the rate of labor force
participation. Since affordable quality housing and
job opportunities are not equally distributed in

urban areas, female heads of households tend to
locate nearer to central business districts where
they are close to female oriented jobs.
This study has been the basis for several later
studies (Brooker-Gross and Maraffa, 1985; Rutherford and
Wekerle, 1988) concerning gender differences.
Spatial Constraints and
Women's Employment in Toronto
Rutherford and Wekerle (1988) studied the gender
differences in transportation use of men and women in
Toronto, Canada. Their results showed that regardless of
household type, women were two to three times more likely
than men to use public transportation from home to work.
Classifying Toronto commuters into three categories -
captive, choice, and automobile riders the authors
found one in five women to be a captive rider with no
alternative mode of transportation. The choice rider was
someone who used part of the public transportation system
because it provides speed, comfort and economy.
As in the Hanson and Johnston study (1985) women
worked closer to home, travelled more slowly to work, and
had lower incomes. As the distance to work increased,
fewer numbers of women commuted in each category. The

number of women traveling less than three miles was 57%
compared to men (43%), and more men were found to travel
distances 20 miles or more (65%) compared to 36% of the
One interesting conclusion of this study was the
analysis of additional income by those who travelled
1-20 miles to work. In general, the authors found men
to earn more additional income than women. Women with
access to an automobile gained $5,500 per year by
travelling 20 miles to work, whereas men's incomes
increased on average, by $7,000 per year. Male choice
riders showed a much larger increase in added income of
$13,500 while women earned only $1,100 per year more.
Women choice riders quite possibly choose the portion of
public transportation offering speed, convenience, and
economic gain. Captive riders are more often women with
a $4,500 increase per year, while men in this category
have economic gains similar to the choice male rider,
$13,000. The authors are not clear about the reasons for
these differences in income between men and women, other
than to argue that female occupations tend to pay the
same regardless of location.

Nonmetropolitan University Employees
Commuting Distance
Brooker-Gross and Maraffa (1985) reported on gender
differences in travel to work in Blackburg, Virginia, a
nonmetropolitan college community, to find if
explanations commonly found in metropolitan areas applied
to rural areas. The authors found women's worktrips in
this nonmetropolitan area are longer than men. In this
study the authors surveyed Virginia Polytechnic Institute
(VPI) faculty and staff, to answer three questions: 1)
how do women's worktrips compare to mens; 2) can
commuting differences between men and women be explained
by the same factors found to be important in metropolitan
such as income and domestic roles; 3) does spatial
structure of nonmetropolitan areas play a role in gender
based differences in the distance to work. Their
findings were:
85% of the respondents lived within 10 miles (one-
way) of the university.
Women's worktrips were 3.7 miles longer than men's
worktrips, a constrast to metropolitan studies
(Hanson and Johnston, 1985). This difference is due

to the spatial structure of housing and employment,
women seek jobs after locating their residence, and
female jobs paying lower wages so they cannot afford
housing in town.
67% of male worktrips and only 39% of female
worktrips were less than 3 miles in distance. The
explanation for this difference is that male faculty
can afford housing in town.
Equal numbers of men and women travelled distances
ranging between 3 to 10 miles. No gender difference
exists in this case because of the greater variety
in housing opportunities further from town.
30% of the female respondents and only 10% of the
males lived further than 10 miles from the
university. This is partly because the women in
this study were more likely to be single than were
men, and were three times more likely to have mid to
lower incomes (<$15,000). Also, in this study
married women were found to travel longer distances
to work than single women, regardless of income,
possibly because their husbands worked off campus.
Unlike the metropolitan studies, in this study

married women travelled longer distances than men.
Evidently, quality of housing rather than the
location of the wife's employment was an important
determinant of travel time.
Some of the issues concerning housing and employment
opportunties are shown here to exist regardless of
Studies Related to Child Care and Energy
This section is reviews studies of the distance from
home to child care. The Toronto study (Michelson, 1983)
was conducted to examine the social patterns created by
women in the workforce. The "Big Six" survey (MacBrine,
1988) by the State of California establish baseline data
on state employees1 commuting patterns. The National
Child Care Survey (Hofferth, Brayfield, Deich, Holcomb,
1990) has a much broader purpose than measuring mileages
to child care, but briefly discusses distance from home
to child care to workplace. The Brooker-Gross and Myer-
Jones study (1990) looks at the time-space geography and
travel patterns of parents using child care in a small,

rural college community. Truelove's study, "Journey to
Day Care Centre in Metro Toronto", (1990) broadens the
picture by looking at commuting distance given three
different types of child care centers.
Toronto Study of Women's Transportation Needs
William Michelson (1983) carried out a detailed
study of commuting patterns of parents comparing the
transportation needs of women in Canada and the United
States in his study, "The Impact of Changing Women's
Roles on Transportation Needs and Usage" (Michelson,
1983). This study focuses on the following specific
travel issues:
The effect of new roles of women on travel patterns
and demands.
Differences in travel patterns of men and women.
Factors affecting conditions of women's travel.
Personal consequences stemming from women's travel
Recommendations of designing facilities to meet the
new needs of women.
The Canadian data were based on a 1980 survey of 538
families in metro Toronto representing a cross section of

the population. The daily patterns of time use and
constraints faced by women were documented, particularly
the changes in travel behavior attributable to women's
participation in the work force and the increase in
female heads of households with children. The U.S. data
were taken from a 1976 survey of families from five
counties conducted by the Southern California Association
of Governments and California Department of
Transportation. Michelson selected only Orange County to
use as a comparison to the Toronto study. Of the 664
households in Orange County, he identified 83 as meeting
characteristics similar to the Toronto sample. Although
the Orange County data were very limited, expected
differences in travel and modes of transportation were
In the Toronto survey travel time and mileages were
positively correlated with the number of hours a woman
works. Women with full-time jobs spent almost twice as
much time travelling on weekdays than did full-time
housewives, and women with part-time employment fell in-
between the two groups. A small difference exists
between the married and single employed women surveyed

because the single women spent more time travelling as
opposed to single non-working women. This is explained
by the fact that poorer single women stay home.
Women's travel needs in Toronto were found to be
closely related to the existence of children in the home.
Regardless of marital status of the women, travel time
increased as children reached ages 10 to 14 years old.
The travel time of employed mothers with very young
children requiring child care was also high. Overall,
Canadian women took their children on a third of their
travel trips and the bulk of these trips were related to
shopping, recreation, and social activities. Of the
trips ending at the mother's work place, about 95% were
made without children.
In Toronto use of public transportation was impor-
tant among women travelling to work:
Women commuting to work made about 25% of the trips
using public transportation.
Many women working full time linked their commuting
with household errands. A third of the trips from
home to work for both full-time and part-time
employed women, were linked with other necessary

trips and the mode of transportation used did not
make a difference.
Travel time to work was higher for women using a
child care center as opposed to home care. Mothers with
full-time jobs using a child care center spent an average
of 25 minutes per day travelling; mothers with child care
in their own home spent an average of 19 minutes
travelling; those using grandparents for child care
travelled 11 minutes. Some of this travel was on foot.
Mothers were primarily responsible for transporting
children to child care. They accompanied their children
to the child care centers 55.6% of the time with a
husband present on the trip only 14.8% of the time. Only
13% of the husbands accompanied children to child care
center alone, and 11% of the children made the trip
Michelson estimated the following mileages:
The average distance from home to work was 4.0 miles
for children of all ages, while the distance from
home, to child care to the parent's place of work
was 4.3 miles (.3 miles a day difference).

