Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The flexible workplace: regional tendencies and daily travel implications
(USC Thesis Other)
The flexible workplace: regional tendencies and daily travel implications
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
The Flexible Workplace: Regional Tendencies and Daily Travel Implications
by
Mohja Lynn Rhoads
Dissertation Submitted in Partial Fulfillment
of the Requirements for the Degree of
Doctor of Philosophy
Policy, Planning and Development
University of Southern California
May 2015
Acknowledgments
I would like to thank my lovely parents Stephen and Mirta Rhoads for providing me with
moral and technical support throughout the duration of my dissertation. You guys are my best
friends and I love you with all my heart.
I would like to thank Professor Genevieve Giuliano for the tireless advice, guidance and
resistance she provided me. The strength of this work is a result of her help.
Finally, I would like to thank all of my other friends who encouraged and listened to me
all the while.
Table of Contents
Introduction .................................................................................................................................................. 1
Literature Review: Flexible workplace options: A literature review of workers and organizations ............. 4
Abstract Literature Review ....................................................................................................................... 4
Introduction Literature Review ................................................................................................................. 4
Supply-Side: Why do Organizations Offer Telework Options? ................................................................. 9
Demand-side Literature: Individual Drives and Motives ........................................................................ 18
Conclusion Literature Review ................................................................................................................. 27
References Literature Review ................................................................................................................. 29
Paper 1: Regional variation in the provision and uptake of flexible workplace options ............................ 38
Abstract Paper 1 ...................................................................................................................................... 38
Introduction Paper 1 ............................................................................................................................... 38
Conceptual Background .......................................................................................................................... 42
Data and Analysis Strategy...................................................................................................................... 53
Empirical Results ..................................................................................................................................... 60
Regression Results .................................................................................................................................. 65
Conclusion Paper 1 ................................................................................................................................. 71
References Paper 1 ................................................................................................................................. 76
Appendix Paper 1 .................................................................................................................................... 85
Paper 2: The un-fixed workplace: Interactions between the workplace and space-time constraints on
daily activities .............................................................................................................................................. 87
Abstract Paper 2 ...................................................................................................................................... 87
Introduction Paper 2 ............................................................................................................................... 88
Theoretical Discussion ............................................................................................................................ 91
Data ....................................................................................................................................................... 104
Analysis Strategy ................................................................................................................................... 107
Differences in socio-demographics and travel behaviors by Flexible Work Options and Gender ....... 111
Differences in trip behaviors by Option to Work at Home and Gender ............................................... 114
Relative Distribution of Trips Made by Time of Day for Ability to Work at Home ............................... 117
Regression Results ................................................................................................................................ 123
Conclusion Paper 2: .............................................................................................................................. 131
Discussion Paper 2 ................................................................................................................................ 135
References Paper 2 ............................................................................................................................... 138
Appendix Paper 2 .................................................................................................................................. 147
Conclusion ................................................................................................................................................. 150
List of Tables
Table 1: .......................................................................................................................................................... 7
Table 2: ........................................................................................................................................................ 16
Table 3: ........................................................................................................................................................ 21
Table 4: ........................................................................................................................................................ 24
Table 5: ........................................................................................................................................................ 52
Table 6. ........................................................................................................................................................ 61
Table 7. ........................................................................................................................................................ 67
Table 8: ........................................................................................................................................................ 70
Table 9: ........................................................................................................................................................ 85
Table 10: .................................................................................................................................................... 111
Table 11: .................................................................................................................................................... 113
Table 12: .................................................................................................................................................... 116
Table 13: .................................................................................................................................................... 117
Table 14: .................................................................................................................................................... 126
Table 15: .................................................................................................................................................... 127
List of Figures
Figure 1: ...................................................................................................................................................... 19
Figure 2: ...................................................................................................................................................... 42
Figure 3: ...................................................................................................................................................... 64
Figure 4: ...................................................................................................................................................... 93
Figure 5: .................................................................................................................................................... 118
Figure 6: .................................................................................................................................................... 119
Figure 7: .................................................................................................................................................... 120
Figure 8: .................................................................................................................................................... 121
Figure 9: .................................................................................................................................................... 122
1
Introduction
Flexible workplace options such as the ability to work remote from the workplace
(telework) or to adjust working start times has interested scholars and policymakers for decades.
In a transportation context, flexible workplace options are proposed as alternatives to peak-hour
commuting as the commute trip can either be eliminated or adjusted. Transportation scholars and
policymakers are also interested in how flexi-workers travel in general. Within human resource
and management domains, flexible workplace options are of interest in understanding how
organizations adjust to competitive climates. Flexible work arrangements allow organizations to
hire talent from a wider geography or through offering their employees benefits in the form of
flexi-work. Flexible workplace practices have also been proposed as tools to engender better
work-life balance to fit the specific needs of households.
Within the last decade renewed interest in flexible workplace practices has arisen due to
the recent and rapid advances in information communication technologies (ICTs). ICTs have
made remote work more possible than ever. The ubiquitous and cheaper availability of portable
devices, cloud technologies and 4G networks allow for the portability and feasibility of one’s
workplace virtually anywhere. Organizations have new labor demands for talent that has the
necessary skills for the emerging digital environment. The global and networked nature of many
business operations means that communication and collaboration is occurring not only within
near proximity, but simultaneously around the world. Not only is a new language being spoken
in the workplace, but the language is being spoken with a new set of tools that are continuously
expanding, evolving and becoming obsolete, causing employers, employees and customers to
communicate, transact, collaborate and coordinate in a host of virtual ways.
2
Overall, the research on flexible workplace practices has not reached a consensus as to
why people engage in flexible workplace practices and why organizations allow these practices.
Studies considering why individuals adopt flexi-work have not comprehensively examined the
occupations and industries of these individuals. Ultimately, one’s organization and type of work
determine whether or not the individual can engage in flexi-work. Concomitantly, studies
considering why organizations provide flexible workplace practices have not comprehensively
considered the larger competitive forces under which the organizations operate. These studies
seek to understand how organizations adjust to their competitive environments through flexible
employment practices, yet have not examined the competitive variations under which different
organizations operate. The literature on individuals and organizations has generally been kept at
small scales of analysis at varying locations rendering results inconsistent and un-generalizable.
Parallel to the study of why individuals adopt flexible workplace practices is the
investigation of how those who adopt flexi-work, travel. Flexible workplace practices are
inherently bound to ICT use in order to conduct remote work and are therefore a measure of ICT
use. ICT use has the potential to substitute trips (e.g. replace a commute trip), compliment trips
(allow one to modify their trip while en route), or incentivize trips (encourage more travel).
Concepts such as fragmentation and fixity have been developed to understand such
behaviors. ICTs can engender more travel by informing or requiring extra travel, or they can
reduce travel by bringing more activities within the domain of the home. The literature on
fragmentation and fixity has focused on individual aspects such as gender and socio-
demographics, generally leaving out how one of the main pivots in an individual’s travel, the
workplace, influences travel. It is still unclear as to how the use of ICTs affect travel and how
3
people who partake in more fragmented work behaviors (e.g. work from home) as a result of
ICTs travel in general. The research is new on this front and large-scale data are hard to come by.
This work explores the above mentioned issues in three papers. The first paper conducts a
literature review on flexible workplace practices. The literature review is followed by two
quantitative papers: the first of which examines flexible workplace practices at the level of the
region and the second, which examines the travel patterns of individuals who have flexible
workplace options.
A regional analysis of the nation has several benefits. Small-scale analyses only convey
results with respect to the institution or area surveyed. A national analysis allows for larger
patterns to be detected. Because there is no national dataset which simultaneously incorporates
an individual’s flexible workplace options and her occupation, examining trends at the regional
level allows for the consideration of industry and occupation.
The second paper also uses a national dataset and looks at the travel patterns of
individuals who have flexible workplace options such as self-employment status, ability to work
at home and the ability to adjust working start times. These individuals serve as a group who can
utilize ICTs to conduct remote work. Comparing these groups to groups who do not have flexible
workplace options and measuring behavior such as number of trips they make and their duration
of activities can shed light on how flexibility and ICT may be resulting in non-traditional travel
patterns.
4
Literature Review: Flexible workplace options: A literature review of workers and
organizations
Abstract Literature Review
This review examines the literature on flexible workplace practices, namely the practice
of remote working or telecommuting/telework. As of yet, we do not have a strong narrative
describing why individuals are offered the option to telework, or what the typical or average
teleworker looks like beyond the fact that they tend to earn higher incomes and are more
educated than non-teleworkers. Mixed results are due to small sample sizes originating from
institutions at varying locations and the different metrics and measurements used to classify
teleworkers and non-teleworkers and their attributes. Moreover, telework behavior begins with
the organization. An individual can only telework if her organization allows her to do so. Only a
small percentage of the workforce is formally allowed the option to work at home. Those
organizations who do allow their employees to formally work at home might be doing so to
adjust their organizational models in order to secure talent. Organization-focused research is
limited and faces problems due to small sample sizes focused in Europe and a lack of a deeper
investigation of industry.
Introduction Literature Review
Given the current state of modern technology, why do we continue to see low rates of
flexible workplace practices? Why are the majority of workers still travelling to the workplace
every day, putting in standard full days of work at the workplace? Modern digital technologies
allow for the portability of work in the form of handheld devices, laptops, ubiquitous wireless
connections and so forth. Yet, we still continue to see low rates of formal flexible workplace
options. The answer is, we still don’t really know.
5
This paper refers to flexible workplace practices as telework- the ability of a worker to
work in a place of her choosing. Given the modern state of technology, many can work remote
from the workplace and do so checking e-mails in the morning and evening, or working while
en-route to the workplace, for example. Therefore, the definition of telework employed in this
paper utilizes another dimension and that is the ability for a worker to work in a place of her
choosing that is directly motivated by her employer, or her type of work. That is, she is not
required every day to show up at a place of work for a standard period of time.
To date, little is known about the typical teleworker or organizations that allow telework.
Little is known about the organizations that don’t allow for telework or why individuals who are
allowed to telework, don’t.
Telework behavior, in the above mentioned context, occurs when an organization allows
her employee the option to conduct flexible work and when she desires to do so. In other words,
flexible work occurs when supply meets demand. Both supply and demand are layered and each
system faces sets of constraints and drives.
The supply-side of flexible workplace practices begins with the organization and the
particularities leading the organization to allow for the practices, or prohibits them. An
organization may wish to allow their employees flexible workplace options as a form of a benefit
so they can attract and retain competitive talent. Organizations may wish to minimize office
space and reduce costs, or hire talent from other geographies. Alternatively an organization may
not be comfortable with flexible workplace practices, as they feel physical presence is required
to better manage their employees.
The demand-side of flexible workplace practices begins with the individual and the
personal particularities leading her to take advantage of, when available, flexible work. An
6
individual may wish to telework in order to save time spent on commuting to the workplace. He
may be inclined to partake in telework so he can spend more time at home with her family, or in
order to be more productive with work in an environment freer from distractions. Alternatively, a
worker’s type of work may require him to be present in the workplace, or she may not have the
necessary resources or space in which to conduct remote work. Or, an individual simply may
prefer to go to the workplace to be seen, or to collaborate with other workers.
There is evidence that the main constraints to flexible workplace provision resides in the
supply of it- most organizations don’t allow for telework. National statistics from the 2009
National Household Transportation Survey (2009) reveal that only around 14% of the workforce
is allowed to work at home, including the self-employed and the non-self-employed. There is
evidence that more people would like to telework, but they are not given the option. Within one
study, 88% of the respondents desired to telecommute, but only 13% actually did so, because
most did not have the option (Mokhtarian & Salomon, 1996).
One may think that the low rates of telework provision may be due to the fact that many
occupations require physical presence, but the virtualization of work is ever increasing and even
for the most physical of jobs, workers tend to do a portion of their work at home. Recent data
from work at home patterns presented by the US Bureau of Labor Statistics for 2003-2007 (table
1) show that, on average, workers in all listed occupations work at home at some point during
the week, although the time spent at home varied by occupations. For all those who are not self-
employed, the average time spent working at home falls at roughly 4 hours per week. Even
occupations for which physical presence is required, such as: food preparation and serving,
cleaning and maintenance, construction and extraction, installation, maintenance and repair,
show some amount of work at home per week: from .5 hours to 1.2 hours per week.
7
Table 1:
Average Hours Worked at Home per Week and Percent of Workforce (US Bureau of Labor
Statistics 2003-2007)
Occupation Average Hours Worked at
Home per Week (2003-2007)
Percent of
Workforce (2012)
Management, Computer and Mathematical
Occupations, Education, Personal Care and
Services, Farming, Fishing and Forestry
7.1 hours and above 17.3%
Business and Financial, Life, Physical and Social
Science, Community and Social Services, Legal,
Arts, Design and Entertainment, Sales
4.1 through 7.0 hours 20%
Architecture and Engineering, Healthcare,
Protective Services, Food Preparation, Building and
Grounds Cleaning and Maintenance, Office and
Administrative Support, Construction, Installation
and Maintenance, Production, Transportation
.5 through 4.0 hours 62%
In order to fully understand flexible workplace practices, several questions must be
answered:
1. Is telework allowed?
2. If so, how much telework occurs?
3. What are the reasons telework is allowed, or not?
4. What are the reasons individuals telework, or not?
To answer such questions, a representative national sample of the workforce is required.
Such a survey would address the above questions by asking the individual whether or not she is
allowed to telework and how often she does. It would ask her detailed questions about her
occupation to gauge, whether or not it allows for virtual presence. The survey would then ask the
individual questions about her organization, its industry, size, and age, for example.
Accompanying this information would be questions assessing socio-demographic and
geographical information, including her place of residence and the location of her employer.
Essentially, questions regarding all the drives and constraints from the supply and demand sides
would need to be addressed.
8
From here several models would be built. Models examining why individuals are offered
the choice to telework would utilize discrete choice modeling as the variable of significance is
binary: the individual either has the option or not, and continuous consumption models do not
apply. A host of variables assessing the organization and the individual’s type of work would be
included as explanatory variables, such as: the individual’s occupation, the industry of her
organization, the place where she resides, the location of her organization, characteristics of the
organization such as its age to measure engrained levels of bureaucracy, the organization’s size,
and her status in the organization (e.g. manager, or executive).
Individuals allowed the option to telework would be separated from the sample to build
models of who actually teleworks and how much. These models could also be discrete, such as
multinomial logit models where quantity is ordered into categories, or Poisson models which are
more appropriate for analyzing counts. The demand-side models would have as the dependent
variable how much one teleworks and the independent variables would be a host of variables
assessing the particularities of the individual. These would include the individual’s commute
times, levels of stress in the workplace, occupation, status in the workplace, ability to work
independently, socio-demographic variables, household variables such as: number of children,
marital status, gender and so forth.
Such detailed data do not exist and researchers have compromised on this front, either
surveying institutions (smaller populations), tailoring surveys to meet specific questions, but not
obtaining enough of a representative sample, or relying on regional or national datasets that
leave many pertinent variables out.
Past research on flexible workplace options generally falls into two categories: research
that focuses on why individuals (demand) adopt telework and research on why organizations
9
(supply) allow their employees to telework. The individual-focused research has mainly come
from the fields of transportation engineering and planning using the conceptual model of
Mokhtarian and Salomon (1994), which consider whether or not an individual teleworks based
on the constraints and drives she faces. Organization-focused research has been sparse and not as
theoretically consistent, mainly conducted in the fields of management and human resources. As
a whole, it has been understudied relative to research on why individuals adopt telework (Bailey
& Kurland, 2002).
The remainder of this paper conducts a more in-depth and critical review of the supply-
side and demand-side bodies of literature beginning with the supply-side of telework.
Supply-Side: Why do Organizations Offer Telework Options?
Organizations may provide the largest explanation for why telework is not more
‘formally’ prevalent in today’s digital world. As mentioned above, US Bureau of Labor Statistics
from 2003-2007 show that around 40% of the workforce works 4 or more hours at home per
week. Yet, the 2009 National Household Transportation Survey reveals that amongst working
individuals, only 14% of their organizations formally allow them to work at home. No doubt, the
reticence in offering formal telework options is due to internal resistance, which may be
explained by the difficulty of supervisorial duties; the belief that face-to-face collaboration is
more efficient, and security issues.
Face-to-face communication is theorized to be more effective and efficient than
computer-mediated communication as rich cues and substance are communicated (90% more
than the actual words being spoken) in the non-verbal portion of communication due to eye
contact, body language, inflections and pitch in voice (Mehrabian, 1981). The non-verbal portion
10
of communication is key to building trust and confidence and for tacit learning and cooperation,
all important factors in working teams (Rhoads, 2010).
Although smart technology and internet connections are ubiquitous, technological
tensions still persist. Central offices provide technical support; software and hardware that may
not be feasibly accessed remotely. Also, perceived security threats are an important consideration
for managers and an impediment to flexible workplace provision (Bonsall & Shires, Employer
expectations for commuting and business-related travel in an environment rich in information
and communication technologies, 2006).
So what about organizations that overcome the barriers associated with telework? These
organizations are overcoming a certain threshold of resistance to meet internal needs or to
respond to external influences.
Skilled labor required to meet the digital demands of the modern workplace is a
continuing issue for many organizations. The labor force required to sustain e-business must be
able to work with information retrieval, processing, and application in virtual settings. In 1999,
93% of information was produced digitally (Castells, 2003). Labor for e-business needs to be
self-programmable (individuals need to consistently reprogram their skills and knowledge) and
flexible. This type of labor, however, is in short supply. “From Silicon Valley to Stockholm, and
from England to Finland, the most important problem for leading companies has become where
to find engineers, computer programmers, e-business professionals, financial analysts, or, for that
matter, anyone with the capacity to develop new skills as required by changing markets”
(Castells, 2003, p. 93).
Teleworking can settle the e-business labor shortage to some extent as it expands a firm’s
labor network by allowing organizations to hire individuals from virtually any location. Offstein,
11
Morick & Koskinen (2010) observe that the benefit in employing telework in an organization
does not come from cost savings, but from allowing companies to create different models of
work to secure the best talent, which is a product of leadership, more than technology. The
ability to attract high performance and skilled knowledge workers is critical to a firm’s ability to
maintain competition in the new economy, and telework is seen as the top recruitment strategy
for groups aged 25 and younger, and groups aged 26-40 (Holtshouse, 2010).
Flexibility in the workplace engenders competitive advantage, and empirical research has
demonstrated that there is a link between firm performance and the adoption of flexible work
practices (Sánchez, Pérez, de Luis Carnicer, & Jiménez, 2007). Telework increases productivity
(reduced distractions and greater concentration), decreases absenteeism, and eliminates the costs
associated with relocation (Rotter, 1999; Tomaskovic-Devey & Risman, 1993; Gajendran &
Harrison, 2007). Employers cite that the main reasons for offering telework options are to
increase employee morale and recruit and retain employees (Moon, 2007).
An organization’s willingness to offer flexible work arrangements has been understudied
by transportation scholars. The small body of research has mainly been conducted by researchers
in the fields of management and human resources. These scholars have used the study of
telework to inform them of workplace evolutions, entrepreneurial activities, and innovation
trends.
The body of literature on organizational provision of telework suffers from theoretical
consistency. Many of the studies have used telework as a lens to study established theories in the
schools of business and management. What the studies share in common is using telework to
understand how organizations are competing in global and uncertain climates.
12
The theory of “fit” has long been used in the context of human resource management to
explain employment practices (variable pay) as responses to outside environments. These
employment practices are fundamentally contingent upon the particular needs of an organization.
Telework has been used as an instrument through which to test such assumptions (Mayo, Pastor,
Gomez-Mejia, & Cruz, 2009; Peters & Heusinkveld, 2010) as it has been theorized to enhance
the resources of an organization through increased dynamic capabilities (Sánchez, Pérez, de Luis
Carnicer, & Jiménez, 2007).
Under competitive climates, the particularities of firm needs may encourage non-
traditional human resource practices. These particularities have been tested through variables
such as firm size, age and geographical composition of its workforce. Smaller firms may be more
likely to adopt flexible workplace practices as they are likely to be less bureaucratized than
larger firms, and more able to adopt innovative Human Resource methods (Tomaskovic-Devey
& Risman, 1993; Mayo, Pastor, Gomez-Mejia, & Cruz, 2009). Smaller firms can face economic
difficulties through offering various incentives in place of higher salaries (Balkin & Gomez-
Mejia, 1987). Larger organizations may dedicate more resources to supervisory systems,
instituting more rules and regulations as it becomes increasingly challenging to monitor
individuals in larger firms (Mayo, Pastor, Gomez-Mejia, & Cruz, 2009).
Firm age may also affect a firm’s willingness to adopt flexible workplace practices.
Younger firms tend to have fewer financial resources and lack status. This may lead to a greater
willingness to adopt informal recruitment practices which expand their networks and make the
younger organizations more appealing (Leung, 2003; Mayo, Pastor, Gomez-Mejia, & Cruz,
2009). A younger organization is more likely to invoke employee participation in the design and
planning process of an organization, and flexible workplace practices tend to arise when
13
employees are part of this process (Sánchez, Pérez, de Luis Carnicer, & Jiménez, 2007) or the
organization is characterized by a flatter hierarchical structure (Peters, Tijdens, & Wetzels,
2004).
Studies examining why organizations adopt flexible workplace practices have mainly
been conducted in the Netherlands and Spain to test for ‘fit’ variables (Table 2). Overall, the
studies utilize relatively small samples, which are problematic when testing firm attributes such
as size and age, where not enough variation is present in the sample. Moreover, most of the
studies argue that innovative HR practices such as flexible workplace provision are methods
used to face uncertain climates, as they allow organizations to adjust employment models to
secure talent and make it more productive, yet, these studies do not adequately measure these
competitive climates. Measures of industry can serve as indications to competitive and uncertain
climates, yet, the studies do not account for industry sufficiently.
Mayo, Pastor, Gomez-Mejia, & Cruz (2009) used a sample of 122 Spanish firms,
surveying the CEOs of such firms to assess whether the attributes of the firm, particularly firm
size and age, revealed why the firm adopts telecommuting through discrete choice modeling.
Results conclude that smaller firms, with a high percentage of international employees, are more
likely to offer telecommuting and that firm size is inversely correlated with telecommuting, yet,
age is inconclusive. The study suffers from several drawbacks. Firstly, it limits the sample size to
firms employing more than 50 individuals. This excludes many smaller start-ups that may be
more inclined to adopt flexible workplace practices to compete, as they may be resource-limited.
As the study is specifically testing firm size, leaving out small firms and organizations limits the
ability to test whether smaller, less bureaucratized firms, adopt more informal practices. Industry
sector is only measured through a dummy variable signifying whether or not the firm is in the
14
services industry based on the assumption that this sector is more likely to adopt telecommuting,
because its work fits better with the use of IT. There are plenty of industries apart from the
service sector whose nature of work fits with the use of IT.
The Mayo et al study controls for competitive or growing industries through a measure of
the coefficient of variation on Return of Assets (ROA) to total invested capital for the industry.
This measure does not reveal the underlying distribution of an entire industry. Some
organizations within an industry may see high rates of return, while others lower; yet, all face the
same level of competition. The measure of ROA was also only captured over a 5-year time
period (2001-2005). If the sample size were large enough, capturing enough variation, this
measure could simply have been replaced by industry as a proxy for environmental conditions.
Moreover, how an industry uses, and is dependent upon technology, is a better measure of
competition and the need for the industry to innovate, as technologies evolve.
Peters and Heusinkveld (2010) performed a larger-scale study on 476 Dutch companies
and found that the larger a firm, the more likely it is to adopt telework options. This study did not
examine the provision of flexible workplace practices specifically, but rather how senior
management of the firms perceived telecommuting as thwarting, or enhancing productivity based
on the pressures of their respective communities. The study controlled for sector in the
assumption that beliefs are based on industry-wide commonalities, but only through 6 broad
categories such as services, manufacturing, building, trade, transport, logistics and
communications, financial and health, and well-being sectors.
Other studies find that firm size and age may not matter, and performance is more likely
to be relevant (Sánchez, Pérez, de Luis Carnicer, & Jiménez, 2007; Peters, Tijdens, & Wetzels,
2004). Peters, Tijdens, & Wetzels (2004), use a sample size of 849 individuals from the
15
Netherlands to measure their opportunity to telecommute, based on a host of characteristics from
their organizations and job types. The study neglects to include the industry sectors of the
individuals’ organizations and the individuals’ job types. It finds that the flatter the hierarchical
structure, the more likely the organization is to offer telework practices, but that the size of the
organization has no impact.
Sánchez, Pérez, de Luis Carnicer, & Jiménez (2007), also find that an organization’s size
does not matter in the provision of telecommuting. This study, however, only looks at small and
medium-sized firms located in Galicia and larger firms are not examined. Industry sector was not
accounted for in this study and telecommuting was not actually tested- the conclusions made in
this paper are based on t-statistics. The study found that firms adopting other forms of flexible
workplace practices, such as the ability to adjust start-times, are also likely to incorporate
telework.
Other studies have examined telework under the diffusion of innovation theory, proposed
by Rogers in 1962, to examine how telework, an innovation, is used to temper uncertain business
climates. Rogers proposes that new ideas and technology proliferate as they are communicated
through different channels of communication over time. Sia et al. (2004) and Karnowski and
White (2002), assess the perspectives of upper-level management on telework via their channels
of communication. Concomitantly, although not using the diffusion of innovation theory, Peters
and Heusinkveld (2010), examine institutional behavior through managerial perspectives as
influenced from institutional environments and normative beliefs.
16
Table 2:
Methodological Issues with Supply-Side Literature
Study Sample Size
and
Location
Dependent
Variable and
Method
Survey Independent Variables Methodological Issues
(Mayo, Pastor,
Gomez-Mejia,
& Cruz, 2009)
122 Firms
Spain
Whether or not a
firm adopts
telecommuting.
Logistic
Regression
Interviews of
CEOs
Firm Age
Firm Size
Number of non-Spanish employees
Contingent Reward Leadership
CEO Tenure
Service Industry
Past firm performance
Female Workforce
Firm Risk (coefficient of variation on Return of
Assets to Total Invested Capital for industry)
Small sample size
Sample limited to firms with
more than 50 employees
Firm industry risk assessed for
only 5 years (2001-2005)
Firm industry only examined
whether or not the firm was in
the service industry
Firm risk insufficiently measured
through ROA of industry
(Peters &
Heusinkveld,
2010)
476 Firms
Netherlands
Perceptions of
telework.
Factor Analysis
OLS
Interviews
with Senior
Management
Mimetic Pressures and Normative Pressures
Sector; Size; Educational Level of Workforce;
Output Rewards; Percent of Mobile Workforce;
Presence of Flexible Hours; Organizational
Culture
Small sample size
Perceptions not actual adoption
are measured
Sector is measured in only 6
categories
(Peters,
Tijdens, &
Wetzels, 2004)
849
Individuals
Netherlands
Opportunity to
telework,
preference and
practice
Logistic
Regression
Computer
Interviews of
individuals
Organizational Characteristics
Number of Business Localities; Size;
Organizational Hierarchy; Supervisorial
Status; Women in unit; Atmosphere
Job Characteristics
Computer Usage; Career Opportunities;
Internet Courses; IT skill
Household Characteristics
Individual Characteristics
Small sample size
Industry sector not assessed
Job occupation not assessed
(Sánchez,
Pérez, de Luis
Carnicer, &
Jiménez, 2007)
479 Small
and
Medium-
Sized Firms
Galicia,
Spain
Firm Performance
OLS
Surveys of
Managers
Use of Teleworking; Use of Flextime
Percent of Employees involved in the design and
planning process; Degree of flexible monitoring;
Percent of variable compensation;
Percent of subcontracting;
Degree of Spatial decentralization; Size
Small sample size
Degree of telework adoption was
not tested in a regression
Industry sector not measured
17
Summary of Supply-Side Literature
A major lacuna in the research on organization’s willingness to offer flexible workplace
practices is based on the premise that organizations are using innovation through employment
practices to better compete in competitive and uncertain climates. Most of the organization-
centered research has looked from within organizations interviewing managers and CEOs of a
handful of organizations attempting to project outward trends. No doubt this provides insights;
yet, the research should be supplemented by aggregate studies on how external forces
(competition, high growth climates) are shaping behaviors within organizations, particularly
those firms and organizations more dependent on technology. Organizations facing pressure to
innovate through technology, especially as some forms of technology rapidly evolve, face higher
and more rapid competitive pressures.
An organization’s adoption of telework is not only contingent upon their own culture, but
also on other structural factors such as the competitive environment of the industry, as well as,
larger corporate and national cultures (Peters & Heusinkveld, Institutional explanations for
managers' attitudes towards telehomeworking, 2010). Industries facing rapid growth and entry
are pressured to adjust their organizational models in order to maintain competitive. When folks
have knowledge of how other ‘successful’ organizations are behaving, they are more likely to be
influenced (DiMaggio & Powell, 1983). The organization-focused literature has not sufficiently
addressed how the external trends of industries are influencing organizations on a wider scale.
In general, the literature focused on why organizations offer telework lacks theoretical
consistency, mainly because the goals are different within each study and also, because telework
is used as an application to study established organizational theories. The studies principally
argue that telework allows certain firms to better compete in a global climate, expanding form
18
performance through employment practices. Yet, the larger, global competitive environment is
not assessed in these studies, only the specific attributes of small samples of firms that may not
be necessarily a consequence of the larger climate in which they operate.
The literature review now turns to the demand-side of flexible workplace practices.
Demand-side Literature: Individual Drives and Motives
Scholars from the fields of planning and transportation have mainly focused on the
demand-side of telework- why individuals telework or not. The literature tries to capture the
reasons individuals engage in telework and how often they do so. Most of the studies borrow
from a conceptual model of the decision factors involved in adopting telework proposed by
Mokhtarian and Salomon (1994), see Figure 1 below.
The model assumes that telework behavior is a product of constraints that prohibit people
from teleworking, facilitators that enable telework and drives that encourage telework. Drives are
considered as long-term objectives and ‘positive’ elements that aid a person in achieving certain
goals related to work, family and leisure time and are internal to the decision making process,
(Mannering & Mokhtarian, 1995). Constraints limit the ability to carry out preferences and are
exogenous to the decision- making process coming from the physical environment, other people
and institutions. They either act directly on the choice (precluding it to be made altogether), or
influence preferences. The empirical literature has sought to test and expand this initial
conceptual model.
19
Figure 1:
Telecommuting Adoption Model adapted from (Mokhtarian & Salomon, 1997)
The first wave of studies utilized stated preference surveys to examine the drives and
constraints affecting the preference to telework (see Bernardino, Ben-Akiva, & Salomon, 1993,
Mahmassani, Yen, Herman, & Sullivan, 1993 and Mokhtarian & Salomon, 1997). Following the
stated preference literature, arose many studies modeling whether or not a person teleworks and
how frequently they do so using demographic and household data, as well as observed work-
related variables, to determine the underlying attributes explaining adoption and frequency (see
Mannering & Mokhtarian, 1995, Drucker & Khattak, 2000, Ellen & Hempstead, 2002, Walls,
Safirova, & Jiang, 2007). Very few studies have analyzed workers who have the option to
telework, versus those who don’t (dependent on the organization), which is a pivotal dimension
to assessing actual telework practice (see Popuri & Bhat, 2003, Sener & Bhat, 2010).
Table 4 summarizes the revealed preference literature mentioned here and a more in-
depth discussion follows. Most importantly, the telework literature on the demand side suffers
from the use of small sample sizes that vary in locations. The Mannering and Mokhtarian study
20
from 1995 highlights the problems of comparing the results of small samples collected from
different locations. The study used data from three California state agencies: City of San Diego,
n=433, San Francisco Public Utilities Commission, n=90, and the Sacramento Franchise Tax
Board, n=90. These data have been used for a variety of studies following the initial (see
Mokhtarian & Salomon, 1996; Mokhtarian, et al., 1996; Vana, Bhat, & Mokhtarian, 2008).
Table 3 shows the various inconsistencies in variable significance across this study. The
analyses of these three data sets demonstrate that the particularities of one organization may
result in different behaviors, due to factors such as: workplace status, gender proportions, and
location. Income per capita positively influences both choice and frequency, only for
Sacramento. The ages of children in the household positively influence the choice to
telecommute, if the child is under five years old for San Diego, and if the child is under two
years old, for San Francisco. Length of time at present occupation positively affects choice for
San Francisco respondents, while length of time at present employer negatively affects choice.
Supervisory status pans out differently for San Diego and Sacramento. Supervisory status is not
significant for San Francisco respondents. Whether or not an individual is involved in clerical
work, only negatively affects San Diego employees.
21
Table 3:
Mannering and Mokhtarian (1995): Results of same survey applied to 3 California agencies
examining adoption and frequency of telecommuting
Variable San Diego (n=433) San Francisco (n=65) Bay Area (n=65)
Adoption Frequency Adoption Frequency Adoption Frequency
Number of people in Household + +
Income per capita in household + +
Female with children under 5 years +
Female with children under 2 +
Children under the age of 5 present + +
Home office +
Vehicles per Capita in household + +
Vehicles per licensed driver +
Length of time at present
occupation
+ +
Length of time at present employer -
Supervisor Status + +
Clerical occupation - -
Full-time +
Family orientation - - +
Key
Not
significant
+
Positive
relationship
-
Negative
relationship
The Mannering and Mokhtarian study suffers from other methodological issues as well.