For parents with children 0 to 3 years old the
average mileage from home to work was 3.3 miles and
the distance from home to child care to work was
5.53 miles, (2.23 mile a day difference.)1
Children of single mothers spent the most amount of time
per day travelling. Michelson argued that additional
travel causes an increase in travel tension, usually
experienced by women, and that tension increases with
Turning to the Orange County, California data as a
comparison, this study differs from the Toronto sample in
several ways. First, only 14 out of the 83 households
were headed by single mothers. Second, most Orange
County female heads of household were employed full-time,
while the Toronto sample had more part-time workers and
full-time housewives. Third, most women in the Orange
County group had access to automobiles. Ninety-nine
percent of their households owned one or more cars, and
76% owned two or more cars. While 89% of the Orange
County women held a driver's license and had access to a
'since a straight-line approach to measuring distances was
ud in this study the mileages may be underestimated.

car, only 62% of the women in the Toronto sample held a
driver1s license.
As expected, differences in results between Toronto
and Orange County reflect differences in the availability
of public transportation and the spatial layout of these
two cities. Toronto is a city with an excellent public
transportation system, while Orange County, a Los Angeles
suburb, has an inadequate system. The mean distance from
home to work for women in Orange County was 16.7 miles
(more than double the 7.5 miles travelled by men), and
averaged 36 minutes to work (males averaged 27 minutes).
A considerably greater distance exists for women in
Orange County than for women in Toronto.
The basic conclusions which can be drawn from this
study are:
Women are primarily responsible for meeting the
transportation needs of the children in the
Given the time-budget constraint of hours in a day,
a greater amount of travel tension is created for
women to complete the necessary daily activities.

Single women with children and employed full-time
spend the greatest amount of time in daily travel,
followed by married women working full-time.
Women employed full time travel longer distances
than men in their daily travel because the greatest
numbers of jobs are usually centrally located and
child care providers are not.
Children in child care centers spend 6 to 20 minutes
longer per day commuting than children in family
child care.
The development of faster public transportation
systems has a significant influence on women's
choices and the need for access to a car. However,
choices might also reflect car availability.
"Big Six" Survey
The "Big Six" survey conducted by the California
Department of General Services in 1988 (MacBrine, 1988),
was designed to provide information to help implement the
Governor's 1988 "Executive Order of Traffic Improvement"
reducing peak hour solo driver traffic ten percent by
January, 1990. The main objective of this study was to
develop baseline data on how State employees working in

congested urbanized areas commute to their jobs. The
survey was designed to identify some of the reasons
behind commuters choices of transportation during peak
and non-peak hours, including the relationship between
transportation choices and the use of child care. The
survey respondents were full-time employees of Califor-
nia's six largest state agencies located in various
heavily populated urban areas of the state. Five
important findings resulted from this telephone survey:
Women are predominately the transporters of children
to child care regardless of the child's age.
Parents with child care responsibility are most
likely to use their car to transport children, and
' £
mass transit was the least likely choice among
Flextime was more commonly used by employees who
transport their children to child care than by other
employees who do not have this responsibility.
Employees added an average of four excess miles to
their one way commute to work when the child was
zero to four years old and 3.6 miles when six to
eleven years old. However, 46% of the parents with

children from infants to preschoolers added only one
mile to a one way commute almost every day.
Respondents with children zero to four use child
care arrangements without any preference to home or
work if the commute is under ten miles. However,
78% parents of children six to eleven preferred
their child to be closer to home regardless of the
parent's or the child's commute.
Although this study provided pathbreaking findings
there are some limitations to the results. First, the
respondents were interviewed over the phone and their
answers put directly into a database, leaving no paper
trail to correct for possible errors created by the
interviewer. Second, mileage estimates were based on the
respondents' estimates and no explicit information as to
the location of their home, child care provider, or
worksite was given in order to verify independently the
respondent's mileage estimates. The lack of this
information would result in an inaccurate determination
of the margin of error. The Department of General
Services reported the margin of error to be plus or minus
one mile (Doyle, 1989). Third, respondents were not

asked whether or not they shared transportation of their
children to child care with another transporter.
In conclusion, this report has helped pave the way
for studying excess mileage created by child care in this
decade. The results indicate the significant impact
child care can create on energy useage and added traffic
congestion by not having better urban planning.
National Child Care Survey, 1990
The 1990 National Child Care Survey provides a
comprehensive picture of family choices related to child
care arrangements, parental expenditures for child care,
perceptions of alternative forms of child care, and
issues related to employers and child care services
(Hofferth, Brayfield, Deich, Holcomb, 1991). While not
much information is reported concerning travel arrange-
ments and its impact on family responsibilities, the
study reports distances travelled by parents who use
center based child care, family day care, and in-home
Respondents were asked their travel time to both
their primary provider as well as any alternate provider.
Perceptions of time for parents who use some type of

child care arrangement was very similar to those of
parents who stay home with their children and use child
care only occasionally. Both groups (parents both work,
one parent works) reported that family day care providers
were located closer to home; parents using child care
centers (74%) reported their providers to be within 10
minutes of home.
No significant differences in distance to family day
care providers appeared based on parents income level.
However, families using center-based care, where only one
or both parents works, there were significant travel time
differences by family income. For parents using center-
based care, 66% of those earning more than $50,000 were
within 10 minutes of a provider while only 49% of the
parents with incomes less than $15,000 were within the
same distance. The reasoning would appear to be that
child care providers are more commonly located in
neighborhoods where parents can afford higher prices.
Blacksburg, Virginia Journey to Daycare Study
Brooker-Gross and Myers-Jones, 1990 conducted a
study of parent trips to child care and time-space
geography of three local centers used by faculty and

staff at Virginia Polytech Institute (VIP), in
Blacksburg, Virginia. The authors chose the three
centers with the largest enrollments. One center was
centrally located and the other two were at each end of
Parents with higher incomes were found to live close
to the major employer (VIP). Overall, trips to child
care were short with 75% of the respondents living in
this small nonmetropolitan town. Of the parents using
these three centers 58% were VIP employees. Mothers were
predominately responsible for dropping off (48%) and
picking up children (55%), a smaller percentage (23%) of
parents jointly shared responsibility. When respondents
were asked why they chose the center they used, most
parents answered quality of care. Although educational
value and cost were also ranked as important.
Together these centers make up the biggest portion
of child care in Blacksburg. The difference in distance
to these centers is small, but over one-third of the
parents indicated distance as an important factor in
choosing a center. Parents using the centrally located
center cited distance from home as important more often

(46%) while parents using the center farthest from the
center of town ranked distance the least often, 20%.
The average excess mileage to transport children to
the centers was less than 2 miles, and maximum added
mileage was 6 miles. A cross-town trip in this small
town is between 4 and 6 miles. Higher income families
disproportionately used the three centers, and were found
to make shorter trips to work and child care.
The authors did not compare their differences
between centers, but indicate that distances travelled is
important even in a small town. Households appear not to
travel out of their way to use child care.
Journey to Day Care Centres
in Metro Toronto
A study by Truelove (1990) captured differences in
transportation of children to child care based on income
status of the families. Truelove sampled 1,619 children
in the metro Toronto area and their use of child care
centers. The children were categorized into two groups
based on how their child care fees were paid, full-fee or
subsidized. Providers were categorized by type of center
ownership: non-profit, municipal government, and
commercial. Similarities and differences in choice of

location were examined as well as the parent1s travel
For all three types of centers combined, 80% of the
children travelled an average distance of less than 3
miles, and 22% of this group travelled less then .3
miles. The children with subsidized child care travelled
shorter distances than children with parents paying the
full fee. The average distance from home to center for a
subsidized child was a 1.8 miles (s=2.4 miles). Children
attending municipal government centers, most of whom were
subsidized, tended to travel the shortest distance of the
three categories of centers, an average of 1.3 miles (37%
commute less than .3 miles). There was more variation in
distance travelled by families to non profit centers but
68% travelled less than 3 miles. The key explanation for
the low mileage for these parents is that most of the
centers are community based, although some parents did
travel out of their community to other non profit centers
they believe to be higher quality. Parents using
commercial centers travelled the greatest distance from
home to child care provider. This longer commute implies

that this type of center is created for suburban families
with higher incomes who travel by car to work.
Truelove also looks at the distance between the
center located nearest to the family home and the center
actually attended. Almost half of the surveyed families
(45%) chose the center nearest to their home. Subsidized
children were most likely to attend the center nearest to
home. The municipal centers showed the highest
percentage (61%) of the children attending their nearest
child care center.
The author further looks at whether "full fee" chil-
dren attending lower priced centers did so to avoid
higher priced centers nearby, but no significant
difference in distance was found. Only 8% of parents
used centers with fees $10 cheaper than the closest
center to their home. Thus, parents did not seem to be
motivated to travel further for a lower priced child care
The Guiliano (1979), Florida State (Rhodes, 1982),
Santa Clara (Crain, 1984), and Toronto (Michelson, 1983)
studies were all very early studies conducted to

determine the transportation needs of women as they began
entering the labor force in larger, numbers. However,
child care transportation needs were not issues clearly
foreseen in the mid 1970s. What can be concluded from
the early studies as well as the "Big Six," (MacBrine,
1988) National Child Care Surveys (Hofferth, Brayfield,
Deich, Holcomb, 1991), Truelove (1990), Brooker-Gross and
Myers-Jones (1990) are the following:
The extra miles in commuting from home to child care
to work create time constraints for the parent.
This time constraint is an important consideration
because of the stress it creates in commuting.
These studies do not show much evidence in the U.S.
for the use of public transit. Many U.S. cities
have been designed to accommodate the use of
automobiles to get people to their destinations.
The result is generations of adults growing up with
a preference to drive rather than use other forms of
transportation in their daily commute. People today
are interested in speed and convenience and the
automobile is the answer to these needs.

The "Big Six" study showed that women are
predominantly responsible for transporting children
from home to child care. The Michelson/ Brooker-
Gross and Myers-Jones, and Hanson and Johnston
studies also show that women are mainly responsible
for performing the daily routine activities of the
household. Since 57% of the women in the U.S. are
in the labor force and mostly responsible for
handling the transportation needs of their children
along with household needs, the issue of stress
related to time constraints is an important issue.
Brooker-Gross and Marrafa (1985), Truelove (1990),
Hanson and Johnston (1985), Rutherford and Wekerle
(1988) studies identify the linkages between gender,
work time travel distance, and income. These
authors all appear to imply that the real gains need
to be to women's incomes, and improvement of public
transportation will not increase income levels but
can change employment opportunities. The work
opportunties created by the spatial division of
labor often causes the mismatching and underem-
ployment of women.

Although the data are limited and not conclusive,
these studies in combination suggest that parents do
travel extra miles to child care and hence worksite child
care could save miles and gasoline. The amount of excess
miles is highly variable and only partially explained by
income, marital and job status. Excess miles are more
often incurred by women, who are primarily responsible
for getting a child(ren) to the child care provider.

The major goal of this study is to estimate the
average mileage parents of preschool children travel from
home to child care, to work, and the additional gasoline
consumption created by add-on trips to a child care
provider. To answer this and related guestions survey
data were collected through questionnaires completed by
parents of children in Denver area centers considered to
be worksite or near-site child care and from a sample of
Denver area parents not using worksite or near worksite
care. The group of parents not using worksite child care
were employed by three Denver area employers and used
various types of child care arrangements.
The sample of worksite or near worksite child care
parents used the following child care centers: Auraria,
Fitzsimons Hospital, Porter Hospital, Swedish Hospital,

Denver Place Montessori, St. Mary's Place, and the Denver
Bulk Mail Center. We distributed surveys to all parents
dropping off their children before 9:00 a.m. Parents
were handed a questionnaire at each center while
delivering their child on a particular day between 0630
to 0900. Additional questionnaires equal to the
remaining number of families were left on a table in the
lobby near the entrance for parents to pickup. Parents
were asked to complete and return the survey to a box
placed in the center lobby. The sample of 302 parents
represented approximately 40% of the distributed surveys,
including those left in the lobby to be picked up by
parents arriving later. The two centers with lower
response rates (25%) had larger numbers of people working
non-standard hours; hence, a larger fraction of the
questionnaires were not distributed directly but by the
center's director. Thus, this study samples primarily
parents who drop off their children between 0630 and
0900. Some bias may exist because of the absence of
parents who work non standard hours or part-time,
although the parents with non standard hours who did

respond had virtually the same excess miles as those
starting work between 0630 and 0900.
TABLE I. Respondent Response Rates by Center
Number of Surveys Distributed Number of Respondents Percentage Returned
Fitzsimons 113 29 26%
Porter 114 48 42
Post Office 63 25 40
St. Mary's 150 60 40
Swedish 150 39 26
Auraria 150 53 35
Denver Place 110 43 39
A comparison group of parents without worksite or
near worksite child care available were also surveyed.
This sample of 167 was obtained through the cooperation
of three large Denver area employers: US WEST, Colorado
National Banks, and Manville Corporation. Different
methods were used to distribute the surveys at each
company in this sample group, based on availability to
the company of information as to which employees had
children under the age of five. One company used their
electronic mail system to collect names and then mailed a

copy of the survey to each parent. Two used their
medical records to identify employees with children
meeting the age requirement.
Parents using the worksite centers completed a
slightly different questionaire for the survey than
parents in the comparison group. The worksite
questionaire included 34 questions, the nonworksite
questionaire included 32. The survey used for the
worksites parent included three additional questions
relating to what the parent would do if worksite child
care were not available. The nonworksite survey had an
extra question concerning whether parents would use
worksite child care if it were available to them. The
questions for both versions of the survey are reproduced
in Appendix C and D.
The respondents were asked to provide information
for the primary parent transporting the child(ren) to a
child care center and for a second person in the
household who shares this responsibility. The questions
asked for:
The number and ages of children in the household
being transported.

The street intersections nearest the family home,
work place, and the child care center for both tran
sporting parents.
Estimates of mileage from home to work, home to
child care, and child care to work along with the
forms of transportation used to get to these desti-
nations .
Number of days per week each adult in the household
picks up and drops off children at the child care
Parent's preference between a child care provider
closer to home or closer to work.
For parents using worksite care only, mileage to,
and location of alternate child care arrangements
that would be used if worksite care were not
Mileage to and location of second child care provid
er if one is used and either parent delivered a
Number of times per week parents run errands before
or after picking up or dropping off children at
their child care provider.

Whether job opportunities are limited because of
responsibility for transporting children.
Attitudes about public transportation.
Awareness of park-n-rides near their home and their
willingness to use a child care facility located at
Age, race, gender, income, marital status, relation-
ship to the child and education level.
One difficulty experienced with the survey responses
was the failure of some respondents to give complete
information about the secondary transporter sharing
responsibility. Most often they would complete the first
page of for both primary and secondary transporters,
indicating responsibility was shared, but then fail to
complete the remaining pages for the second transporter.
This study differs from the "Big Six" and other related
studies (Michelson, 1988, Brooker-Gross and Myer-Jones,
1990, Truelove, 1990) because it includes data on the
secondary transporter and their excess mileage. The
other important difference between this study and the
"Big Six" is a printed questionaire versus a telephone

survey where the responses are directly inputed to the
computer. The questionaire allowed the actual mileages
to be calculated and a more accurate margin of error.
Survey Results
Characteristics of Respondents
Demographic and other background information about
the respondents is shown in Table II. The mother
generally responded to the survey (82% for parents with
worksite child care and 85% for the non worksite
families). The average respondent's age is 30 35 years
old, white (90% worksite, 84% non worksite) and married
(84% worksite, 81% non worksite). The parents in general
had more education than anticipated. Sixty percent of
the worksite group had a bachelor's degree or better and
another 22% had some college. In the nonworksite group,
36% had a bachelor's degree or better and 45% some
college. Worksite parents' modal income was $75,000 and
over and the nonworksite parents' $30,000 to $50,000 per
year. In both groups, over 70% of families had incomes
greater than $30,000 and the education levels may have a

correlation. The educational qualifications necessary
for many jobs in hospitals and the high cost for child
care at the surveyed centers may indicate the type of
employees able to afford this arrangement.
Transportation Modes and Preferences
Table III summarizes responses to survey questions
relating to transportation modes, options and
preferences. As expected, overwhelmingly, parents use
their automobile to travel to child care and work, and
would still use their auto to commute to work even
without child care responsibilities. Parental preference
for their automobile may be partially related to the
practice of running errands at least once a week before
or after their trip to child care (59% of the worksite
and 58% nonworksite). The specific reasons of the
respondents for use of their automobile for this commute
were the inconvenience of travelling with children on
public transportation (27% worksite, 10% nonworksite) and
the inconvenience of public transportation coupled with
it being too slow (26% worksite, 5% nonworksite). Most
often (37% of respondents) the nonworksite parents
indicated combinations of reasons for not using public