Firstly, the authors employ a multinomial logit model on the frequency of telecommuting. In
order to do so the authors made three, discrete independent categories: never, infrequent (less
than once a month to 1 to 3 times a month) and frequent telecommuting (at least one day a
week). The separation of frequency into these three categories, greatly limits the range and
variation within which people telework. More importantly, the study does not assess constraints
sufficiently as it is more motivated by looking at how individual drives (willingness,
independence) influence telework. Left out of the models are key transportation constraints such
as the distance one has to commute to the workplace and the cost of living. Key workplace
constraints are also not considered, such as: a person’s occupation, status in the workplace, how
long a person has worked at the organization and a person’s age and gender (apart from whether
22
it is a female with a small child). Finally, the study only surveyed known telecommuters and did
not investigate individuals who did not telecommute.
Using a national sample, Drucker and Khatak (2000), determine that education
attainment and presence of children increase the propensity to work at home. This study utilized
the 1995 National Personal Transport Survey resulting in a sample size of 23,712 observations of
working individuals. The authors used ordered and unordered models to regress three categories
of work at home frequency (frequent, infrequent and never) on explanatory variables, including
age, gender, household characteristics, demographics and commute considerations. The study did
not examine choice directly. Individuals were not separated by whether or not they had the
option to work at home; rather, individuals who never worked at home were combined as the
base category. Occupation was not considered in the study.
Walls, Safirova and Jiang (2007) is the only study to look at industry and occupation
comprehensively. In general, the study determines that workers in jobs that are more conducive
to telecommuting (sales, architecture and engineering and education and training) are more likely
to telecommute than workers in jobs that are less conducive to telecommuting (healthcare and
construction). The authors use the size of one’s organization to explain propensity as well and
find that individuals employed in smaller firms are more likely to telecommute.
The findings are slightly suspicious, because around 25% of individuals in the sample
telecommuted at least one in the past 2 months. This is contrary to most studies, which find that
less than 10% of the sampled population actually conducts remote work. The 2009 National
Household Transportation Survey (NHTS) reveal that only 8% of the national working
population works 1 or more days a month at home. The large percentage in the Walls et al. study
is due to the fact that telecommuting is defined at three levels: if a person works at home, if a
23
person works at a telecenter , and, if a person works at a satellite office. The inclusion of work at
a satellite office most likely inflates the numbers and is not considered technically remote work.
The Walls, Safirova and Jiang (2007) study also does not appropriately separate job
occupations and industries for those more conducive to telecommuting. For example,
information services are grouped with public relations and customer services. Information
services are not examined at the industrial level either. Finally, choice is not examined. Grouping
all individuals together to examine likelihood (whether or not they have the choice, but rather,
whether or not they actually telecommuted in the last 2 months) is to assume that all individuals
have the same option to conduct flexible workplace practices.
Whether or not one has the choice to telework reveals information about the
organizations allowing/disallowing them to conduct flexible work, informing the supply-side of
telework. Whether or not one has the choice to telework, may inform how frequently she
teleworks (Sener & Bhat, 2010). For instance, those who are allowed to telework may be more
technically literate and will influence their frequency of teleworking. The latter stream of
telework literature has jointly modeled this process.
24
Table 4:
Methodological Issues with Demand-Side Literature
Study Sample Size
and
Location
Dependent
Variable and
Method
Independent Variables Methodological Issues
(Mannering &
Mokhtarian,
1995)
65;65;433
San
Francisco;
Sacramento;
San Diego
Frequency of
Telecommuting
Ordered
Response
Number of People in HH; Female w/small children; Home
office space; Vehicles per capita; Automobile as a status
symbol; Hours worked; Supervises others; Clerical occupation;
Full-time; Level of control over work; Productivity in the
workplace; Familiarity with Telecommuting; Lack of self-
discipline; Family orientation; Satisfaction with Life
Frequency separated as Never, Infrequent,
Frequent
Key constraints not assessed: Distance from
the Workplace Occupation, Status in the
workplace, Time at the workplace, Age and
Gender
Only telecommuters were surveyed
(Drucker &
Khattak,
2000)
23,712
Nation
Frequency of
Work at home
Ordered and
Unordered
Models
Age; Gender; Single; Age of children in HH; Education; HH
Income; Driver?; Number of Vehicles; Time to Work; Rural?;
Pays to Park; Availability of Bus, Train, Streetcar and Rail
Choice is not examined, individuals are all
grouped together regardless of choice
Occupation and job status not included
Frequency separated as Never, Infrequent,
and Frequent
(Walls,
Safirova, &
Jiang, 2007)
2,315
Southern
California
Likelihood and
Frequency of
Telecommuting
Probit and
Ordered Probit
Age less than 30; No College; White; Kids between the ages of
0 and 5 and 6 and 17; Gender; Part-time or Full-time; 11
categories for organizations industry; 11 categories for
worker’s occupation; Size of firm; Commute
Regional Sample of the Southern California
Income is not examined
Education examined through no college
Job types and industries are not separated
enough to examine which ones are more
reliant on technology
Choice is not examined
(Popuri &
Bhat, 2003)
6,523 and
1,018
New York
Choice and
Frequency of
Telecommuting
Binary and
Ordered
Response
Gender; Female w/children; Age; Marital Status; Education;
Drives to Work?; Many Vehicles?; Licensed Driver; Takes
Transit to Work?; Works for a private company?;F2F contact is
needed at work?; Part-time status; Pays to park at work; Length
of employment; Income; Fax machine?; Multiple phone lines
Regional sample
Occupation is not examined
Status in the workplace is not examined
Distance to work is not examined
Frequency assessed through ordered groups
(Sener &
Bhat, 2010)
9,264 and
1,534
Chicago
Choice and
Frequency of
Telecommuting
Binary and
Ordered
Response
Gender; Female w/children; Younger than 30 years; Education;
Driver’s license; Full-time; Workplace flexibility; Sector:
Communications Service portions of Finance, Real Estate,
Professional, Scientific or Technical, Management, Arts,
Education and Health Care, Government; Income between 75
and 100K; Income > 100K; # of vehicles, #of workers in HH;
Commute > 25 miles; Walk, bike or take transit to work
Regional sample
Sector is mainly examined through services
Status in the workplace not examined
Age not properly accounted for
Income not properly accounted for
Frequency assessed through ordered groups
25
Only a few studies have examined choice, comparing individuals who have the choice to
telework, versus those who don’t (see Popuri & Bhat, 2003, Sener & Bhat, 2010). For these two
studies choice has been analyzed separately through binary modeling and frequency as
contingent upon choice through ordered response modeling. Popuri & Bhat (2003) use a sizeable
sample from the New York region and Sener and Bhat (2010) from the Chicago region. These
studies make progress in analyzing the underlying processes of why people are allowed to
telework, yet, there are a few drawbacks. Although the sample sizes are adequately large, they
still come from regions not allowing for a larger, universal pattern to be detected. Both of the
studies do not measure the frequency of telecommuting on a continuous scale, but rather as
ordered groups. Popuri & Bhat (2003) do not address occupation or status in the workplace.
Sener & Bhat (2010) do address occupation, but not in a straightforward manner. Individuals in
the manufacturing, transportation, retail and other occupations are treated as the base (the authors
believe workers must travel to these occupations). Communications and Government related jobs
are given their own dummy variables and for all other occupations only the service related jobs
are included as dummies. The study also does not measure age or income on a continuous scale
(only individuals under 30 and those earning between 75 and 100K and over 100K).
The variables described are transportation-related variables (commute distances and
vehicle availability); occupation and industry; type of work (part-time or contracted); status in
the workplace; socio-demographics (gender, presence of children, education and income).
For each of the constraints and facilitators tested in one study there is concomitantly a
contradictory one. This is due to the fact that the studies employ differing sample sizes, times,
and locations of the samples and to the methodological issues mentioned above. The studies as a
whole only come to the conclusion that income, education and type of work (part-time or
26
contracted) are significantly and positively correlated with telework uptake and frequency of use.
These are measures of skill and status in the workplace further reinforcing the notion that
occupation and organizational culture are understudied in the individual-centered telework
research.
Summary of Demand-Side Literature
The above section reviews telework demand studies conducted over the past two decades.
The review highlights the mixed results of demand models based on surveys that are location
specific, vary in size and measurements. Telework seems to be contingent upon occupation
(although not clearly understood), type of work, income and education. Overall, there seems to
be evidence that part-time, self-employed and contract work, income, and education, are
associated with telework uptake and frequency. These factors indicate that flexible forms of
employment are correlated with uptake and frequency, and that the higher skilled, as determined
by levels of income and education, are more likely to telework.
Using models of constraints and drives to assess why individuals adopt telework is a
useful platform to understand teleworking trends. These models must contain pertinent
information however, and be conducted at scales through which larger generalizations can be
made. Occupation and organizational culture have been understudied in the individual-focused
literature and this information is pivotal for drive and constraint models. Very few studies have
examined the reasons for which some individuals have the choice to telework, while others
don’t.
As mentioned previously, the percentage of the workforce ‘formally’ allowed the option
to telework is small, yet, more people want to telework and around 37% of the non-self-
employed workforce works 4 or more hours at home per week. This means that the lack of
27
formal flexible workplace provision lies within organizations. More concrete and in-depth
analyses of individuals including their occupations and industries and their telework options
would reveal trends occurring in industries and what types of occupations are more amenable to
flexible work. Such occupational and industrial analyses need to occur at larger scales. Smaller-
scale studies lack enough variation to reveal widespread trends. Even a larger-scale study at a
singular region only communicates what is happening at the regional level. A national scale
study would be the most informative on flexible workplace practices and occupational trends.
Conclusion Literature Review
The study of flexible workplace options, namely the ability to conduct work remote from
the workplace, reveals trends about the organizations and industries accommodating flexibility in
the workplace and the demographics accessing flexible options. Each of these facets is important
for understanding regional competition, how to accurately forecast and pinpoint urban economic
trends and imagine how ICT will continue to shape our urban environments.
The literature mainly comes to a consensus that the highly educated and skilled workers
are those accessing flexible workplace options. Yet, this is only part of the story. Larger
workplace forces are at play and the literature has not examined sufficiently the interplay
between individuals and their places of work through larger-scale datasets. Whether or not
telework will increase or decrease, may be inconsequential. ICTs are naturally transforming the
way in which we engage in activities, particularly work activities, and will continue to morph the
traditional patterns. Understanding this new world order is necessary for planners considering the
future of the spaces they are tasked to assess. Studying flexible workplace options is an avenue
for understanding how ICTs are influencing the status quo.
28
Most of the demand and supply-side studies examining single institutions only reveal the
trends occurring at the institution surveyed. Regional studies inform us of trends occurring at the
regional level. Regions and their industrial or demographic trends are also influenced by a larger
national and global landscape. To truly learn about the workplace and its relationship with
industry, it is necessary to compare regions or areas.
In order to continue exploring the interactions between ICT and behavior, or how ICTs
highlight engrained motivations, better data and better theories are needed. A study of an
institution, small group of people, or a city, tells us only about that institution, small group of
people, or city. One or two-day travel diaries tell us only about what that individual did on the
given day of her survey. Much caution should be taken in generalizing results. National datasets
become increasingly important and many of our national datasets do not combine questions on
work, travel and occupation sufficiently. The burden and cost of data collection will always be a
problem, and focusing on the local is usually more feasible than large-scale studies, and thus,
questions must be matched appropriately with available data and with desired outcomes. In the
era of big data generated from mobile devices, passive GPS devices, accelerometers, consumer
behavior, employment behavior, and so forth, more sophisticated and thorough analyses are
possible.
29
References Literature Review
Alexander, B., Ettema, D., & Dijst, M. (2010). Fragmentation of work activity as a multi-
dimensional construct and its association with ICT, employement and sociodemographic
characteristics. Journal of Transport Geography , 18, 55.
Asensio, J. (2002). Transport mode choice by commuters to Barcelona's CBD. Urban Studies ,
39, 1881-1895.
Bagley, M., & Mokhtarian, P. (1997). Analyzing the preference for non-exclusive forms of
telecommuting: Modeling and policy implications. Transportation , 24, 203-226.
Bailey, D., & Kurland, N. (2002). A review of telework research: Findings, new directions, and
lessons for the study of modern work. Journal of Organizational Behavior , 23, 383-400.
Balkin, D., & Gomez-Mejia, L. (1987). Toward a contingency theory of compensation strategy.
Strategic Management Journal , 8, 169-182.
Bernardino, A., Ben-Akiva, M., & Salomon, I. (1993). A stated preference approach to modeling
the adoption of telecommuting. Transportation Reseach Record: Journal of the Transportation
Research Board, No. 1413 , 22-30.
Bonsall, P. (2004). Traveller behavior: Decision making in and unpredictable world. Journal of
Intelligent Transport Systems , 8 (1), 45-60.
Bonsall, P., & Shires, J. (2006). Employer expectations for commuting and business-related
travel in an environment rich in information and communication technologies. Transportation
Research Record: Journal of the Transportation Research Board, No. 1977 , 268-276.
Brannen, J. (2005). Time and negotiation of work-family boundaries. Time and Society , 14, 113-
131.
Breedveld, K. (1998). The double myth of flexibilization: trends in scattered work hours, and
differences in time-sovereignty. Time and Society , 7, 129-143.
Cameron, S. (1993). Why is the R squared adjusted reported? Journal of Quantitative Economics
, 183-186.
Castells, M. (2003). The Internet Galaxy: Refections on the Internet, Business and Society.
Oxford: Oxford University Press.
Chatman, J., & Jehn, K. (1994). Assessing the relationship between industry characteristics and
organizational culture: How different can you be? The Academy of Management Journal , 37,
522-553.
30
Couclelis, H. (2003). Housing and the new geography of accessibility in the information age.
Open House International , 28, 7-13.
Covin, J., Slevin, D., & Covin, T. (1990). Content and preformance of growht-seeking strategies:
A comparison of small firms in high and low-technology industries. Journal of Business
Venturing , 5, 391-412.
Crampton, S., & Hodge, J. (2009). Generation Y: Unchartered territory. Journal of Business and
Economic Research , 7.
Cullen, I., & Godson, V. (1975). Urban networks: The structure of activity patterns. Progress in
Planning , 4, 1-96.
Cullen, I., Godson, V., & Mayor, S. (1972). The structure of activity patterns. In A. Wilson,
Patterns and Processes in Urban and Regional Systems (p. 286). London: Pion.
DiMaggio, P., & Powell, W. (1983). The iron cage revisited: Institutional isomporhism and
collective reationality in organizational fields. American Sociological Review , 48, 147-160.
Drucker, J., & Khattak, A. (2000). Propensity to work from home: Modeling results from the
1995 Nationwide Personal Transportation Survey. Transportation Research Record: Journal of
the Transportation Research Board, No. 1706 , 108-117.
Ellen, I., & Hempstead, K. (2002). Telecommuting and the demand for urban living: a
preliminary look at white-collar workers. Urban Studies , 39, 749.
Farag, S., Dijst, M., & Lazendorf, M. (2003). Exploring the use of e-shopping and its impact on
personal travel behavior in the Netherlands. Transportation Research Record 1858 , Paper No.
03-3058, 47-58.
Faulconbridge, J., & Beaverstock, J. (2008). Geographies of international business travel in the
professional service economy. In D. Hislop, Mobility and technology in the workplace (pp. 87-
102). Abingdon: Routledge.
Florida, R., Gulden, T., & Mellander, C. (2088). The rise of the mega-region. Cambridge
Journal of Regions, Economy and Society , 1, 459-76.
Frank, L. (2000). Land use and transportation interaction: Implications on public health and
quality of life. Journal of Planning Education and Research , 20 (6).
Frank, L., Bradley, M., Kavage, S., Chapman, J., & Lawton, T. (2008). Urban form, travel time,
and cost relationships with tour complexity and mode choice. Transportation , 35, 37-54.
31
Gao, S., Mokhtarian, P., & Johnston, R. (2008). Exploring connections among job accessibility,
employment, income and auto ownership using structural equation modeling. Annals of Regional
Science , 42, 341-356.
Garhammer, M. (1995). Changes in working hours in Germany: The resulting impact of daily
life. Time and Society , 4, 167-203.
Giuliano, G. (2003). Travel, location and race/ethnicity. Transportation Research Part A: Policy
and Practice , 37, 351-372.
Gujarati, D. (2004). Basic Econometrics 4th Edition. The McGraw-Hill Companies.
Handy, S. (2005). Smart growth and the transportation land-use connection: What does the
research tell us? International Regional Science Review , 28, 146-167.
Handy, S., & Mokhtarian, P. (1995). Planning for telecommuting: Measurement and policy
issues. Journal of the American Planning Association , 61, 99-111.
Handy, S., & Mokhtarian, P. (1996). The future of telecommuting. Futures , 28, 227-240.
Handy, S., & Mokhtarian, P. (1996). The future of telecommuting. Futures , 28 (3), 227-240.
Hanson, S., & Pratt, G. (1995). Gender, work and space. London: Routledge.
Haynes, B. (2011). The impact of generational differences on the workplace. Journal of
Corporate Real Estate , 13 (2), 98.
Hipple, S. (2010). Self-Employment in the United States. Monthly Labor Review. Bureau of
Labor Statistics.
Hjorthol, R. (2002). The relation between daily travel and use of the home computer.
Transportation Research PArt A , 36, 437-452.
Hjorthol, R., & Gripsrud, M. (2009). Home as a communication hub: The domestic use of ICT.
Journal of Transport Geography , 17, 115-123.
Holtshouse, D. (2010). Knowledge work 2020: Thinking ahead about knowledge work. On the
Horizon , 18 (3), 193-203.
Hubers, C., Dijst, M., & Schwanen, T. (2008). ICTs coping strategies and gender differences in
the temporal and spatial fragmentation of paid labor. In: Paper presented at the Fourth
International Specialist Meeting on ICT, Everyday Life and Urban Change, Bristol, UK.
Janelle, D. (1986). Metropolitan expanision and the communications-transportation tradeoff. In
S. Hanson, The geography of urban transportation (pp. 357-385). New York: The Guilford
Press.
32
Janelle, D., & Hodge, D. (2000). Information, Place and Cyberspace Issues in Accessibility.
Berlin: Springer.
Jang, T. (2005). Cout data models for trip generation. Journal of transportation engineering ,
131, 444-450.
Karnowski, S., & White, B. (2002). The role of facility managers in the diffusion of
organizational telecommuting. Environment and Behavior , 34, 322-24.
Kenyon, S., & Lyons, G. (2007). Introducing multitasking to the study of travel and ICT:
Examining its extent and assessing its potential importance. Transportation Research A , 41,
161.
Khandwalla, P. (1976). Some top management styles, their context and performance.
Organization and Administrative Sciences , 7 (4), 21-51.
Kitamura, R., Mokhtarian, P., & Laidet, L. (1997). A micro-analysis of land use and travel in
five neighborhoods in the San Francisco Bay Area. Transportation , 24, 125-158.
Koskela, H. (1999). "Gendered exclusions':Women's fear of violence and changing relations to
space. Geografiska Annaler B, Human Geography , 81, 111.
Kwan, M.-P. (2007). Mobile communications, social networks, and urban travel: New topoligies
of spatial interaction. The Professional Geographer , 59 (4), 434-446.
Learner, E., & Storper, M. (2001). The economic geography of the internet age. Journal of
International Business Studies , 32, 641-65.
Lenz, B., & Nobis, C. (2007). The changing allocation of activities in space and time by the use
of ICT- "Fragmentation" as a new concpet and empirical results. Transportation Research Part A
, 41 (2), 190-204.
Leung, A. (2003). Different ties for different needs- recruitment practices of entrepreneurial
firms at different development phases. Human Resources Management , 42 (4), 303-320.
Line, T., Jain, J., & Lyons, G. (2011). The role of ICTs in everyday mobile lives. Journal of
Transport Geography , 19, 1490-1499.
Littlejohn, S., & Foss, K. (2005). Theories of human communciation 7th ed. Belmont, CA:
Wadsworth.
Lyons, G., & Urry, J. (2005). Travel time use in the Information Age. Transportation Research A
, 39, 257-276.
Madge, C., & O'Connor, H. (2006). Parenting gone wired: empowerment of new mothers on the
Internet? Social and Cultural Geography , 7 (2), 199-220.
33
Mahmassani, H., Yen, J., Herman, R., & Sullivan, M. (1993). Employee attitudes and stated
preferences towards telecommuting: an exploratory analysis. Transportation Research Record:
Journal of the Transportation Research Board, No. 1413 , 31-44.
Mannering, J., & Mokhtarian, P. (1995). Modeling the choice of telecommuting frequency in
California: An exploratory analysis. Technological Forecasting and Social Change , 49, 49-73.
Martins, L., Gibson, L., & Maynard, M. (2004). Vitual teams: What do we know and where do
we go from here? Journal of Management , 30, 805-35.
Mayo, M., Pastor, J., Gomez-Mejía, L., & Cruz, C. (2009). Why some firms adopt
telecommuting while others do not: A contingency perpective. Human Resource Management ,
48 (6), 917-939.
McQuail, D. (2000). Mass communication theory. London: Sage.
Mehrabian, A. (1981). Silent messages: Implicit communication of emotions and attitudes.
Belmont, CA: Wadsworth.
Miller, H., & Shaw, S.-L. (2001). Geographic Information Systems for Transportation:
Principles and Applications. New York: Oxford University Press.
Mokhtarian, P. (2004). Telecommuting, residential location, and commute-distance traveled:
evidence from State of California employees. Environment and Planning A , 36, 1877-1897.
Mokhtarian, P., & Salomon, I. (1996). Modeling the choice of telecommuting: 3. Identifying the
choice set and estimating binary choice models for technology-based alternatives. Environment
and Planning A , 28, 1877-1894.
Mokhtarian, P., & Salomon, I. (1994). Modeling the choice of telecommuting: Setting the
context. Environment and Planning A , 28, 1859-1876.
Mokhtarian, P., & Salomon, I. (1994). Modeling the choice of telecommuting: Setting the
context. Environmental Planning A , 26, 749-766.
Mokhtarian, P., & Salomon, I. (1997). Modeling the desire to telecommute: the importance of
attitudinal factors in behavorial models. Transportation Research A. , 30 (1), 35-50.
Mokhtarian, P., Handy, S., & Salomon, I. (1995). Methodological issues in the estimation of
travel, energy and air quality impacts of telecommuting. Transportation Research A , 29, 283-
302.
Mokhtarian, P., Salomon, I., Saxena, S., Sampath, S., Cheung, P., Le, K., et al. (1996). Adoption
of Telecommuting in Two California State Agencies. Research Report UCD-ITS-RR-96-5,
Institute of Transportation Studies, University of California, Davis, September.
34
Moon, N. (2007). Private sector telework and its implications for economic development:
Results of a case survey and policy assessment. Atlanta, Georgia: Enterprise Innovation Institute
and School of Public Policy's Science, Technlology, and Innovation Policy (STIP) Program
Georgia Institute of Technology.
Newman, D. (2012, May 22). 6 Technology trends transforming the workplace. Retrieved June
28, 2013, from millenialceo.com: http://millennialceo.com/leadership/6-technology-trends-
transforming-workplace/
Odih, P. (1999). Gendered time in the age of deconstruction. Time and Society , 8 (1), 9-38.
Offstein, E., Morwick, J., & Koskinen, L. (2010). Making telework work: Leading people and
leveraging technology for competitive advantage. Strategic HR Review , 9 (2), 32-37.
Olson, G., & Olson, J. (2000). Distance matters. Human Computer Interaction , 15, 130-79.
Olszewski, P., & Mokhtarain, P. (1994). Telecommuting frequency and impacts for State of
California employees. Technological forecasting and social change , 45, 275-286.
Pain, R. (1997). Social geogrphies of women's fear. Transactions of the Institute of British
Geographers , 22, 231.
Pennings, J., & Gresov, C. (1986). Technological networking and innovation implementation.
Organization Science , 3, 317-334.
Peters, P., & Heusinkveld, S. (2010). Institutional explanations for managers' attitudes towards
telehomeworking. Human Relations , 63, 107.
Peters, P., Tijdens, K., & Wetzels, C. (2004). Employees' opportunities, preferences, and
practices in telecommuting adoption. Information and Management , 41 (2), 469-482.
Piccoli, G., & Ives, B. (2003). Trust and the unintended effects of behavior control in virtual
teams. MIS Quarterly , 27, 365-95.
Pickup, L. (1988). Hard to get around: a study of women's travel mobility. In J. Little, L. Peake,
& P. Richardson, Gender and the Urban Environment (pp. 98-116). New York: New York
University Press.
Pinjari, A., Pendyala, R., Bhat, C., & Waddell, P. (2007). Modeling residential sorting effects to
understand the impact of the built environment on commute mode choice. Transportation , 34,
557-573.
Popuri, Y., & Bhat, C. (2003). On modeling choice and frequency of home-based
telecommuting. Transportation Research Record: Journal of the Transportation Research
Board, No. 1858 , 55-60.
35
Rheingold, H. (2003). Smart Mobs: The Next Social Revolution. New York: The Perseus Book
Group.
Rhoads, M. (2010). Face-to-face and computer-mediated communication: What does theory tell
us and what have we learned so far? Journal of Planning Literature , 25 (2), 111-122.
Sánchez, A., Pérez, M., de Luis Carnicer, P., & Jiménez, M. (2007). Teleworking and workplace
flexibility: A study of impact on firm performance. Personnl Review , 36 (1), 42-64.
Schrank, D., Lomax, T., & Eisele, B. (2011). 2011 Urban Mobility Report. College Station, TX:
Texas Transportation Institute.
Schwanen, T. (2007). Matter(s) of interest: Artefacts, spacing, timing. GeografiskaAnnaler,
Series B: Human Geography , 89 (1), 9-22.
Schwanen, T. (2006). On 'arriving on time', but what is 'on time'? Geoforum , 37 (6), 882-894.
Schwanen, T., & Dijst, M. (2003). Time windows in workers' activity patterns: Empirical
Evidence from the Netherlands. Tranpsortation , 30 (3), 261-283.
Schwanen, T., & Kwan, M.-P. (2008). The Internet, mobile phone and space-time constraints.
Geoforum , 1362-1377.
Schwanen, T., & Mokhtarian, P. (2005). What affects commute choice: Neighborhood physical
structure or preferences toward neighborhoods? Journal of Transportation Geography , 13, 83-
99.
Schwanen, T., Dielema, F., & Dijst, M. (2004). The impact of metropolitan structure on
commute behavior in the Nethwrlands: a multilevel approach. Growth Change , 35, 304-333.
Schwanen, T., Kwan, M.-P., & Ren, F. (2008). How fixed is fixed? Gendered rigidity of space-
time constraints and geographies of everyday activities. Geoforum , 39, 2109-2121.
Schwanen, T., Pokwan, M., & Ren, F. (2008). How fixed is fixed? Gendered rigidity of space-
time constraints and geographies of every day activities. Geoforum , 39, 2109-2121.
Sener, I., & Bhat, C. (2010). A copula basedsample selection model of telecommuting choice
and frequency. Environment and Planning A , 43 (1), 126-145.
Senju, A., & Johnson, M. (2009). The eye contact effect: Mechanisms and development. Trends
in cognitive sciences , 13, 127-34.
Sia, C., Teo, H., Tan, B., & Wei, K. (2004). Effects of environmental uncertainty on
organziational intention to adopt distributed work arrangements. IEEE Transactions on
Engineering Management , 51, 256-67.
36
Slattery, M. (2013, March 7). How consumer technology is remaking the workplace. Retrieved
June 28, 2013, from wired.com: http://www.wired.com/insights/2013/03/how-consumer-
technology-is-remaking-the-workplace/
Susilo, Y., & Maat, K. (2007). The influence of the built environment to the trends in commuting
journeys in the Netherlands. Transportation , 34, 589-609.
Swanberg, J., Pitt-Catsouphes, M., & Drescher-Burke, K. (2005). A question of justice:
disparities in employees' access to flexible schedule arrangements. Journal of Family Issues , 26,
866.
Tang, W., Mokhtarian, P., & Handy, S. (2008). The role of neighborhood characteristics in the
adoption and frequency of working at home: Empirical evidence from Northern California.
(Paper, UCD-ITS-RR-08-21). UC Davis: Institute of Transportation Studies.
Tomaskovic-Devey, D., & Risman, B. (1993). Telecommuting innovation and organization: a
contingency theory of labor process change. Social Science Quarterly , 74, 367-385.
Tulgan, B., & Martin, C. (2001, October 3rd). Managing generation Y- part 2. Business Week
Online. Retrieved June 13th, 2012, from http://www.businessweek.com/stories/2001-10-
03/managing-generation-y-part-2
Valentine, G. (1989). The geography of women's fear. Area , 21, 385.
Valentine, G., & Holloway, S. (2002). Cyberkids? Exploring children's identities and social
networks in on-line and off-line worlds. Annals of the Association of American Geographers , 92
(2), 302-319.
Vana, P., Bhat, C., & Mokhtarian, P. (2008). On modeling the choice of work-hour arrangement,
location and frequency of telecommuting.
Wajcman, J. (2010). Feminist theories of technology. Cambridge Journal of Economics , 34,
143-152.
Walls, M., Safirova, E., & Jiang, Y. (2007). What drives telecommuting? The relative impact of
worker demographics. employer characteristics, and job types. Transportation Research Record
Journal of the Transportation Research Board, No. 2010 , 111-120.
Wiggins, S. (1995). Entrepreneurial enterprises, endogenous ownership, and the limits to firm
size. Economic Inquiry , 33, 54-69.
Yeraguntla, A., & Bhat, C. (2005). A classification taxonomy and empirical analysis of work
arrangements. Transportation Research Record: Journal of the Transportation Research Board,
No. 1926 , 233-241.
37
Yu, H., & Shaw, S.-L. (2007). Revisiting Hägerstrand's Time-Geographic Framework for
Individual Activities in the Age of Instant Access. In H. Miller, Societies and Cities in the Age of
Instant Access (pp. 103-118). Dordrecht, Netherlands: Springer.
38
Paper 1: Regional variation in the provision and uptake of flexible workplace options
Abstract Paper 1
The relationship between occupation, industry norms and flexible workplace options such
as the ability to work at home, or “telework” is fundamental to understanding how flexibility is
absorbed in the workplace and practiced by individuals. However, few studies have explored this
relationship in-depth. Rather, research has focused on the underlying attributes of small sample
sizes of individuals, or organizations, to explain the uptake and provision of flexible workplace
options. Using a national dataset from the 2009 National Household Transportation Survey, I
investigate how the occupational mix of a region is associated with the prevalence of flexible
workplace practices.
Results suggest that regions with a greater share of workers, who have the ability to work
at home, also have higher proportions of workers in high-tech and managerial occupations.
Regions with more highly educated workforces also see higher rates of telework provision.
Flexible workplace practices are innovative. They signal that an organization is
purposefully accommodating the needs of its workforce, or that the organizations are using the
practices to better compete in a global economy. The high-tech industry faces rapid growth and
entry, causing organizations to adjust their models in order to remain competitive. Flexible
workplace practices may therefore reflect this accommodation.
Introduction Paper 1
Recent advances in information and communication technologies (ICTs) have provided
individuals with more choices than ever, particularly in the domain of work. Now, many
individuals have choices as to when, where and how work is conducted, resulting in
unprecedented flexibility. Workplace flexibility (work at home, or flexible start times), can bring
many benefits to both individuals and organizations. In terms of the individual, the ability to
39
work from home, or nearer to home and the option to adjust start times, can help to
accommodate greater work-life balance, alleviate commute burdens and enhance work
productivity, by allowing the individual to better concentrate. Organizations can benefit from
workplace flexibility through alleviating office costs associated with work- space and through
the increased ability to secure needed talent from a wider geography, or, in other words ‘whoever
is the best person for the job.’ Offering workplace flexibility is an incentive to employees,
enhancing the appeal of organizations.
Workplace flexibility, particularly the ability to work at home, has interested scholars for
decades. Transportation scholars have sought to understand how working at home, or closer to
home (telecommuting or telework), can be used as an alternative to commuting during peak-
hours. These studies explore the attributes that explain who teleworks (Mokhtarian & Salomon,
1996; Popuri & Bhat, 2003; Walls, Safirova, & Jiang, 2007). From business and management
fields, scholars have attempted to pinpoint the characteristics of firms that offer telework as an
option to their employees (Tomaskovic-Devey & Risman, 1993; Mayo, Pastor, Gomez-Mejia, &
Cruz, 2009; Leung, 2003).
Telework provision and uptake manifest when the supply of telework meets the demand
for telework (Handy & Mokhtarian, 1996). The supply of telework relates to whether or not the
type of work can be feasibly conducted remotely and the organizations’ willingness to allow for
remote work. The demand for telework generally refers to an individual’s attributes, which drive
or constrain her to telework.