TABLE II. Demographic Information About Survey Respondents
Question Worksite Nonworksite
Respondent's Relationship to Child
Mother 82% 85%
Father 16 13
Other 1 0
Mean Age Range of Respondent 30-35 30-35
Mean Age of Children in Family 3.2 3.6
Mean Number of Children in Household 1.3 1.4
Ethnic Background of Respondent
White (Non Hispanic) 90% 84%
Black (Non Hispanic) 4 6
Hispanic 2 7
Native American, 0 0
Asian and Other 1 2
Marital Status of Respondent
Married 84% 81%
Single 16 18
Highest Level of Education of Respondent
Some High School 1% 1%
High School or GED 5 9
Some College 22 45
Bachelor's Degree More than a 30 20
Bachelor1s Degree 30 16
Annual Family Income Level
Under $10,000 2% 7%
$10,000 but < $20,000 7 9
$20,000 but < $30,000 10 9
$30,000 but < $50,000 22 28
$50,000 but < $75,000 26 24
$75,000 and over 28 20

transportation. Another expected characteristic of the
parents was the fact they worked full time jobs during
traditional working hours. The data showed 55% of
worksite and 61% of and nonworksite respondents started
work between 7:00 and 8:00 a.m. indicating a higher
percentage of parents than expected working part-time or
non traditional hours.
A surprising and important result was that 13% of
worksite parents and 35% of nonworksite parents would use
a child care center located along a bus route, and that
37% of the nonworksite and 16% of the worksite
respondents would consider using a bus from the center to
work. These findings may be related to the fact about
50% of respondents in both groups live within 1 to 5
miles of a park-n-ride. The "Big Six" survey showed that
parents with school age children prefer child care closer
to home. Parents with children ages zero to four had no
preference when the commute was under 10 miles but
preferred care closer to home when the commute was
longer. This study's finding is parents with worksite
child care prefer their child to be closer to work and
nonworksite parents prefer children closer to home.

TABLE III. Family Transportation Modes, Options, and
Preferences Primary Transporter
Survey Question Work- site Non- work- site
1. Percentage using automobile to get the child to the child care provider. Auto 92% Auto 85%
2. Percentage using automobile to go to work. Auto 92% Auto 77%
3. Percentage using automobile or bus when not dropping off a child at the child care center.
automobile bus 92% 1 68% 14
4. Days and times worked during the week?
Primary Transporters starting work between 7:00 am to 8:00 am. 55% 61%
Primary Transporters with full time jobs. 59 61
Primary and Secondary Transporters both with full time jobs. 26 42
Average number of days per week primary transporter works. 4.6 3.9
5. Percent who would use a child care center and bus to work if good affordable child care were located at a park-n-ride close to home. 13% 35%

Table III Continued
Survey Question Work- Non-
site work- site
6. Distance from home to park-n-ride.
percentage less than 1 mile. 13% 10%
1 to 5 miles. 38 37
over 5 miles. 7 12
don11 know. 37 35
7. Average number of times per week respondent runs errands on way home.
Percentage at least once. 59% 58%
Rarely or never run errands. 36 34
Didn't answer. 4 8
8. Percent who have limited their job location due to convenience to 22% 14%
their child care provider.
9. Child care location preferences:
percentage preferring closer to home. 21% 58%
prefer closer to work. 53 15
doesn't matter. 22 18
did not answer question. 3 7
10. Type of transportation which would be used if a respondent
used a child care provider at a park-n-ride or along a bus route.
Bus 16% 37%
Car/Vanpool 8 12
Bus and Car or Vanpool 5 4
None of the above. 64 37
Did not answer question. 7 10

Table III continued
Survey Question 11. Reasons for not using busr or carpool to get to work: Percent saying inconvenient travel with children using vanpool, to these Work- site Non- work- site
forms of transportation. 27% 10%
too slow. 3 5
too expensive. 0 1
inconvenient to travel with children and too slow. 26 5
combinations of the above. 13 37
doesn't like them. 1 3
other. 22 22
all of the above. 4 1
did not answer question. 4 16
Distances to Child Care
The remainder this chapter gives the results of
several approaches to estimating excess mileage and
potential gasoline consumption savings from providing
child care at the worksite. The results are discussed
around several tables.
Tables IV through VII show the commuting distances
and excess mileage due to child care of both the primary
and secondary transporters in the worksite and

nonworksite groups, separated into suburban and downtown
child care centers. These four tables primarily show the
amount of average excess mileage for each child care
center and the differences between the worksite and non
worksite parents by location of the child care center in
the suburbs or downtown. The primary suburban trans-
porters in the worksite group travel an average of 1.6
excess miles which is fewer excess miles than the non
worksite group (2.3 miles). The worksite and nonworksite
downtown transporters travel 1.3 and 2.0 average excess
miles respectively.
Table VIII summarizes the average excess mileage of
each center for the primary and secondary transporters of
the worksite parents. This table allows the reader to
see where the concentration of excess mileage exists
given the center location. The highest excess mileage
exists for the primary transporter at St. Mary's (5.5
miles) and for the secondary transporter at the Post
Office (27.8 miles). Such high average excess mileage
seems inconsistent with worksite child care. However,
Table IX helps clarify the degree to which centers
provide worksite or near worksite child care. A true

worksite center is defined by this study to be within .2
miles of the parent's workplace, and near worksite,
within 1.0 miles. This table shows which centers are
true worksite or near worksite by our definition.
Tables X and XI show the alternate child care
arrangements of worksite or near worksite parents and the
estimated excess miles saved relative to that particular
arrangement. Note the average mileage savings from
having worksite compared to the parent's chosen alternate
child care center is near zero for primary transporters
and negative secondary transporters. This indicates that
worksite or near worksite costs the secondary transporter
extra miles.
Table XII summarizes the distances of those families
using a second child care provider for both the primary
and secondary transporters in the worksite groups. This
table shows the dilemma parents face with children
needing child care at different child care locations.
For the daily commute, worksite child care saves an
average of .3 miles per one way trip for the suburban
parents, but adds 1.8 additional miles for downtown

Tables XIII and XIV show the number of times per
week that the primary and secondary transporters drop off
and pick up children at their worksite center. This
information is important in calculations for Table XV
which estimates the savings in mileage and gasoline
consumption per child per day and year if parents use
true worksite care by our definition. This calculation
is made for families using child care at one of the
worksite centers, families using an alternate child care
provider and families with nonworksite child care. This
table also summarizes the gasoline consumption related to
child care.
Extra Miles Travelled to Transport Children to Child Care
Excess Miles. Tables IV through VII present the
survey's findings with regard to commuting distances
traveled and excess mileage due to child care for both
the worksite and nonworksite primary and secondary trans-
porters. There does not appear to be a large difference
in mileage savings of parents having worksite or near
worksite child care when compared to parents in the
nonworksite group.

Table IV shows the distances that primary-
transporters travel from home to child care, home to work
and home to child care to work. The final column shows
the difference between the direct trip from home to work
TABLE IV. Distance to Child Care and Work
Near or Worksite Child Care Primary Transporter
mean miles one way, standard deviation in parentheses
Home to Home to Home Differ-
Child Care Child Care to ence*
Center to Work Work
1 2 3 (2)-(3)
Suburban Centers
Fitzsimons 3.7 5.2 4.2 +1.0
(2.9) (4.2) (3.2)
Porter 8.4 9.9 9.7 + .2
(6.4) (7.4) (7.5)
Post Office 12.5 16.4 13.3 +3.1
(7.6) (10.0) (7.3)
St. Mary1s 7.8 12.0 9.4 +2.6
(6.0) (7.0) (6.8)
Swedish 9.0 11.1 9.7 +1.4
(5.5) (7.0) (6.6)
Average Miles 8.2 10.9 9.3 +1.6
in Suburbs
Downtown Centers
Auraria 7.4 9.2 7.8 +1.4
(5.3) (5.9) (5.2)
Denver Place 9.0 10.6 9.5 +1.1
(6.7) (8.0) (7.8)
*Home-to-Child Care-to-Work minus Home-to-Work.

and the trip from home to child care to work, which we
define to be excess miles due to child care. This
difference is the extra one way mileage for just primary
transporters in the sample. Excess mileage due to child
care for primary transporters using worksite child care
averages1.6 miles in suburban Denver and 1.3 miles
downtown. Parents with the highest mileage use St.
Mary's and the Post Office, with 2.6 and 3.1
respectively. Higher mileage for parents using St.
Mary's is due to the fact this worksite center is located
in an office complex covering a large area and not
attached to a single employer. For the Post Office, many
primary transporters are not employees, which might be
due to the fact it is a new center. Neither of these
centers are truly a worksite care center for the majority
of its families. Among the worksite centers chosen
Porter, appeared to be the closest to being a true
worksite center with .2 average excess miles. Even at
Porter, however, 18% of the primary transporters surveyed
did not work at the hospital. Nevertheless many of these
non employees did not have to go out of their way to
reach Porter on their way to work.