The drives and constraints tested in the demand-side (or individual-focused) literature
have sought to characterize those who seek and use telework in terms of socio-demographic
variables relating to age, income and levels of education (Ellen & Hempstead, 2002; Sener &
40
Bhat, 2010; Popuri & Bhat, 2003); commute-related variables (Tang, Mokhtarian, & Handy,
2008; Mokhtarian, et al., 1996; Sener & Bhat, 2010); and work-related variables such as status
(Mokhtarian & Salomon, 1996; Ellen & Hempstead, 2002; Walls, Safirova, & Jiang, 2007) and
occupation (Vana, Bhat, & Mokhtarian, 2008; Yeraguntla & Bhat, 2005; Sener & Bhat, 2010).
The organization-focused (or supply-side) literature borrows from established
organizational theories to test how telework explains employment recruitment processes in
uncertain business climates employing theories of ‘fit’ (how aspects of a firm engender particular
needs in an environmental context) and diffusion of innovation (how innovation is
communicated through channels). These studies have tested how age, size and geographical
composition of the organization’s labor force affects levels of telework provision ( (Mayo,
Pastor, Gomez-Mejia, & Cruz, 2009; Moon, 2007; Sánchez, Pérez, de Luis Carnicer, & Jiménez,
2007; Peters & Heusinkveld, Institutional explanations for managers' attitudes towards
telehomeworking, 2010), as well as managerial perceptions (Peters & Heusinkveld, Institutional
explanations for managers' attitudes towards telehomeworking, 2010; Sia, Teo, Tan, & Wei,
2004).
Research examining the supply of flexible work arrangements is mixed. The propensity
to offer flexible work is associated with the competitive environment firms face in securing
talent. Most of the studies looking at why organizations adopt flexible workplace practices have
not sufficiently considered this competitive climate. Use of flexible work is associated with
status in the workplace and the type of work one conducts amongst other personal attributes. The
literature examining why individuals adopt flexible workplace practices has not sufficiently
considered individual’s occupations and status in the utility models. Both bodies of literature
have relied on small sample sizes rendering results un-generalizable.
41
This paper addresses these issues by more closely assessing, on a large scale, how the
nexus between the individual, her occupation and her organization interact, utilizing a national
dataset. Large-scale datasets that survey an individual’s teleworking behaviors and her
occupation do not exist. The National Household Travel Survey (NHTS) for 2009 is the most
appropriate national dataset examining flexible workplace practices as it asks individuals if they
have the option to work at home. Yet, it does not sufficiently address occupation, using only 4
broad categories. This paper resolves the issue by analyzing the 2009 NHTS data at the regional
level. A region’s mix of occupations serves as a proxy for prevalent industry sectors in the
region, while proportions of demographics such as age and education serve as a proxy for types
of workers within the region.
Cursory evidence suggests that opportunities for flexible work place practices vary across
metropolitan areas. A quick glance at the NHTS data reveals that there is an uneven distribution
of employed individuals who are offered the option to telework across the 51 US metropolitan
regions (see Figure 2). The highest percentage of workers offered work at home options lies in
San Jose, CA, at around 31%, while the lowest lies in Las Vegas, NV, at around 6%.
42
Figure 2:
Option to Work at Home by Top 5 and Bottom 5 US Metropolitan Regions
The question this paper seeks to address is why such regional variation exists. It begins
with a conceptual background of the motivations and barriers of flexible workplace practices.
The conceptual background is followed by a description of the dataset and formulations of the
models to be tested. The paper concludes with an explanation of the results and a conversation
on their significance.
Conceptual Background
In the digital age, work place flexibility, namely, working at home or remote from the
workplace, is far more feasible than it has ever been in the past. True, with the advent of the
telephone, workers could communicate easier from various locations and work from home has
always been possible to some degree. However, the digital age has decreased the necessity of a
workplace that provides shared resources such as computers, printers, fax machines, networks
and so forth. The digital age has brought sophisticated technology at much lower costs, so that
personal laptops, handheld devices and smart phones are affordable for many. 3 and 4 G
30.9%
25.4%
24.9%
24.5%
23.0%
10.9% 10.8%
9.9% 9.9%
6.3%
43
networks have rendered the ability to communicate through these devices from virtually any
location. The combination of these effects has already caused many of us to engage in
fragmented activity- activity which is broken up in space and time. Many shop while engaging in
work, or work while shopping in the aisles of grocery stores. ICTs have expanded the traditional
8 hour workday by allowing those with smart phones and laptops to check e-mails at home
before travelling to work, after work and on weekends. It would seem that as flexible work is
already taking place more individuals would be allowed to formally replace traditional time
spent at a place of work with remote work.
Although work at home is more feasible than ever, statistics continue to show relatively
low rates of work from home. The 2009 National Household Transportation Survey (2009)
reveals that only around 14% of the workforce is allowed the option to work at home. Only
around 9% of the workforce works at home one or more days a month.
What can explain the aggregate preference to continue to travel to work and the limited
uptake of flexible work options? Several tensions and barriers exist. Face-to-face communication
and contact provides efficiency that may never be replaced by technology. Technology itself is
still limited and some occupations require complete, if not partial physical presence.
Face-to-face Communication
Some argue that the efficiencies of face-to-face communication exceed and will continue
to exceed efficiencies in digital communication (Learner & Storper, 2001; Olson & Olson,
2000).
Computer-mediated communication has several indisputable benefits. It can equalize
hierarchies among conversants by obscuring status (McQuail, 2000) and increase opportunities
for participation, as the more vocal during face-to-face encounters have diminished physical
44
capacity to exert dominance (Martins, Gibson, & Maynard, 2004). Computer-mediated
communication can allow organizations to harness talent from virtually anywhere through
alleviating the costs of travel and time, expanding their networks to a global reach (Martins,
Gibson, & Maynard, 2004; Piccoli & Ives, 2003). Computer-mediated communication allows for
instant interaction from virtually anywhere.
However, face-to-face communication still dominates the playing field when it comes to
work. From an urban theory perspective, this occurs for several reasons. Face-to-face
communication enhances productivity and innovation, as it eases the transfer of information
spillovers, the backbone to innovation and learning (Rhoads, 2010). It also facilitates the
networking and collaboration necessary to support industries and workers.
Face-to-face communication inspires the above processes, because it provides cues not
possible through computer-mediated communication. Face-to-face communication allows us to
assess the non-verbal portion of communication, which may hold more value and substance than
the verbal portion. Critical to assessing the non-verbal portion is eye contact, which moderates
our social cognitive processes (Senju & Johnson, 2009). Eye contact and the ability to read body
language allow us to interpret tacit knowledge, which conveys meaningful information.
In fact, the nonverbal portion of communication such as: body language, eye contact,
inflections and pitch in voice, conveys attitudes and feelings 90% more than the actual words
being spoken (Mehrabian, 1981). Nonverbal communication projects nonverbal codes, which
many times occur spontaneously and emit universal meaning, while eliciting automatic responses
(Littlejohn & Foss, 2005). Digital signals conveyed through computer-mediation communication
obscure the above-mentioned cues and conveyances. Moreover, face-to-face communication
45
builds trust and engenders a better resolution of conflict; two aspects which are critical to
working in teams (Rhoads, 2010).
The benefits of face-to-face communication are such that most of us still prefer to work
alongside our team members, or travel long distances to meet with clients and collaborators.
Computer-mediated communication may act more as a compliment to face-to-face
communication, serving as a mode which fills in the gaps when needed, such as teleconferencing
meetings when collaborators cannot all travel to the same place.
Technology
In an increasingly virtualized workplace, technology becomes ever more important in our
ability to conduct work at home.
Handy and Mokhtarian (1996) cite the telephone as the key technology in conducting
work at home. The authors also mention that answering telephones and fax machines can
improve the home to office link. They envision that having a computer will eventually determine
the ability to work at home and that costs will be a deciding factor.
Since the time of this paper, technology has evolved dramatically. Both advances in
technology and decreasing costs have allowed computers and networks to penetrate most
American households. Estimates from the International Telecommunications Union (ITU) World
Telecommunication Indicators Database show that the proportion of US households with a
computer is around 76% in 2010, and 71% of households have internet access. Internet
subscriptions totaled 93,194,874 in 2010 and cell phone subscriptions 290,304,000. Although
many of these subscriptions may be attributed to the same household, given that in 2010, US
households totaled an estimated 116.7 million, it is safe to say that many, if not most US
households, have access to a computer, a network and cell-phone.
46
Technology trends in the workplace are shifting to accommodate more mobility and
flexibility. Whereas technology previously catered towards the workplace, eventually
transitioning into the consumer market, the trend is reversing and the consumer market is
penetrating the workplace (Newman, 2012). Businesses are creating platforms to accommodate
consumer market technologies, such as smart phones and I-pads, and IT companies are
incorporating BYOD, ‘Bring your own device’ movement (Newman, 2012). Around 61% of
surveyed companies have their employees bring their own personal laptops to work, 54% of
employees use smart phones for basic work tasks and 33% use tablets (Slattery, 2013). Tablets
are being used for more sophisticated tasks such as: data and project management, content
creation and customer relationship management (Slattery, 2013).
However, simply having available technology at home does not erase technological
tensions that may result from working at home. Some work may require substantial internet
connections and other types of access that may only be feasible from a shared workspace.
Central work offices may also provide technical support that is also not available from home,
which may be necessary for older workers who need more assistance and on-hand support. Other
resources such as: software, computing power and hardware may only be available at a work
place. Heavy technology such as that found in labs and other places of work require physical
presence. Finally, an organization may be better able to maintain security issues from a central
location.
The general arguments of the trust and efficient communication engendered by face-to-
face communication and the benefits of shared resources within a workplace, would explain low
rates of ability to work at home equally across regions if: 1) these factors do not vary across
industry sectors and if they do, 2) industry mix between regions is invariant.
47
The data reveal different patterns: 31% of workers living in San Jose have the option to
work at home while only 6% of residents in Las Vegas have the same option. One of the reasons
for not having the option may have to do with the particularities of these places: demographic
composition, rents and congestion levels, for instance. Another set of reasons may have to do
with the unique attributes of the industries composing these regions such as, the industry’s
propensity to take advantage of ICTs and a stronger ability of workers within the industries to
demand flexible workplace practices. In the case of San Jose and Las Vegas, San Jose is known
for its high-tech industries, requiring a lower physical presence from its employees, while Las
Vegas is known for its gaming industry, requiring a high level of physical presence.
Place
Demographics
Aggregate demographics may explain rates of telework. Regions that, on aggregate
house, older, more educated individuals may see more rates of telework as older individuals
have, more control over their work and status within the workplace and they are more likely to
be managers. However, younger individuals may be more comfortable with technology and
inclined toward flexible lifestyles, thereby, demanding more telework than older populations and
regions who house on average more youth, may see more rates of telework.
Results of telework studies are mixed with respects to age. Some conclude that the older
an individual, the more likely they are to telework (Walls, Safirova, & Jiang, 2007; Sener &
Bhat, 2010; Ellen & Hempstead, 2002). Others determine that younger individuals may be more
likely to telework (Drucker & Khattak, 2000).
Most telework studies find a positive correlation between education and the propensity to
telework (Yeraguntla & Bhat, 2005; Tang, Mokhtarian, & Handy, 2008; Walls, Safirova, &
48
Jiang, 2007; Sener & Bhat, 2010; Ellen & Hempstead, 2002; Drucker & Khattak, 2000). The
more highly educated individuals may possess greater skills and therefore, the ability to demand
flexible workplace options.
H1: The proportion of individuals who have the option to work at home is positively
correlated with regions composed of higher educated workforces.
H2: Regions with higher shares of older/younger populations will see higher proportions
of flexible workplace options.
Density and Accessibility
It is well known that aspects of travel to work vary spatially due to the geographic
distribution of activities and the supply of transportation systems (Giuliano, 2003). Although
telework is the absence of travel to work and will be influenced, in part, by a different set of
attributes than traditional travel to work, the two are likely to share common influences.
Researchers may not be in agreement on the significance and magnitude of the causes
affecting travel in general; however, consensus exists that the built environment influences travel
in some way (Frank, 2000), particularly travel to work. Aspects such as accessibility, density and
land-use mix influence what type of mode one chooses when they travel to work (Pinjari,
Pendyala, Bhat, & Waddell, 2007; Schwanen & Mokhtarian, 2005).
What is it about spatial variation in activities and transportation networks that may
influence telework? Firstly, one would expect that areas with greater transportation accessibility
(number of transit options and frequency) would see more residents who utilize these services as
a mode to work. Secondly, the price one has to pay to park their car at work, the cost of fuel, and
the cost of using the transportation system, would affect the decision to own a car and or drive it
49
to work. Also, the number of available destinations both at home and at work is likely to
influence the decision to commute by car.
In general, accessibility is found to be associated with higher use of modes, other than a
car (Gao, Mokhtarian, & Johnston, 2008; Kitamura, Mokhtarian, & Laidet, 1997) and shorter
commuting times (Susilo & Maat, 2007). These findings are enhanced by conclusions that
densities (a proxy for accessibility and the transit network), affect travel in some form
(Schwanen, Dielema, & Dijst, 2004; Handy, 2005) notably the work commute (Asensio, 2002).
Travel time and cost are also significant predictors of travel choice to work (Frank, Bradley,
Kavage, Chapman, & Lawton, 2008; Asensio, 2002).
Telework is likely to be dependent on some of the same factors. The more accessible a
region, the shorter the average distance to activities. Denser places are likely to have more transit
services. Therefore, dense and accessible places should provide more job options within shorter
distances and the demand for telework would decrease, all else equal.
The higher the time and cost to travel and reside within a region, the more telework
should be seen. Higher parking rates, higher rents and higher commute burdens, would increase
the likelihood to telework, as it would alleviate, or eliminate the associated time and cost
burdens. The telework literature validates some of these statements. Higher parking rates and
greater transit accessibility increase the likelihood to telework (Drucker & Khattak, 2000). The
likelihood to telework also rises, as perceived frustrations with commuting increases
(Mokhtarian, et al., 1996; Mokhtarian & Salomon, 1996; Vana, Bhat, & Mokhtarian, 2008) and
as actual commute distance increases (Tang, Mokhtarian, & Handy, 2008; Walls, Safirova, &
Jiang, 2007; Sener & Bhat, 2010).
50
H3: Commute costs and Density will be positively correlated with proportions of flexible
workplace options.
Industrial Composition
Business operations are being redefined constantly in response to technological
innovations and society’s adaptations to the digital world. No doubt, organizations cannot
survive without some degree of virtualization. 80% of on-line transactions, in 2001, were
business-to-business. Virtual information grows at a rate of 7.3 million web pages per day
(Lyman and Varain, 2000). E-business, defined as business where the components of
management, finance, innovation, production, distribution, and sales are conducted virtually, is
dominating the economic playing field engendering a need for organizations to conform to
internet-based technology (Castells, 2003).
The labor force required to sustain e-business must be able to work with information
retrieval, processing, and application in virtual settings. Labor for e-business needs to be self-
programmable (individuals need to consistently reprogram their skills and knowledge) and
flexible. This type of labor, however, is in short supply. “From Silicon Valley to Stockholm and
from England to Finland, the most important problem for leading companies has become where
to find engineers, computer programmers, e-business professionals, financial analysts, or, for that
matter, anyone with the capacity to develop new skills, as required by changing markets”
(Castells, 2003: 93).
Firms that innovate successfully and rapidly are likely to be high-growth firms. High-
growth industries are generally characterized by industries more reliant upon technological
progress (Chatman & Jehn, 1994) and high-tech industries are characterized by the use of
sophisticated technologies and complex operations (Khandwalla, 1976). When firms innovate in
51
technology, they are poised to reap large global market profits. In order to maintain competitive
and innovate, these organizations must secure the best talent, which is in short supply. Those
who adapt their organizational models to modern-day business climates are those who succeed.
Flexible workplace practices can mitigate workplace shortages, to some extent, through
expanding a firm’s labor network, by allowing them to hire individuals from virtually any
location. It can also be provided as a form of a benefit (the same as pay incentives, for example)
to recruit and retain competitive talent. Offstein, Morick & Koskinen (2010) observe that the
benefit in employing telework in an organization does not come from cost savings, but from
allowing companies to create different models of work to secure the best talent, which is a
product of leadership, more than technology.
Given that high-tech firms are competing for talent, which is critical to their ability to
innovate, they may be using more flexible workplace strategies, than the average industry, to
secure such talent, whether it is in the form of hiring talent, regardless of location, or providing
flexible workplace practices as an incentive to attract talent.
Table 5 lists the correlations between regional shares of various occupations and the
region’s share of workers who can work at home. These categories range from Arts, Design,
Entertainment, Sports, and Media Occupations to Farming, Fishing, and Forestry Occupations.
52
Table 5:
Work at Home and Occupation Correlation
Occupations Pearson Correlation
Architecture and Engineering 0.579
Arts, Design, Entertainment, Sports, Media 0.336
Building and Grounds Cleaning and
Maintenance -0.279
Business and Financial Operations 0.563
Community and Social Service 0.038
Computer and Mathematical Science 0.743
Construction and Extraction -0.161
Education, Training and Library 0.066
Farming, Fishing and Forestry -0.064
Food Preparation and Serving Related -0.007
Healthcare Practitioner and Technical -0.308
Healthcare Support -0.281
Installation, Maintenance and Repair -0.367
Legal Occupations 0.390
Life, Physical, and Social Science 0.610
Management Occupations 0.654
Office and Administrative Support -0.374
Personal Care and Service -0.354
Production -0.278
Protective Service -0.160
Sales and Related -0.183
Transportation and Material Moving -0.441
It is evident through the table that some industries allow more flexible workplace options
than others; for example, the share of computer and mathematical science has the highest
correlation with work at home. Other high ranking categories such as life, physical and social
science occupations and architecture and engineering also have high proportions of knowledge
work.
H4: High-tech occupation mix will be positively correlated with proportions of flexible
workplace options.
The share of a region’s management occupations is also highly correlated with the share
of the region’s workforce allowed flexible workplace options. This could be explained several
53
ways. It could be that managers are better able to express and demand flexibility in the
workplace. Some telework studies find corroboratory evidence to such a notion (Vana, Bhat, &
Mokhtarian, 2008; Mokhtarian & Salomon, 1996; Ellen & Hempstead, 2002).
Entrepreneurial theory may also help to explain the correlation. Wiggins (1995) proposes
that some innovations are engendered by small, owner-managed firms, as innovators have more
control over products, than if they were to remain at their former place of employment. In these
cases, an employee makes a valuable discovery, while at a firm and for a variety of reasons
‘spins-off’ of the firm creating her own firm, where control over the creative process is more
feasible. Owner-managed firms are likely to incorporate more flexible management and hiring
processes, for the reasons of competition mentioned above.
H5: The share of workers in management occupations is positively correlated with
proportion of flexible workplace options.
In summary, a model based on regions allows me to test for place; a model with
sufficiently fine data on occupation mix allows me to test industry sector arguments, as
occupational intensity is a proxy for industries more present in regions. My model tests the place
variables mentioned above (demographics and land use variables) with the industry/occupation
variables on a region’s proportion of flexible workplace provision.
The analysis below explores the systematic differences between regional provision and
uptake of flexible workplace options and underlying characteristics of a region.
Data and Analysis Strategy
Data
This study uses a national sample from the 2009 National Household Transportation
Survey. The 2009 NHTS was conducted by the US Department of Transportation from March
54
2008 to May 2009. The survey was completed for 150,147 households and collected information
on travel patterns, work-related information, and household demographics. Datasets are available
at the household level, person level, vehicle level, and daily trip level.
The following areas (cities and states) requested increased sample sizes: Phoenix AZ,
Tucson, AZ, California, Florida, Georgia, Iowa, Cedar Rapids IA, Indiana, North Carolina,
Piedmont Region NC, Omaha NE, New York, South Carolina, South Dakota, Tennessee, Texas,
Virginia, Vermont Chittenden County VT, and Wisconsin.
The sample is intended to be nationally representative through the random sampling of
list-assisted dialing banks, so that each household with a landline telephone had equal probability
of being selected. The sampling frame includes 100-banks with sets of 100 land-line telephone
numbers with the same first eight digits. Telephone numbers were then sorted by the ten Census
Divisions and then by metropolitan/non-metropolitan areas. Metropolitan areas were then sorted
by population of the Metropolitan statistical area (MSA) and primary metropolitan statistical area
(PMSA). Within these, telephone numbers were sorted by those in the central city, then those in
the non-central cities.
The geographic unit of analysis for this study is the core based statistical area (CBSA). A
CBSA is defined around an urban center of at least 10,000 people where the adjacent areas are
connected to the core by commuting. The NHTS groups its respondents into 51 CBSAs. Those
who are not located in a CBSA are placed together in one group.
Several questions regarding flexible work conditions were asked in the 2009 survey. The
following question serves as the basis for whether the respondent has the opportunity to
telework: “Do you have the option of working at home, instead of going into your primary
workplace?” A yes response leads to a question on frequency of work at home: “How many
55
times in the last month did you work only at home for an entire work day, instead of traveling to
your usual (primary) workplace? The following question serves as the basis for whether the
respondent has flexibility in their work schedule: “Do you have the ability to set or change your
own start work time?
Individuals who are both 18 and over and employed (self-employed and employed by
others, as well as part-time and full-time individuals) are used and separated into two groups for
flexible workplace: those who have the option to work from home and those who indicate that
they do not have the option to work from home. These same groups are used for flexible work
time: those who have the option to set their work start time and those who indicate that they do
not have the option. These individuals will then be examined on a regional level through
proportion of the workforce within a region who is allowed to telework and who has flexible
start time options.
One of the issues to consider when using the NHTS to analyze regional behavior is
whether the derived sample statistics from the survey are a true representative sample of the
region. Sampling frames of land-line telephones tend to be biased toward upper income home
owners who have resided in a home for a long-time; while address-based samples captures lower
income, minority, renters, new residents, and cell-only households (Sener & Bhat, 2010).
Therefore, landline surveys will over-represent high income, non-minority and older households
and address based surveys will do the opposite.
The NHTS accounts for this through their weighting mechanisms that adjust for
misrepresented demographic factors such as household size, race and ethnicity. However,
weights should solely be used for demographic purposes, not travel behavior, as inflating travel
behaviors based on weights produces unlikely outcomes. When they are extended to reflect
56
behavior, assumptions become distrustful. For example, if a weight of three is given to a certain
household or individual, this entity’s travel behavior is multiplied by three and three more
individuals or households automatically display the same behavior in the weighted sample.
Although, not using weights for an analysis not at a national scale results in an unrepresentative
sample, weights were not used in this study.
Although the potential unrepresentativeness of the sample is likely to affect aggregate
regional statistics, the underrepresentation of low- income households and younger populations
should be consistent across regions. The comparison of flexible workplace proportions across
regions should still reveal patterns as to underlying regional attributes that help to explain
regional differences, as all regional sampling was random, based on the MSAs phone bank and
therefore the same types of populations were accessed across regions.
Table 9 in the Appendix for paper 1 (pg. 85) validates such an assumption. The table
compares the proportion of individuals in each of the chosen categories from the 2010 Census
and the 2009 NHTS. Although, to slightly varying degrees, the NHTS under represents for each
MSA, individuals aged 18 to 24 and 25 to 34 compared to the Census. It over represents those
aged 45 to 54, 55 to 64, 65 to 74 and 75 to 84. The NHTS over represents for each MSA ‘whites’
and under represents ‘blacks’. It over represents the population with a graduate degree or higher.
Finally, the NHTS, for each MSA, under represents households earning under 10K and over
represents households earning over 100K, compared to the Census.
The number of CBSAs is 51 and this is the resulting sample size for the regression
analysis. Although aggregating the data to the level of the region results in a smaller sample size
than the NHTS makes available, the level of aggregation is needed in order to explore industry
57
effects. The NHTS separates individuals by 5 arbitrary occupational categories rendering
occupational analyses impossible.
Analysis Strategy
Ordinary least squares regressions are conducted on the dependent variables (flexible
workplace provision and use):
1. Regional percent of individuals allowed the option to work at home
2. Regional percent of individuals who have the ability to adjust their working start-
times
3. Regional percent of individuals who work at home at least once a month and
4. The regional weighted average of days worked at home.
Dependent variable relationships are tested against:
Geographical context: population density, log of population, median gross rent. These
variables come from 2009 American Community Survey (ACS) data.
Log of Population estimates the size of the region. It is measured as the 2010 census
population of each metropolitan region.
Population Density is measured as the 2010 total population of the region divided by the
square miles of the metropolitan region. It should be negatively related to the provision of
telework as higher density indicates higher overall accessibility of a region.
Median Gross Rent for a region are data provided by the 2009 census. Higher rents may
be associated with more options to telework as people move further from the central city, in
order to seek lower rents and use telework as a replacement to the commute.
Transportation factors: public transit network, central business parking rates, commuter
stress index. These variables come from 2009 Texas Transportation Institute (TTI) data and the
2011 Colliers International Parking Rate Survey data.
Public Transit per Capita is the Texas Transportation Institute’s (TTI) measure of
public transit by region divided by the population of the region in order to compare across
58
regions. The institute calculates public transportation based on regular route services from all
public transportation providers in an urban area (Schrank, Lomax, & Eisele, 2011). It is expected
that this measure be negatively correlated with rates of telework provision in an area. The more
option one has to get to work the less likely they are to use telework as a commuting option.
Commuter Stress Index is a 2009 TTI measure of peak hour congestion. It utilizes the
same formula for calculating the Travel-Time Index (peak period travel time compared to free
flow travel time), yet, only includes travel in the peak direction during peak periods (Schrank,
Lomax, & Eisele, 2011). With regards to the travel-time index, a value of 1.3 indicates that a 30-
minute trip would take 39 minutes to complete. It is expected that the higher the index, the more
telework provision in a region.
Mean Weekday Central Business District Parking Rates are pulled from the 2011
Colliers International Parking Rate Survey and represent mean weekday parking rates of the
central business districts (CBD) of metropolitan regions. Higher parking rates should incentivize
more telework provision within a region. Parking rate data are only available for all US
metropolitan regions at the level of central-business district. Although central-business districts
share only a portion of the employment in a region, CBD parking rates serve as a proxy to
average regional parking rates.
Occupation: high-tech shift share, managerial - shift share, business - shift share
Occupation
Borrowing from Barbour & Markusen’s (2007) classifications, I will define high-tech
workers via the Standard Occupational Classification Codes (SOC) as:
Computer and information technology professionals: computer engineers, systems
analysts, database administrators, computer support specialists, computer programmers,
59
computer programmer aides, programmers of numerical, tool, and process control, and all other
computer scientists
Selected engineers: chemical, civil, electrical, electronics, and mechanical
Selected natural scientists: agriculture and food scientists, biological scientists,
conservation scientists, foresters, medical scientists, all other life scientists, geologists,
geophysicists, oceanographers, physicists, astronomers, chemists, atmospheric scientists, and all
other physical scientists
This work will also consider the Business and Financial Occupation and Management
occupational categories.
The occupational measurements will compare the proportion of regional workers in a
sector relative to the proportion of national workers in a sector:
( 𝑂𝑖 /𝐸𝑖 ) /( 𝑂𝑗 /𝐸𝑗 )
Where O represents total employment in a sector and E represents total employment
(i=region; j=nation).
In the descriptives (Table 6) High Tech shift share has the largest range and standard
deviation of the three shift share measures.
Demographics: share of individuals aged 20to 30; share of individuals aged 30 to 40;
share of individuals aged over 40. These variables are calculated with 2010 Census Data.
Education
To measure the level of skill in a region I include Percent of Population with a Bachelor’s
Degree, which is calculated from census 2009 data on number of individuals within a region
holding a BA, divided by the total number of adults (population over the age of 25). The more
educated a region, the more likely telework provision will be seen in that region.
60
Age
Share of Population aged 20 to 30; 30 to 40; and 40+ are gathered from 2009 census data
as the total of the population of certain age groups, divided by the total population in that region.
If a region is composed of greater shares of younger individuals, they may see more telework, as
this population represents a workforce more suited to telework.
Empirical Results
Descriptive Analysis
For the national sample, around 14.2% of the employed population has the option to work
at home (17,165 individuals). Around 8.7% of the population (or 61.6% of the population who is
allowed the option), actually worked one or more days at home for the prior month interviewed
(10,566 individuals).
For the 51 regions the average proportion of the population who has the option to
telework is around 17% and the average proportion of the population who can adjust the time of
their working hours is around 47%. The region with the highest share of workers allowed the
option to work at home is San Jose (31%) and the region with the lowest share is Las Vegas
(6.3%). This pattern also holds for workers allowed the ability to adjust their work start times.
San Jose ranks the highest in percentage (62%) and Las Vegas the lowest (32%).
Regional rankings differ with respect to who actually takes advantage of working at
home. In terms of the percentage of individuals allowed the option, who actually worked at home
one or more times a month, the region with the highest percentage is New Orleans (85%) and the
lowest is Louisville-Jefferson (46%). Cincinnati has- the highest weighted average of days
worked at home (6.73 days) and Cleveland the lowest (1.69 days).
In terms of occupation, the average region has a higher share of high tech workers than
the nation (11% higher). The nation’s share of high-tech workers relative to total workers is
61
3.7% (4,141,060 high-tech workers/113,041,390 total workers). Figure 2 shows the regions
ranked by their high tech shift share scores. As expected, San Jose has the largest score (3.5 or
250% greater than the nation) followed by Seattle and San Francisco (2.25 or 125% greater) and
Raleigh-Cary (1.75 or 75% greater). San Jose’s share of high-tech workers relative to total
workers is 13.0% (115,970/888,960). The region with the lowest share is Las Vegas (.5 or 50%
less) along with New Orleans, Miami and Buffalo. Las Vegas’ share of high-tech workers
relative to total workers is 1.5% (12,680/858,540).
Table 6.
Regional Variables Summary Statistics for Regressions (n=51)
Variable Mean Std. dev. Min Max
Percent of workforce who can work at home 16.82% 4.47% 6.25% 30.91%
Percent of workforce who can work flexible hours 46.85% 5.91% 32.03% 61.92%
Occupation
High Tech Shift Share 1.11 0.55 0.4032 3.56
Management Shift Share 1.03 0.25 0.6329 1.69
Business Shift Share 1.08 0.24 0.58 1.88
Accessibility
Public Transit per Capita 0.178 0.19 0.016 1.11
Commuter Stress Index 1.25 0.09 1.08 1.54
Mean Weekday CBD Parking 14.72 8.03 3.00 41.00
Population Density 690.41 576.63 117.65 2825.99
Total Population
3,166,845
3,236,163
115,788
18,897,109
Log of Population 6.35 0.36 5.1 7.3
Median Gross Rent 901 178 643 1414
Demographics
Percent of population with a Bachelor ’s degree 19.81% 3.41% 12.54% 28.89%
Share of population aged 20 to 30 14.00% 1.04% 11.84% 17.07%
Share of population aged 30 to 40 14.00% 1.38% 11.17% 17.72%
Share of population over the age of 40 44.67% 3.59% 36.00% 52.82%
62
The average region also has a higher share of management occupations than the nation
(3%) and a higher share of business occupations (8%). The region with the highest management
shift- share score is San Jose (69% higher than the national average) and the lowest is
Jacksonville (37% lower than the national average). The region with the highest business-shift
share is Washington (88% more than the nation) and Las Vegas has the lowest (42% lower than
the nation).
Regions differ rather widely in terms of transportation measures. The average score for
public transit availability per capita is quite low, at 0.18. The regions see much variation in
public transit per capita. New York ranks as the highest at 1.11, followed by San Francisco at
.61. The region with the lowest transit per capita score is Oklahoma at .02. Mean weekday
central business parking rates are $14.72. The region with the most expensive mean central
business parking rates is New York at $41, followed by Boston at $31.00. Raleigh, NC sees the
lowest mean rates at $3.00. For a 20-minute commute during free flow speeds, the average
region’s peak hour commute takes 25 minutes, or an additional 5 minutes. There is slightly less
variation amongst the regions with respect to the commuter stress index. The region with the
highest index sees an additional 10.8 minutes to a 20-minute free flow commute (Los Angeles)
and the lowest 1.6 minutes (Richmond, VA).
The average median household gross rent is $901 a month for a region (based on 2009
census data). The highest costly rent is $1,414 per month (San Jose) and the least costly rent is
$643 (Pittsburgh). An average of around 20% of region’s populations have a Bachelor’s degree
or greater. The most educated region’s proportion is around 29% (Raleigh, NC, followed by San
Francisco and then Austin) and the least educated proportion is around 13% (Riverside-San
Bernardino).
63
There is little variation in proportions of ages within regions. On average, 14% of the
regions have populations between the ages of 20 to 30 and 30 to 40. The highest percentage is
around 17% (Austin sees the highest percentage of individuals aged 20 to 40), while the lowest is
around 12% (Cleveland and Rochester see the lowest percentages of individuals aged 20 to 40).