Table V shows that slightly higher excess mileages
are experienced by nonworksite primary transporters when
compared to worksite primary transporters with 2.3 and
2.0 miles, respectively, for suburban and downtown
centers. These estimates suggest only minimal savings
TABLE V. ' Distance to Child Care and Work
Non Worksite Child Care Primary Transporter
(mean miles one way, standard deviation in parentheses)
Home to Home to Child Home to Differ-
Child Care Care to Work Work ence*
(1) (2) (3) (2)-(3)
Suburban Centers
U.S. WEST 5.6 (6.7) 18.0 (6.3) 14.5 (7.5) +3.5
Auraria 4.6 9.4 7.5 +1.9
Waiting List (5.3) (6.2) (6.5)
Average Miles
in Suburbs 4.8 11.5 9.2 +2.3
Downtown Centers
Colorado .7 9.7 9.1 + .6
National Bank (.6) (4.0) (3.7)
Manville 3.7 (3.8) 13.8 (6.2) 11.8 (5.2) +2.0
Average Miles
of Downtown 3.5 13.4 11.4 +2.0
*Home to Child Care to Work minus Home to Work.

Distance to Child Care to Work
Near or Worksite Child Care Secondary Transporter
(mean miles one way, standard deviation in parentheses)
Home to Home to Home to Differ-
Child Care Child Care to Work Work ence*
(1) Suburban Centers (2) (3) (2) (3)
Fitzsimons 2.7 (1.7) 4.6 (3.8) 2.9 (2.1) +1.7
Porter 8.2 (7.6) 15.2 (5.2) 12.5 (6.2) +2.7
Post Office 17.2 (6.6) 27.3 (5.9) 13.4 (9.3) +13.9
St. Mary1s 6.7 (4.9) 15.0 (6.9) 9.7 (6.9) +5.3
Swedish 8.2 (7.2) 14.9 (13.6) 8.1 (8.8) +6.8
Miles in 7.7 Suburbs Downtown Centers 15.1 9.6 +5.5
Auraria 7.7 (5.7) 10.9 (5.5.) 6.9 (5.2) +4.0
Denver 8.3 9.4 8.5 + .9
Place (8.2) (7.6) (7.8)
*Home-to-Child Care-to-Work minus Home-to-Work.
from using worksite child care. The unexpected low
savings are traceable to the number of cases where the

Distance to Child Care to Work
Non Worksite Child Care Secondary Transporter
(mean miles one way, standard deviation in parentheses)
Home to Home to Child Home to Differ
Child Care Care to Work Work ence*
(1) (2) (3) (2)-(3)
Suburban Centers
U.S. WEST 6.8 18.3 14.0 +4.3
(7.3) (8.0) (6.8)
Auraria 4.7 11.0 8.8 +2.2
Waiting List (3.2) (6.5) (5.8)
Average Miles in Suburbs 5.3 12.9 10.2 +2.7
Downtown Centers
Colorado 19.6 21.8 15.7 +6.1
National Bank (27.1) (21.3) (13.8)
Manville 4.2 13.4 9.8 +3.6
(3.9) (5.0) (5.2)
Average Miles in Downtown 5.5 14.2 10.5 +3.7
*Home-to-Child Care-to-Work minus Home-to-Work
parents apparently using worksite care did not, in fact,
work at the center site or even very close to it.
Tables VI and VII reports findings for the secondary
transporters. A major improvement in this study over the
Big Six and Toronto studies is information about the
secondary transporter to child care transportation. An

important finding about the secondary transporter is on
average he/she travels farther out of their way to
dropoff or pick up children. The difference column in
Tables VI and VII show a higher excess mileage, overall
than in Tables IV and V. Tables VI reveals that the
excess miles for secondary transporters to worksite child
care average shows greater excess mileage overall than
in Tables IV and V. Tables VI reveals that the excess
miles for secondary transporters to worksite child care
average 5.5 miles in the suburbs and 4 miles for the
Auraria center downtown. Table VII shows that the
secondary transporter using nonworksite child care
arrangements adds fewer miles to their commute only 2.7
or 3.7 miles than does the secondary transporter when
he/she takes the child(ren) to the primary transporter's
Denver Place stands out as one center as not
creating large excess mileage for the secondary
transporter. Our data indicates that both parents using
Denver Place usually work downtown, so this is a true
case of near worksite care. Interestingly, primary
transporters travel more excess miles to Denver Place,

indicating the primary transporter (usually a woman)
works further from the child care center than her
husband. This might be a situation in which both the
child and spouse are dropped off and picked up at Denver
Are worksite centers truly worksite? Table VIII
summarizes the excess miles per trip accrued by the
primary and secondary transporters with worksite child
care, as shown in Tables IV and VI, and converts them
into miles per round trip. The high excess mileage for
both primary and secondary transporters at the Post
Office (5.4 and 27.8 miles respectively) reflects the
fact that many primary transporters using this center do
not work for the Post Office. St. Mary's high excess
miles (5.5 primary and 10.6 secondary) may be due to two
factors: 1) the location of the center in a business
complex covering a very large area; 2) its attractiven-
ess to parents who work in other parts of Denver.
Swedish shows a relatively low excess mileage for primary
transporters (2.8), but high excess mileage for secondary
transporters (13.6). Apparently the separation of work
places between spouses is greater than at most locations.

At the other end of the spectrum, primary transporters at
Porter average just .4 excess miles, reflecting its more
purely worksite character. Even here, the secondary
transporters average 5^3 excess miles per week, indi-
cating that many do not necessarily work at Porter Hospi-
tal. As for the downtown centers, the high excess
mileage for both primary and secondary transporters to
Auraria suggests it is used only partly as a worksite
center. Students have the largest portion
TABLE VIII. Excess Miles Parents Travel Daily (Round Trip)
to Transport Child to Child Care Worksite
Primary Transporter's Excess Mileage Other Transporter1s Excess Mileage
Suburban Centers
Fitzsimons 1.9 3.5
Porter .4 5.3
Post Office 5.4 27.8
St. Mary's 5.5 10.6
Swedish 2.8 13.6
Downtown Centers
Auraria 2.8 5.5
Denver Place 2.3 1.8
of children in attendance and usually work off campus.
The center at Denver Place shows low excess mileage,

suggesting that the downtown area is compact enough to
allow this center to function effectively as a near
worksite center for dual working families when both
parents work near downtown. Even so, these parents do
not have true worksite child care.
The high excess mileage of primary transporters to
centers thought to be worksite or near worksite child
care, led to the question whether or not primary
transporters actually worked at or near those centers.
Table IX shows the percentage of primary transporters at
each of the worksite centers who travel farther than .2
miles and those who travel over one mile (a subset of
those traveling over .2 miles). The results show suggest
that St. Mary's, is not functioning as a worksite center
at all; over three-fourths of its clients work more than
a mile from the center. Denver Place and Auraria both
show substantial groups in all travel ranges; from data
it is somewhat surprising that Auraria parents averaged
more excess miles than Denver Place parents. The Post
office and Fitzsimons both show a minority of parents
travelling more than a mile from child care to work, and
in those cases, the affected parents travelled unusually