64
Figure 3:
High Tech and Shift Share Rankings by Region
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
Shift Share High Tech Management Shift Share
65
Regression Results
Supply Regressions: Option to Work at Home and Option to Adjust Work Start
Times
Table 7 shows the results of least squares regressions on 2009 rates of work at home and flexible
start time provision by metropolitan regions in the US in standardized beta coefficients in order
to interpret the individual coefficients and to be able to compare the coefficient responses
equivalently. The estimated coefficients are generally consistent with my expectations, with
some exceptions. Demographics other than education played no significant role in telework
variation. Overall, the explanatory and control variables together explained around seventy
percent of the regional differences in telework provision.
In terms of working at home results, high-tech shift share and management -shift share
are both strongly positive and statistically significant in all the models as predicted. The
coefficient on business- shift share is not significant and rather small, indicating that the
occupational phenomenon really has to do with the share of a region’s high-tech and
management occupations, not general share of knowledge workers, as many employed in the
business fields deal in knowledge, not physical material. The standardized coefficient on high
tech is on average 30% larger than the coefficient on management. High tech is the most
powerful coefficient in the models.
The age measures are small and statistically insignificant. However, Share of Population
with a Bachelor’s Degree and Share of the Population Aged 30 to 40 are highly correlated. The
share of the population aged over 40 is also negatively correlated with the younger shares,
indicating that regions that have higher shares of youth have lower shares of older generations.
The coefficient on Median-Gross Rent is small and insignificant; indicating that share of
telework is not explained by higher rents in a region. In other words, telework does not seem to
66
be a travel substitute for people, or employers seeking lower rents. This is also substantiated by
the fact that the coefficient on commuter-stress index is also insignificant, highlighting that the
magnitude of the commute does not influence rates of telework. Moreover, although
insignificant, the commuter-stress index is negative, which is counterintuitive and reads that
regions with worse commutes see lower rates of telework. The higher the commute burden, the
more transit a region has and Public Transit per Capita is significant and negative; indicating that
the more alternative options one has to commute to work, the lower the rates of telework within a
region. The more likely explanation is that New York is biasing the transit results. New York
has the highest transit per capita at 1.11, which is .50 points higher than the second highest
region, San Francisco, at .61, followed by Washington D.C., at .54. There is only a .07 difference
between San Francisco and D.C. New York, however, only ranks 26 out of the 51 regions in
telework provision.
Log of Population and Population Density, although insignificant, are positive
contributing to the argument that regions with greater amenities see more telework.
Results for percent of individuals who have the ability to set or change their own work
times mainly mimic the work at home regressions with a few exceptions. High tech shift share is
strongly correlated with the proportion of a region that is allowed the option to adjust their
working hours. The coefficients on high tech shift share are slightly higher in the work start time
flexibility model than in the work at home availability model. Public transit per capita is again
significant and negative indicating that organizations within regions holding more travel options
offer less workplace flexibility. The results here substantiate the results in the work at home
models indicating that regions seeing higher proportions of individuals who have the option to
work at home have higher proportions of high-tech workers and limited transit options.
67
Table 7.
Regression Results Dependent Variable: Percent of Individuals who WAH and FLEX
Standardized Coefficients and p-values (n=51)
WAH FLEX
Adjusted R Squared 0.628 0.558
Shift Share High Tech 2.612 4.151
0.044* 0.028*
Shift Share Management
5.156 5.294
.025* 0.105
Public Transit
-6.655 -10.106
0.045* 0.036*
CBD Parking Rates
0.075 0.056
0.319 0.608
Commuter Stress Index
-11.079 -10.547
0.177 0.37
Population Density
0.001 0.001
0.419 0.599
Log of Population
1.113 0.485
0.528 0.849
Bachelor’s Degree
0.398 0.415
0.127 0.268
Median Gross Rent
0.002 0.000
0.541 0.997
20 to 30
0.063 1.347
0.925 0.171
30 to 40
0.437 1.058
0.56 0.331
Business Shift Share
0.93 1.832
0.709 0.612
40+
0.11 0.522
0.724 0.248
68
Demand Regressions: Proportion of Individuals who take the Option and Weighted
Average of Number of Days Worked at Home
Table 8 shows the results of the regional work at home quantity regressions. The results
from these two regressions (percent of the workforce who works at home at least once a month
and regional average days worked at home), are different from the option regressions.
For the regional workforce who actually takes advantage of working from home, the
explanatory and control variables together explain around seventy percent of the regional
variation. However, only two variables are significant.
In this regression, the region’s high-tech-shift share ceases to become a significant
contributor, although its management shift share remains significant. This indicates that key to
the uptake of telework is one’s position within the workplace. So, although in the previous
regressions, high-tech composition explained variation in telework provision, it is still not
commonplace within these organizations to actually allow people to work at home. Rather,
employees with higher work status are those who are taking advantage.
Concomitantly, a region’s education level remains significant. Therefore, regions
composed of highly educated managers see more actual telework. High status employees only
take advantage if they are highly educated.
For the final regression, regional averages of days worked at home, the explanatory and
control variables together explain around forty percent of the regional variation, thus far, the
lowest of the regressions. Here several other variables become significant.
A region’s high-tech and management-shift share cease to become significant.
Conversely, the higher a region’s business-shift share, the lower the average days worked at
home.
69
Regional descriptors such as median gross rent and demographics hold more weight in
explaining average regional variation. The higher the regional median-gross rent, the higher rates
of telework a region sees. This may indicate that individuals taking more advantage of working
at home do so in areas that see higher rents and may be seeking less expensive housing on the
peripheries of regions. A region’s size is positively correlated with higher averages. This finding
plays well with median gross rents, as larger regions tend to see cheaper housing on its
peripheries. A region’s density is negatively correlated with higher averages.
The implication here is that whether or not one has the option to work at home is a
measure of the supply of telework, driven more by workplace attributes. The actual amount of
work at home may be more demand driven, reflecting conditions of the person.
70
Table 8:
Regression Results Dependent Variable= % Who WAH at Least Once a Month and Regional
Weighted Average of days WAH Standardized Coefficients and p-values (n=51)
R Squared 0.715 0.422
Shift Share High Tech
0.001 -0.601
0.757 0.160
Shift Share Management
0.02 -1.161
.011* 0.123
Public Transit
-0.009 1.768
0.462 0.139
Log of Population
0.003 1.078
0.601 0.094
Bachelor’s Degree
0.003 0.139
0.005* 0.112
Business Shift Share
-0.002 -1.606
0.82 0.061
Commuter Stress Index
0.002 1.032
0.948 0.712
Median Gross Rent
0.000 0.004
0.29 .009*
CBD Parking Rates
0.000 -0.038
0.715 0.137
20 to 30
0.000 0.125
0.939 0.589
30 to 40
0.000 0.146
0.944 0.559
40+
0.001 0.075
0.541 0.469
Population Density
0.000 -0.001
0.671 0.030*
Average Office Price per
sq foot
0.000 -0.002
0.599 0.276
71
Conclusion Paper 1
The analysis of regional variation in the opportunity of workplace flexibility using
aggregate regional data from the 2009 National Household Transportation Survey (NHTS)
suggests that industry norms, in the form of aggregate occupational intensity, matter. Direct
focus on an individual’s circumstances and attributes has not led us too far in understanding the
general trends of flexible workplace practices. Factors such as household constraints and
personal views toward working at home are most definitely important elements in considering
who works at home and why. However, without a national dataset that explores these issues in-
depth, results from smaller-scale studies will only correspond to the location under consideration
and cannot be generalized to larger structural trends.
Ultimately, whether one works at home or not, is fundamentally contingent on if she is
allowed to do so, whether it be the ability of her work to be conducted at home, or whether her
organization allows the practice. Larger-scale analyses examining regional tendencies can
illuminate larger industrial patterns or trends. If some regions see more flexible workplace
opportunities explained by their underlying industrial and occupational intensities, this may be
indicative of a larger cultural trend occurring in certain industries.
What we may intuitively think affects rates of telework provision such as: congestion,
high rents, and dense-metropolitan environments, generally does not seem to be associated with
high rates of work at home opportunities amongst regions. High rates of work at home are not
present in some of our largest metro areas such as Los Angeles and New York, but rather in new,
upcoming and smaller cities such as San Jose and Austin, which have become high tech
incubators. Work here merits further research. The reasons these cities are high tech incubators
may be explained through the nexus of nearby university high-tech programs, level of
entrepreneurial funding and resources and quality of life provided in these cities.
72
Work at home is an innovative workplace practice as the organization must have trust
that employees allowed to work at home will be as productive as in a central location. The ability
to work at home is also an innovative workplace benefit and can serve to attract and retain
employees. Therefore, whether or not an organization offers the option, reflects the culture of the
organization, which is in part influenced by industry norms (Pennings & Gresov, 1986).
Industries use technology differently and experience different rates of growth, both features,
which help to define an industry and shape organizational culture (Chatman & Jehn, 1994).
Industries facing rapid growth and entry are pressured to adjust their organizational models in
order to maintain competitive.
Results here demonstrate that rates of telework provision seem to be associated with
areas that have high levels of education and house more high-tech workers and managers than
the average region. The provision of flexible work start-time also follows the same pattern. This
finding corroborates the concept that telework is fundamentally an organizational tendency,
perhaps a buzz, or a trend of the ‘new order’, or a reflection of the industrial tendencies of certain
regions such as the increasing use of distributed work. The use of telework by organizations can
accommodate and placate its workforce, who on average may demand more flexibility in their
lifestyles. Individuals are drawn to these telework friendly regions, because of the job
opportunities they offer. The organizations reflect the culture the individuals are bringing.
Such a finding also sheds light on the innovative culture of organizational practices.
Given that regions with higher proportions of high-tech workers also see the most telework
provision, something about the nature of these organizations is demanding innovative work place
practices. Hi-tech organizations face the most pointed competition, as they are poised to reap
large and global market shares as long as they innovate. Innovation in high technology requires
73
that organizations hire the freshest and most skilled talent who are able to keep a pulse on the
latest technological trends, wherever they may be located. Therefore, high technology
organizations respond to the culture of their employees and attempt to create an environment that
fosters the work product of their employees and meets their lifestyle demands.
Results differ when the analysis turns to who is actually taking advantage of work at
home. The 2009 NHTS data reveal that only around 9% of the workforce takes advantage of
working at home, while 14% are allowed to take the option. Regions that see more rates of
individuals working at home at least once a month are explained by the region’s managerial-shift
share and education. These results indicate that one’s position of authority in a place of work
explains her propensity to work at home more than her occupation. Although high-tech
organizations may be offering more flexible workplace options individuals (who are not in
positions of management) may still feel the need to show up at the workplace to be noticed. In
terms of how often one works at home, factors such as rent and the size of a region play a role
indicating that people more affected by higher rent and longer commutes are likely to engage in
more flexible work. Future work on this front is needed.
Examining option is still important to understand. An individual surveyed may not have
exercised the option in the month prior to being surveyed. However, this same individual may
take the option at different times of the year for different purposes. Understanding who has the
option, may also shed more light on how likely an individual fragments her work episodes. She
may take the flexibility provided by working at home to stagger her work throughout the day;
bunching it in the mornings, evening or weekends, while also visiting the office during the day.
The investigation of flexible work place practices within the context of organizational
culture and work type merits future research. Firstly, larger datasets that tie workplace flexibility
74
to occupation are needed. This study sacrificed sample size in order to explore regional effects
(particularly industry) on flexible workplace practice. An analysis could have been performed at
the individual level using multilevel modeling to control for regions; however, regions are of key
interest in this study and the hypothesis is that regional context such as industry plays a
substantial role. Analyzing at the regional level is also important as the individual-based studies
on telecommuting and telework have rendered inconsistent and mixed results.
Better, representative data sets are needed that capture young populations more
accurately, along with tighter measures of occupation and organizational descriptives (types, size
and age). If, in fact, the regions mentioned here are indeed offering more flexibility, the
attributes of these regions need to be better understood, as well as the migration and immigration
patterns of their workers. Results on industry effects of flexible workplace options can inform
planners of regional and neighborhood travel patterns.
The more we can understand how to engender efficiency in the workplace, the better
placed our economies will be to compete globally. Understanding how future workforces may
look and the barriers to entry can help economies prepare for accommodating these workforces.
For example, in Japan, which is facing a shrinking workforce, the Prime Minister Shinzo Abe is
creating a movement to make the workforce more accommodating to women, who are reluctant
to join the workforce, due to the taxing hours expected from most jobs. In the US, more women
are attending college than men and are soon to be a larger part of the professional workforce.
Studies examining workforce trends and how the flexible workplace can accommodate such
trends are needed.
Closer qualitative studies of industry more disposed to incorporating the flexible
workplace should be conducted as successful models. It is important to understand why
75
individuals, who are offered the option to conduct flexible work, don’t do it. This paper showed
that only a small portion of the workers who are allowed the option to conduct flexible work,
actually do so once or more times a month. Is this because the nature of the work truly does not
allow it to be virtualized, or that presence in the workplace is critical to getting promotions and
so forth? Although, face-to-face communication is a more efficient method, increasing evidence
shows that working in solitude produces more productive and creative work. Understanding the
reluctance of organizations to offer flexible workplace practices and the reluctance of individuals
to remotely, even when they are offered the option to work flexibly, is important for
understanding the barriers to such processes. Concomitantly, the reasons why individuals who
are offered the option take it, needs to be further explored.
Finally, more in-depth analysis of shared workspaces, or informal workspaces, such as
coffee shops, can be conducted. Researchers can examine how the provision of shared
workspaces is growing and where and what are the types of individuals accessing such spaces.
This type of research can also be conducted at informal workspaces, such as coffee shops and
libraries. Research on this front is important, as informal-flexible work is likely to grow, as more
individuals demand more of it and as our work becomes more portable through evolving
technologies.
76
References Paper 1
Alexander, B., Ettema, D., & Dijst, M. (2010). Fragmentation of work activity as a multi-
dimensional construct and its association with ICT, employement and sociodemographic
characteristics. Journal of Transport Geography , 18, 55.
Asensio, J. (2002). Transport mode choice by commuters to Barcelona's CBD. Urban Studies ,
39, 1881-1895.
Bagley, M., & Mokhtarian, P. (1997). Analyzing the preference for non-exclusive forms of
telecommuting: Modeling and policy implications. Transportation , 24, 203-226.
Bailey, D., & Kurland, N. (2002). A review of telework research: Findings, new directions, and
lessons for the study of modern work. Journal of Organizational Behavior , 23, 383-400.
Balkin, D., & Gomez-Mejia, L. (1987). Toward a contingency theory of compensation strategy.
Strategic Management Journal , 8, 169-182.
Bernardino, A., Ben-Akiva, M., & Salomon, I. (1993). A stated preference approach to modeling
the adoption of telecommuting. Transportation Reseach Record: Journal of the Transportation
Research Board, No. 1413 , 22-30.
Bonsall, P. (2004). Traveler behavior: Decision making in and unpredictable world. Journal of
Intelligent Transport Systems , 8 (1), 45-60.
Bonsall, P., & Shires, J. (2006). Employer expectations for commuting and business-related
travel in an environment rich in information and communication technologies. Transportation
Research Record: Journal of the Transportation Research Board, No. 1977 , 268-276.
Brannen, J. (2005). Time and negotiation of work-family boundaries. Time and Society , 14, 113-
131.
Breedveld, K. (1998). The double myth of flexibilization: trends in scattered work hours, and
differences in time-sovereignty. Time and Society , 7, 129-143.
Cameron, S. (1993). Why is the R squared adjusted reported? Journal of Quantitative Economics
, 183-186.
Castells, M. (2003). The Internet Galaxy: Refections on the Internet, Business and Society.
Oxford: Oxford University Press.
Chatman, J., & Jehn, K. (1994). Assessing the relationship between industry characteristics and
organizational culture: How different can you be? The Academy of Management Journal , 37,
522-553.
77
Couclelis, H. (2003). Housing and the new geography of accessibility in the information age.
Open House International , 28, 7-13.
Covin, J., Slevin, D., & Covin, T. (1990). Content and preformance of growht-seeking strategies:
A comparison of small firms in high and low-technology industries. Journal of Business
Venturing , 5, 391-412.
Crampton, S., & Hodge, J. (2009). Generation Y: Unchartered territory. Journal of Business and
Economic Research , 7.
Cullen, I., & Godson, V. (1975). Urban networks: The structure of activity patterns. Progress in
Planning , 4, 1-96.
Cullen, I., Godson, V., & Mayor, S. (1972). The structure of activity patterns. In A. Wilson,
Patterns and Processes in Urban and Regional Systems (p. 286). London: Pion.
DiMaggio, P., & Powell, W. (1983). The iron cage revisited: Institutional isomporhism and
collective reationality in organizational fields. American Sociological Review , 48, 147-160.