long distances. It is interesting to note the large
difference in results between Swedish and Porter
hospitals, which are similar in many ways. At Swedish
28% of the parents travel more than .2 miles to work and
21% travelled more than one mile, whereas for Porter
parents, only 10% travelled over .2 excess miles. An
unusual arrangement with one family at Swedish center
illustrates a problem with worksite child care as an
energy saver. One parent works at Swedish and takes the
child home after her shift each day, incurring almost no
excess miles. But, because that parent's shift starts at
3 a.m., the spouse brings the child to child care at 8
a.m on his way to work. In doing so, the spouse adds 24
miles per day to his one way commute. In this situation
the worksite child care saves for the primary
transporter, but the secondary transporter incurs large
excess mileage. A similar case was found at Fitzsimons
although the difference in mileage was not so
significant. At Fitzsimons many of the families
surveyed were residents of the military base and in some
cases the parent doing the primary transporting worked
off the base, frequently at Lowry Air Force Base. Effec-

tively, for some families, Fitzsimon's became a close to
home, rather than close to work, center.
TABLE IX. Percentage of Primary Transporters Who
Travel Over .2 Mile and Over 1 Mile From Worksite
Child Care to Work
Travelling over .2 miles Travelling over 1 Mile
Suburban Centers
Fitzsimons 17% 17%
Porter 10 10
Post Office 28 24
St. Mary1s 88 68
Swedish 28 19
Downtown Centers
Auraria 32 19
Denver Place 61 27
Mileage to alternate provider. Another way to
estimate mileage savings is to compare mileage to
worksite child care centers with the mileage that would
be incurred to alternate arrangements. In general, our
results suggest that if the worksite parents were not
using the centers in this study, they would be finding
child care closer to home. The sample of parents that
could identify alternative arrangements were small and
the averages may be unreliable, particularly for the

Table X
Distance to Alternate Child Care Provider
Worksite Child Care Primary Transporter
(mean miles one way, standard deviation in parentheses)
Home to Home Home to Home to Savings
Alternate to Worksite Alternate From Hav-
Child Care Work Child Care Child Care ing Worksite
to Work to Work
Child Care
(1) (2) (3) (4) (4) (3)
Suburban Centers
Fitz- 1.3 4.2 5.2 5.6 + .4
simons (1.3) (3.2) (4.2) (3.2)
Porter 3.0 9.7 9.9 10.3 + .4
(4.3) (7.5) (7.4) (5.6)
Post 2.1 13.3 16.4 16.8 + .-4
Office (3.2) (7.3) (10.0) (8.2)
St. Mary's 4.4 9.4 12.0 10.1 -1.9
(4.4) (6.8) (7.0) (7.8)
Swedish 4.1 9.7 11.1 12.1 +1.0
(5.9) (6.6) (7.0) (7.3)
Average of Suburbs 3.2 9.3 10.9 10.9 0
Downtown Centers
Auraria 1.6 7.8 9.2 8.7 -.5
(1.8) (5.2) (5.9) (5.5)
Denver 2.3 9.5 10.6 11.5 + .9
Place (3.7) (7.8) (8.0) (7.5)
* Column 4 minus Column 3.
secondary transporter. The alternative arrangements
differ dramatically in their effects from center to
center, but overall there does not appear to be any

substantial amount of savings in miles from having
worksite child care available for the suburbs for primary
TABLE XI. Distance to Alternate Child Care Provider *
Worksite Child Care Secondary Transporter
(mean miles, standard deviation in parentheses)
Home to Alternate Child Care (1) Home to Work (2) Home to Worksite Child Care to Work (3) Home to Alternate Child Care to Work (4) Savings from Worksite Child Care* (4)(3)
Suburban Centers
Fitzsimons 7.8 (6.6) 4.1 (1.3) 7.6 (1.6) 13. (14.3) +5.6
Porter 3.9 (6.6) 11.9 (4.1) 12. (4.6) 12.2 (4.2) -.4
Post Office 1.4 (1.0) 13.4 (9.3) 27.3 (5.9) 14.4 (8.8) -12.9
St. Mary's 2.9 (2.0) 6.3 (5.5) 12.3 (7.0) 11.2 (6.6) + .9
Swedish 9.5 (8.9) 6.8 (8.5) 22.5 (20.1) 18.4 (15.0) +4.1
Average Miles in Suburbs 4.0 8.1 15.4 12.9 -2.5
Downtown Centers
Auraria 1.0 (.7) 9.2 (2.4) 14.3 (4.5) 10.6 (3.4) -3.7
Denver Place 2.4 (2.0) 5.9 (3.4) 6.1 (3.2) 7.9 (3.3) +1.8
* Column 4 minus Column 3.

TABLE XII. Mileage to Second Child Care Provider
For Worksite Parents Transporting Children to Two Providers, Child Care Primary and Secondary Transporters (miles)
Home to Home to Child Work Care (1) (2) Home to Worksite Child Care to Work (3) Home to Second Child Care to Work (4) Savings from Worksite Child Care* (4)-(3)
Worksite Centers - Primary Transporters
Suburbs 4.4 9.3 10.5 11.1 + 6
Downtown 4.1 8.6 8.9 1.3 -7.6
Average of Primary 4.3 9.1 9.7 6.2 -3.5
Worksite Centers - Secondary Transporter
Suburbs 3.9 9.5 14.3 12.2 -2.1
Downtown 3.5 8.0 10.6 11.0 + .4
Average of Secondary 3.8 9.1 13.7 11.9 -1.8
* Column 4 minus Column 3
transporters (Table X). For secondary transporters, the
worksite care adds somewhat to excess mileage over the
alternate arrangements in about half the centers and on
average. For families using downtown Denver centers,
excess miles would be reduced by the alternative
childcare arrangements for Auraria parents and increased
for Denver Place. Apparently travel convenience in not

an important factor for these parents in choosing a
Travel for parents using two child care providers.
Table XII provides a breakdown of differences in excess
mileage for parents having children at two different
child care providers. The table showsthat on average,
the primary transporter travels about .3 miles less in
transporting children to near worksite centers than to
the second provider. On the other hand, the secondary
transporter travels an aditional 1.8 miles to the
worksite child care center than to the second provider,
The second transporter's workplace is more convenient to
the secondary provider than to the worksite center.
However, these results may not be reliable as only 13% of
primary transporters reported using a second provider.
Number of times primary and secondary transporters
pick up or drop off their child(ren). Tables XIII and
XIV show how parents divide the responsibility for
transporting their child to a child care provider. On
average the primary transporter (usually the female)
dropped off and picked up the child seven time or more

TABLE XIII. Number of times Per Week Primary
Transporter Picks Up and Drops Off Child(ren) at
Child Care
Child Care Center 1-2 3-4 5-6 7 or more
Worksite Centers Suburban Centers Fitzsimons 0 1 5 19
Porter 6 10 8 22
Post Office 0 2 4 19
St. Mary's 0 4 19 36
Swedish 0 10 11 19
Downtown Centers Auraria 1 9 10 33
Denver Place 0 0 9 35
Nonworksite Suburban U.S. WEST 0 0 0 17
Auraria Wait List 6 11 16 22
Downtown Colorado National Bank 0 0 0 23
Manville 1 4 11 22
each week in both the suburban and downtown centers.
This result is similar to the Big Six
and Toronto studies which found that women are mainly
responsible for the transportation needs of their

Table XIV. Number of Times Per Week Secondary
Transporter Drop Off or Pick Up Child(ren) at
Child Care
1-2 3-4 5-6 7 or More
Suburban Centers
Fit2simons 2 2 3 0
Porter 8 3 2 0
Post Office 1 1 2 1
St. Mary's 6 6 15 2
Swedish 3 5 3 2
Downtown Centers
Auraria 6 4 7 1
Denver Place 3 3 10 2
Nonworksite Suburban
US WEST 0 0 7 1
Auraria Wait List 14 8 10 14
Colorado National Bank 0 0 6 0
Manville 6 4 7 2
children. A large number of both primary and secondary
transporters at St. Mary's pick up or drop off their
childa total of five to six times per week suggests that

a significant number of parents share the transporting
their of children at this center.
Table XIV shows several cases where the secondary
transporter dropped off and picked up the child at least
five times per week. Cases where the secondary
transporter makes 5-6 trips a week are generally
caseswhere parents split the transportation of children
to child care. The few cases where the secondary
transporter is reported making 7 or more trips per week
are mostly those in which both parents commute together.
A few times the primary transporter performs this task
only two or three times per week, and are cases where the
respondents is employed part-time or worked longer
shifts. The secondary transporters, in general, do much
less commuting with children than the primary
transporters which supports findings of the Toronto
Gasoline savings from true worksite care. One reason
for the low mileage savings reported in this study is
that even for primary transporters, "worksite" child care
is not really adjacent to their place of work. To find
the savings from true worksite care, we estimated how

much savings would exist if the primary transporter's
trip from child care to work were only .2 miles. Table
XV shows the potential reductions in mileage and gasoline
consumption for this hypothetical true worksite child
care given the same sample of parents. This projection
was developed by assuming that all primary transporters
in this study worked within .2 miles of their child care,
TABLE XV. Mileage and Gasoline That Would Be Saved with
True Worksite Child Care
(Excess Gasoline per day and per year per child)
Actual Excess Mileage (1) Estimated Excess Mileage Worksite Care* (2) Differ- ence** (1) (2) Excess Gasoline Consumption Per Day (gallons) (4) Excess Gasoline Con- sumption Per Year (gallons) (5)
Worksite 3.2 1.4 +1.8 .12 30.0
Alternate Child Care 4.9 2.3 +2.6 .17 43.3
Non- worksite 3.2 1.3 +1.9 .13 31.7
* Computed assuming that all primary transporter
works within .2 miles child care and that the
secondary retains same excess miles and number of
trips per week
** Column 1 minus Column 2.

but secondary transporters retained their excess mileage
and frequency of taking the child to child care. Excess
mileage for both transporters was weighted by their
respective numbers of drop offs and pick ups per week and
averaged for each respondent. It should be noted that
excess miles would be reduced somewhat more if the true
worksite care were zero miles from work or led to the
primary transporter taking still more responsibility for
transporting the child.
Column 1 in Table XV shows the actual extra mileage
and column 2 shows extra miles to transport children if
the center were only .2 miles from the primary
transporter's work. The "difference" (column 3) is the
actual excess miles minus the estimated excess miles for
true worksite child care. Using 1992 E.P.A. gasoline
consumption estimates of 15 miles per gallon, the miles
were converted to excess gasoline consumption per day
(Column 4). Since the majority of parents in this study
worked full-time, excess gasoline consumption per year
was computed by multiplying the daily excess gasoline
consumption by 250 working days per year.

The amount of energy saved with true worksite child
care is projected to be 30 gallons per year per child for
the group that now has nominally worksite child care.
Families of the nonworksite sample would save 31.7
gallons per year. When compared with the alternative
child care arrangements, the parents using nominally
worksite care true worksite child care would save 43.3
gallons per year. The group which indicated an alternate
child care arrangement is not directly comparable to the
group in Tables X and XI because of differences to
weighting the mileage by the number of trips taken per
Results of this study demonstrate the complex
decision making process parents face when choosing child
care and transporting children. Many factors such as
quality, time constraints related to travel, convenience
to the work commute, and excess mileage must be
considered when trying to choose the best arrangement for
their child(ren). Nevertheless, true worksite centers
will produce some gasoline savings as seen in Table XV.
The results also indicate that for dual earner families
with high incomes and in their 30's, near worksite

centers providing good child care will also attract many
parents regardless of the excess mileage.
Does Child Care Create A Perceived Job Limitation?
The previous sections of this chapter have discussed,
the potential savings of gasoline by parents using
worksite child care centers. This section examines the
results of a survey question regarding job limitations
due to child care. This is an issue of particular
interest, given the literature review regarding gender
differences. Both versions of the survey questionaire
included a question, "Have you limited your job location
due to convenience to your child care provider?" (See
Appendices C and D, question 19.) I was interested in
discovering whether proximitiy to a child care provider
affected the job search. Interestingly 75% of worksite
parents, while only 15% of the nonworksite parents,
responsed that they had limited job opportunties. The
purpose of the regression analysis discussed below was to
identify the basis for this difference between the two
groups of parents.

Procedures. The job opportunities question is
explored in a number of ways. The survey responses
provides a large number of variables that may explain or
modify the impact of employment opportunties as it
related to child care responsibilities. Further, the
data were collected from two different samples of
parents: parents using worksite child care, and parents
without worksite care available. Both samples allow an
assessment of how parents perceive limiting employment
opportunities when responsible for child care
transporting. The dependent variable is structured as a
dummy variable equal to 1 if the parent has limited their
job search, 0 is not. The first step was to identify
variables collected in the survey that may significantly
affect the parent's desire for worksite child care, and
therefore employment choices. The variables selected
Excess mileage of primary transporter: The
assumption was the further the commute from home to
child care the greater the time constraint and
opportunity cost (Hanson and Johnston, 1985). The
primary transporter would not want to commute long

distances from home to child care and would choose a
provider convenient to his/her daily commute. This
implies a positive relationship with the dependent
variable, and this variable is continuous.
Age: Younger parents (under 30) have more
limitations placed upon them because of children
being younger (Michelson, 1985). This implies a
postive relationship to the dependent variable. (In
the simple regression each age range individually is
set = 1).
Education: Higher education levels provide more
employment opportunities (Brooker-Gross and Myers-
Jones, 1990). This implies a negative relationship
with the dependent variable. (In the simple
regression each level of education individually is
set = 1).
Gender: As shown in chapter two, (Brooker-Gross and
Myers-Jones, 1990; Rutherford and Wekerle, 1988;
Hanson and Johnston, 1985) women are usually the
primary transporters of children from home to child
care. This implies a positive relationship with the
dependent variable. (Female = 1, male = 0 in data set.)

Family income: Greater family income creates more
opportunities for child care arrangements. This
implies a positive relationship to the dependent
variable. (In a simple regression each income range
individually was set = 1.)
Form of transportation from home to child care:
Access to an automobile allows the parent greater
mobility for job opportunities and more options for
selecting affordable, quality child care providers
(Hanson and Johnston, 1985). In addition, a car
reduces time spent commuting and reduces the time-
budget constraint. This implies a negative
relationship with the dependent variable since
automobile = 1 in the data set.
Additional minutes created by child care
responsibilities in the daily commute: The primary
transporter will limit his/her employment
opportunities in order to reduce travel time from
home to work (Truelove, 1990). This implies a
positive relationship with the dependent variable.
In a simple regression each time range was
individually set = 1).

Preference for a child care provider closer to home
versus work: Depending upon the parents'
preference, they may limit their job opportunities
in order to avoid commuting long distances with a
child. The sign for this relationship is not clear.
(In a simple regression each preference was
individually set = 1).
Percentage of times per week primary transporter
drops off and picks up child(ren) at the child care
provider: If one parent is predominately
responsible, that individual's job opportunities may
be limited to working traditional hours during the
day, when more convenient options for child care
exist. This implies a positive relationship, and
the variables are continuous. (This percentage was
calculated for each primary transporter by summing
his/her total drop offs + pick ups per week and
dividing by the grand total of number of drop offs +
pick ups.)
A series of two variable regression tests were run
on the hypothesized independent variables to determine
their level of significance. The dependent variable is

the "yes" or "no" response to the question of limiting
job opportunties. Given a binary dependent variable,
logit analysis is required. The form of the test is:
L = J30 + J3iXi + e where:
L = the primary transporter answered "yes" they
had limited their job opportunities because
of child care. Dependent variable.
J30 = the Y-axis intercept or constant variable.
J3iXi = independent variables and coefficients,
s = the error vector.
The fi1X1 variable is always excess mileage because of
it being the only continuous variable. All the other
independent variables take values of 0 or 1, and are
included or excluded as the search for relationships was
made from the data. The results of the testing indicated
that the two sample groups had very different reasons
when they limited their employment opportunties. The
remainder of this section discusses the two tests
Results for the nonworksite sample. A number of
regressions were run to find significant variables
showing why primary transporters limit their job
opportunties. The most interesting regression for

nonworksite parents is reproduced in Table XVI. The wish
to avoid excess mileage does not appear to restrict
parents' job search and child care. However, respondents
over 30, may limit their job opportunties. The other
variables tested did not prove to be a factor in limiting
the primary transporters job search.
Further investigation of the nonworksite data set
was necessary to search for other significant independent
variables. The two variable logit regression was used,
and all the remaining variables collected in the survey
were tested to determine their significance. The three
variables showing significance in this set of test
results were: parent indication that they would use
worksite child care if it were available; preference for
their present child care arrangement to be closer to
home; and the percentage of times the primary transporter
dropped off and picked up their child.
A logit model was used which included all four
significant variables. The model uses one continuous
independent variable (excess mileage of primary
nonworksite transporter) and four dummy variables
(willingness to use worksite, preference for child care

closer to home, percentage of the week primary trans-
porter drops off and picks up child(ren), and
respondent's age is over 30). Table XVI summarizes
results, and using the coefficients for the variables the
equation for this regression is:
L = -0.753 + (-0.018)X1 + (-1.167)X2 + 1.898X3 +
-0 O28X4 + 6.
From the equation and Table XVI the coefficient signs of
the variables did not result in the same hypothesized
Excess mileage was used primarily because the model
required a continous variable, and this was the only one
in the data sample. The negative sign of the coefficient
indicates that the greater the excess mileage the less
likely the parent will limit employment opportunities.
This finding is interesting because of its constrast to
the Brooker-Gross and Myers-Jones (1990) study which
concluded that in nonmetropolitan areas parents were
sensitive to distance to child care. Evidently parents
in metropolitan areas are not as sensitive or perceive
distance differently than parents in small towns or have
already reduced their excess mileage.

Table XVI.
Logit Regression Model
(Nonworksite Sample)
Logit // Dependent Variable is JOB
Sample Range: 1 -163
Oberservations excluded because of missing data.
Number of observations: 138
Convergence achieved after 5 iterations
Nonworksite Sample
-0.7892149 0.4314*
-0.1904484 0.8492*
-2.1089572 0.0184
2.3557200 0.0099
-3.5651322 0.0003
* 2-Tailed significance
Log liklihood -45.590729
Cases with JOB =1 22
Cases with JOB =0 116
130 = Constant variable
J31X1 = Excess mileage of primary transporter. A
continuous variable.
J32X2 = Age of respondent was over 30. If age was
greater than 30 = 1, all other ages = 0.
J33X3 = Primary transporter prefers child care closer to
home. Response closer to home = 1, closer to
work or didn't matter = 0.
J34XA = Percentage of times per week primary transporter
drops off and picks up child from child care
center. A continuous variable.

The variable, age of 30 or older, is significant and
bears a negative coefficient, which is consistent with
the hypothesis because it implies the lower the parent's
age the more likely they will limit job opportunity. The
hypothesis was that young parents are more likely to have
young children requiring child care.
The parents preference for their present child care
arrangement being closer to home is consistent with the
hypothesis and can be explained by a finding in
Truelove's (1990) study that parents will place a child
closer to home even when a lesser price is available at a
slightly further distance. Since this study did find a
higher excess mileage from home to child care to work in
the nonworksite group, the positive coefficient implies
that parents that proximity to child care is important.
The percentage of times the primary transporter drops
off and picks up the child(ren) per week has been used in
previous studies (Brooker-Gross and Myer-Jones, 1990).
The negative coefficient in this case is troublesome,
because it implies that the higher the percentage of drop
offs and pick ups the less likely the parent will limit
job opportunities. This result is not consistent with

any other studies (Brooker-Gross and Myers-Jones, 1990;
Michelson, 1988) examining distances to child care and
work. Clearly a primary transporter responsible for
child care arrangements would have limitations with
working hours and type of job.
The results for the worksite sample. About 75% of
worksite parents reporting they limit their job
opportunities due to their child care responsibility
compared to only 15% of nonworksite parents. The number
of cases with JOB = 1 in Table XVII shows the reversal of
frequency when compared to the results in Table XVI of
the nonworksite sample. This finding clearly indicates
that parents who wanted worksite child care found an
employer who provides it.
A number of two variable regressions were also run on
the worksite data sample. The results shown in Table
XVII again reveals that excess mileage does not appear to
affect whether or not the primary transporter limits
employment opportunities because of child care
responsibility. The results corraborate the fact that
reducing excess mileage is important. However, three
factors did limit parents employment opportunities: if

Table XVII.
Logit Regression. Model
(Worksite Sample)
Logit // Dependent Variable is JOB
Sample range: 1 301
Observations excluded because of missing data.
Number of observations: 248
Convergence achieved after 3 iterations
Worksite Sample
% 2.1103934 0.4527427 4.6613529 0.0000*
B.X, -0.0352177 0.0387620 -0.9085629 0.3645*
BpX, -0.6911487 0.4492161 -1.5385660 0.0626
BX -0.6177209 0.3253695 -1.8985212 0.0588*
V* -0.7404930 0.3874523 -1.9111847 0.0572*
* Two tailed significance test
Log likelihood -132.41944
Cases with JOB = 1 188
Cases with JOB =0 60
J30 = Constant variable.
J31X1 = Excess mileage of primary transporter, a
continous variable.
J32X2 = Respondent's gender was female. Female = 1 and
male = 0.
J33X3 = Family income was greater than $75,000/year.
Income range of 75,000 > = 1, all other
ranges = 0.
J34X4 = Primary transporter worked 3 day week. Less
than or equal to 3 day week = 1, greater than
3 day week = 0.

the respondent was female, the family income was $75,000
per year, and the primary transporter worked less than or
equal to three days per week.
The following is the regression equation of the
worksite sample:
L = 2.11 0.04X1 - 0.69X2 - 0.62X3 - 0.74)X4
The coefficients for these variables shown in Table
XVII do not provide much information concerning the
likelihood that child care responsibilities would limit
job selection, the basis of the difference between the
two data samples. The negative sign for the gender proxy
implies that females are less likely than males
transporters to limit employment opportunites. This
result is not consistent with the hypothesis of this
study or findings from other studies (Hanson and
Johnston, 1985).
Higher income levels may be important due to the fact
the highest percent (28%) of the worksite sample falls in
this income bracket. The negative coefficient on the
income variable indicates that as income increases
parents are less likely to limit job opportunities.

Finally, the most surprising variable is the three day
work week. The negative coefficient indicates that
working 3 days per week or less reduces the likelihood of
limiting job opportunties. Previous studies such as
Hanson and Johnston (1985) found that working part-time
did not limit the distance to work. Since the number of
work days for primary transporters was part of the
hypothesis, it was assumed that worksite child care would
create limitations to the job search, especially part-
time work.
Results of the cross-pooling tests. The two data
samples were pooled to identify whether or not the two
samples could be pooled. In Test 1 the worksite sample
is pooled into the nonworksite sample. The logit
regression was run on the pooled sample to determine
whether the variables found for the nonworksite parents
are also significant for the worksite parents (See
Appendix A). Table XVIII shows the results of the cross-
pooling for Test 1. The independent variables cross-
pooling level of significance (28) when compared to the
Chi-Square distribution table (1% level of significance)
value (13.27) reveals that the nonworksite and worksite

Table XVIII. Pooling Tests Results
Log Likelihood Value*
Test 1 Test 2
Nonworksite -45.59 -51.96
Worksite -150.79 -132.42
Both Groups -210.87 -208.92
Cross Pooling** 28 50.00
Chi Square Critical Value 13.27 13.27
(1% level)
*See Appendix A and B
Independent Variables Used in Tests
Test 1. = Excess mileage of primary transporter, age
of primary transporters over 30, preferred
child(ren) closer to home, and percentage
of drop offs and pick ups per week.
Test 2. = Excess mileage of primary transporter,
gender of respondent was female, family
income was $75,000 or more, and worked 3
days per week.
** Formula for Significance in Cross-Pooling
-2[log likelihood of pooled nonworksite and
worksite (log likelihood of nonworksite + log
likelihood of worksite) ]
groups are very different. Thus, the cross-pooling test
reveals the variables significant in the nonworksite
sample are not significant for the worksite group.