Drucker, J., & Khattak, A. (2000). Propensity to work from home: Modeling results from the
1995 Nationwide Personal Transportation Survey. Transportation Research Record: Journal of
the Transportation Research Board, No. 1706 , 108-117.
Ellen, I., & Hempstead, K. (2002). Telecommuting and the demand for urban living: a
preliminary look at white-collar workers. Urban Studies , 39, 749.
Farag, S., Dijst, M., & Lazendorf, M. (2003). Exploring the use of e-shopping and its impact on
personal travel behavior in the Netherlands. Transportation Research Record 1858 , Paper No.
03-3058, 47-58.
Faulconbridge, J., & Beaverstock, J. (2008). Geographies of international business travel in the
professional service economy. In D. Hislop, Mobility and technology in the workplace (pp. 87-
102). Abingdon: Routledge.
Florida, R., Gulden, T., & Mellander, C. (2088). The rise of the mega-region. Cambridge
Journal of Regions, Economy and Society , 1, 459-76.
Frank, L. (2000). Land use and transportation interaction: Implications on public health and
quality of life. Journal of Planning Education and Research , 20 (6).
Frank, L., Bradley, M., Kavage, S., Chapman, J., & Lawton, T. (2008). Urban form, travel time,
and cost relationships with tour complexity and mode choice. Transportation , 35, 37-54.
78
Gao, S., Mokhtarian, P., & Johnston, R. (2008). Exploring connections among job accessibility,
employment, income and auto ownership using structural equation modeling. Annals of Regional
Science , 42, 341-356.
Garhammer, M. (1995). Changes in working hours in Germany: The resulting impact of daily
life. Time and Society , 4, 167-203.
Giuliano, G. (2003). Travel, location and race/ethnicity. Transportation Research Part A: Policy
and Practice , 37, 351-372.
Gujarati, D. (2004). Basic Econometrics 4th Edition. The McGraw-Hill Companies.
Handy, S. (2005). Smart growth and the transportation land-use connection: What does the
research tell us? International Regional Science Review , 28, 146-167.
Handy, S., & Mokhtarian, P. (1995). Planning for telecommuting: Measurement and policy
issues. Journal of the American Planning Association , 61, 99-111.
Handy, S., & Mokhtarian, P. (1996). The future of telecommuting. Futures , 28, 227-240.
Handy, S., & Mokhtarian, P. (1996). The future of telecommuting. Futures , 28 (3), 227-240.
Hanson, S., & Pratt, G. (1995). Gender, work and space. London: Routledge.
Haynes, B. (2011). The impact of generational differences on the workplace. Journal of
Corporate Real Estate , 13 (2), 98.
Hipple, S. (2010). Self-Employment in the United States. Monthly Labor Review. Bureau of
Labor Statistics.
Hjorthol, R. (2002). The relation between daily travel and use of the home computer.
Transportation Research PArt A , 36, 437-452.
Hjorthol, R., & Gripsrud, M. (2009). Home as a communication hub: The domestic use of ICT.
Journal of Transport Geography , 17, 115-123.
Holtshouse, D. (2010). Knowledge work 2020: Thinking ahead about knowledge work. On the
Horizon , 18 (3), 193-203.
Hubers, C., Dijst, M., & Schwanen, T. (2008). ICTs coping strategies and gender differences in
the temporal and spatial fragmentation of paid labor. In: Paper presented at the Fourth
International Specialist Meeting on ICT, Everyday Life and Urban Change, Bristol, UK.
Janelle, D. (1986). Metropolitan expanision and the communications-transportation tradeoff. In
S. Hanson, The geography of urban transportation (pp. 357-385). New York: The Guilford
Press.
79
Janelle, D., & Hodge, D. (2000). Information, Place and Cyberspace Issues in Accessibility.
Berlin: Springer.
Jang, T. (2005). Cout data models for trip generation. Journal of transportation engineering ,
131, 444-450.
Karnowski, S., & White, B. (2002). The role of facility managers in the diffusion of
organizational telecommuting. Environment and Behavior , 34, 322-24.
Kenyon, S., & Lyons, G. (2007). Introducing multitasking to the study of travel and ICT:
Examining its extent and assessing its potential importance. Transportation Research A , 41,
161.
Khandwalla, P. (1976). Some top management styles, their context and performance.
Organization and Administrative Sciences , 7 (4), 21-51.
Kitamura, R., Mokhtarian, P., & Laidet, L. (1997). A micro-analysis of land use and travel in
five neighborhoods in the San Francisco Bay Area. Transportation , 24, 125-158.
Koskela, H. (1999). "Gendered exclusions':Women's fear of violence and changing relations to
space. Geografiska Annaler B, Human Geography , 81, 111.
Kwan, M.-P. (2007). Mobile communications, social networks, and urban travel: New topoligies
of spatial interaction. The Professional Geographer , 59 (4), 434-446.
Learner, E., & Storper, M. (2001). The economic geography of the internet age. Journal of
International Business Studies , 32, 641-65.
Lenz, B., & Nobis, C. (2007). The changing allocation of activities in space and time by the use
of ICT- "Fragmentation" as a new concpet and empirical results. Transportation Research Part A
, 41 (2), 190-204.
Leung, A. (2003). Different ties for different needs- recruitment practices of entrepreneurial
firms at different development phases. Human Resources Management , 42 (4), 303-320.
Line, T., Jain, J., & Lyons, G. (2011). The role of ICTs in everyday mobile lives. Journal of
Transport Geography , 19, 1490-1499.
Littlejohn, S., & Foss, K. (2005). Theories of human communciation 7th ed. Belmont, CA:
Wadsworth.
Lyons, G., & Urry, J. (2005). Travel time use in the Information Age. Transportation Research A
, 39, 257-276.
Madge, C., & O'Connor, H. (2006). Parenting gone wired: empowerment of new mothers on the
Internet? Social and Cultural Geography , 7 (2), 199-220.
80
Mahmassani, H., Yen, J., Herman, R., & Sullivan, M. (1993). Employee attitudes and stated
preferences towards telecommuting: an exploratory analysis. Transportation Research Record:
Journal of the Transportation Research Board, No. 1413 , 31-44.
Mannering, J., & Mokhtarian, P. (1995). Modeling the choice of telecommuting frequency in
California: An exploratory analysis. Technological Forecasting and Social Change , 49, 49-73.
Martins, L., Gibson, L., & Maynard, M. (2004). Vitual teams: What do we know and where do
we go from here? Journal of Management , 30, 805-35.
Mayo, M., Pastor, J., Gomez-Mejia, L., & Cruz, C. (2009). Why some firms adopt
telecommuting while others do not: A contingency perpective. Human Resource Management ,
48 (6), 917-939.
McQuail, D. (2000). Mass communication theory. London: Sage.
Mehrabian, A. (1981). Silent messages: Implicit communication of emotions and attitudes.
Belmont, CA: Wadsworth.
Miller, H., & Shaw, S.-L. (2001). Geographic Information Systems for Transportation:
Principles and Applications. New York: Oxford University Press.
Mokhtarian, P. (2004). Telecommuting, residential location, and commute-distance travelled:
evidence from State of California employees. Environment and Planning A , 36, 1877-1897.
Mokhtarian, P., & Salomon, I. (1996). Modeling the choice of telecommuting: 3. Identifying the
choice set and estimating binary choice models for technology-based alternatives. Environment
and Planning A , 28, 1877-1894.
Mokhtarian, P., & Salomon, I. (1994). Modeling the choice of telecommuting: Setting the
context. Environment and Planning A , 28, 1859-1876.
Mokhtarian, P., & Salomon, I. (1994). Modeling the choice of telecommuting: Setting the
context. Environmental Planning A , 26, 749-766.
Mokhtarian, P., & Salomon, I. (1997). Modeling the desire to telecommute: the importance of
attitudinal factors in behavorial models. Transportation Research A. , 30 (1), 35-50.
Mokhtarian, P., Handy, S., & Salomon, I. (1995). Methodological issues in the estimation of
travel, energy and air quality impacts of telecommuting. Transportation Research A , 29, 283-
302.
Mokhtarian, P., Salomon, I., Saxena, S., Sampath, S., Cheung, P., Le, K., et al. (1996). Adoption
of Telecommuting in Two California State Agencies. Research Report UCD-ITS-RR-96-5,
Institute of Transportation Studies, University of California, Davis, September.
81
Moon, N. (2007). Private sector telework and its implications for economic development:
Results of a case survey and policy assessment. Atlanta, Georgia: Enterprise Innovation Institute
and School of Public Policy's Science, Technlology, and Innovation Policy (STIP) Program
Georgia Institute of Technology.
Newman, D. (2012, May 22). 6 Technology trends transforming the workplace. Retrieved June
28, 2013, from millenialceo.com: http://millennialceo.com/leadership/6-technology-trends-
transforming-workplace/
Odih, P. (1999). Gendered time in the age of deconstruction. Time and Society , 8 (1), 9-38.
Offstein, E., Morwick, J., & Koskinen, L. (2010). Making telework work: Leading people and
leveraging technology for competitive advantage. Strategic HR Review , 9 (2), 32-37.
Olson, G., & Olson, J. (2000). Distance matters. Human Computer Interaction , 15, 130-79.
Olszewski, P., & Mokhtarain, P. (1994). Telecommuting frequency and impacts for State of
California employees. Technological forecasting and social change , 45, 275-286.
Pain, R. (1997). Social geogrphies of women's fear. Transactions of the Institute of British
Geographers , 22, 231.
Pennings, J., & Gresov, C. (1986). Technological networking and innovation implementation.
Organization Science , 3, 317-334.
Peters, P., & Heusinkveld, S. (2010). Institutional explanations for managers' attitudes towards
telehomeworking. Human Relations , 63, 107.
Peters, P., Tijdens, K., & Wetzels, C. (2004). Employees' opportunities, preferences, and
practices in telecommuting adoption. Information and Management , 41 (2), 469-482.
Piccoli, G., & Ives, B. (2003). Trust and the unintended effects of behavior control in virtual
teams. MIS Quarterly , 27, 365-95.
Pickup, L. (1988). Hard to get around: a study of women's travel mobility. In J. Little, L. Peake,
& P. Richardson, Gender and the Urban Environment (pp. 98-116). New York: New York
University Press.
Pinjari, A., Pendyala, R., Bhat, C., & Waddell, P. (2007). Modeling residential sorting effects to
understand the impact of the built environment on commute mode choice. Transportation , 34,
557-573.
Popuri, Y., & Bhat, C. (2003). On modeling choice and frequency of home-based
telecommuting. Transportation Research Record: Journal of the Transportation Research
Board, No. 1858 , 55-60.
82
Rheingold, H. (2003). Smart Mobs: The Next Social Revolution. New York: The Perseus Book
Group.
Rhoads, M. (2010). Face-to-face and computer-mediated communication: What does theory tell
us and what have we learned so far? Journal of Planning Literature , 25 (2), 111-122.
Sánchez, A., Pérez, M., de Luis Carnicer, P., & Jiménez, M. (2007). Teleworking and workplace
felxibility: A study of impact on firm performance. Personnl Review , 36 (1), 42-64.
Schrank, D., Lomax, T., & Eisele, B. (2011). 2011 Urban Mobility Report. College Station, TX:
Texas Transportation Institute.
Schwanen, T. (2007). Matter(s) of interest: Artefacts, spacing, timing. GeografiskaAnnaler,
Series B: Human Geography , 89 (1), 9-22.
Schwanen, T. (2006). On 'arriving on time', but what is 'on time'? Geoforum , 37 (6), 882-894.
Schwanen, T., & Dijst, M. (2003). Time windows in workers' activity patterns: Empirical
Evidence from the Netherlands. Tranpsortation , 30 (3), 261-283.
Schwanen, T., & Kwan, M.-P. (2008). The Internet, mobile phone and space-time constraints.
Geoforum , 1362-1377.
Schwanen, T., & Mokhtarian, P. (2005). What affects commute choice: Neighborhood physical
structure or preferences towards neighborhoods? Journal of Transportation Geography , 13, 83-
99.
Schwanen, T., Dielema, F., & Dijst, M. (2004). Th impact of metropolitan structure on commute
behavior in the Nethwrlands: a multilevel approach. Growth Change , 35, 304-333.
Schwanen, T., Kwan, M.-P., & Ren, F. (2008). How fixed is fixed? Gendered rigidity of space-
time constraints and geographies of everyday activities. Geoforum , 39, 2109-2121.
Schwanen, T., Pokwan, M., & Ren, F. (2008). How fixed is fixed? Gendered rigidity of space-
time constraints and geographies of every day activities. Geoforum , 39, 2109-2121.
Sener, I., & Bhat, C. (2010). A copula basedsample selection model of telecommuting choice
and frequency. Environment and Planning A , 43 (1), 126-145.
Senju, A., & Johnson, M. (2009). The eye contact effect: Mechanisms and development. Trends
in cognitive sciences , 13, 127-34.
Sia, C., Teo, H., Tan, B., & Wei, K. (2004). Effects of environmental uncertainty on
organziational intention to adopt distributed work arrangements. IEEE Transactions on
Engineering Management , 51, 256-67.
83
Slattery, M. (2013, March 7). How consumer technology is remaking the workplace. Retrieved
June 28, 2013, from wired.com: http://www.wired.com/insights/2013/03/how-consumer-
technology-is-remaking-the-workplace/
Susilo, Y., & Maat, K. (2007). The influence of the built environment to the trends in commuting
journeys in the Netherlands. Transportation , 34, 589-609.
Swanberg, J., Pitt-Catsouphes, M., & Drescher-Burke, K. (2005). A question of justice:
disparities in employees' access to flexible schedule arrangements. Journal of Family Issues , 26,
866.
Tang, W., Mokhtarian, P., & Handy, S. (2008). The role of neighborhood characteristics in the
adoption and frequency of working at home: Empirical evidence from Northern California.
(Paper, UCD-ITS-RR-08-21). UC Davis: Institute of Transportation Studies.
Tomaskovic-Devey, D., & Risman, B. (1993). Telecommuting innovation and organization: a
contingency theory of labor process change. Social Science Quarterly , 74, 367-385.
Tulgan, B., & Martin, C. (2001, October 3rd). Managing generation Y- part 2. Business Week
Online. Retrieved June 13th, 2012, from http://www.businessweek.com/stories/2001-10-
03/managing-generation-y-part-2
Valentine, G. (1989). The geography of women's fear. Area , 21, 385.
Valentine, G., & Holloway, S. (2002). Cyberkids? Exploring children's identities and social
networks in on-line and off-line worlds. Annals of the Association of American Geographers , 92
(2), 302-319.
Vana, P., Bhat, C., & Mokhtarian, P. (2008). On modeling the choice of work-hour arrangement,
location and frequency of telecommuting.
Wajcman, J. (2010). Feminist theories of technology. Cambridge Journal of Economics , 34,
143-152.
Walls, M., Safirova, E., & Jiang, Y. (2007). What drives telecommuting? The relative impact of
worker demographics. employer characteristics, and job types. Transportation Research Record
Journal of the Transportation Research Board, No. 2010 , 111-120.
Wiggins, S. (1995). Entrpreneurial enterprises, endogenous ownership, and the limits to firm
size. Economic Inquiry , 33, 54-69.
Yeraguntla, A., & Bhat, C. (2005). A classification taxonomy and empirical analysis of work
arrangements. Transportation Research Record: Journal of the Transportation Research Board,
No. 1926 , 233-241.
84
Yu, H., & Shaw, S.-L. (2007). Revisiting Hägerstrand's Time-Geographic Framework for
Individual Activities in the Age of Instant Access. In H. Miller, Societies and Cities in the Age of
Instant Access (pp. 103-118). Dordrecht, Netherlands: Springer.
85
Appendix Paper 1
Table 9:
Percentage Differences between Census 2010 and NHTS 2009 Data
Region 18-24 25-34 45-54 55-64 65-74 75-84 White Black Grad < 10K > 100K
Atlanta-Sandy Springs-
Marietta 5.1% 8.2% -3.0% -7.5% -9.9% -5.0% -25.6% 20.7% -4.0% 1.4% -6.1%
Austin-Round Rock-San
Marcos 8.4% 8.3% -5.3% -7.8% -7.4% -5.1% -8.9% 4.0% -7.9% 4.2% -10.9%
Baltimore-Towson 6.6% 7.9% -3.6% -2.5% -6.8% -4.3% -16.4% 16.4% -6.5% 4.5% -5.1%
Birmingham-Hoover 3.0% 10.9% -6.0% -5.3% -5.9% -8.6% -16.6% 16.9% -4.7% 1.7% -3.9%
Boston-Cambridge-Quincy 6.0% 6.5% -2.5% -6.6% -8.5% -5.0% -9.6% 4.7% -3.5% 2.7% -0.9%
Buffalo-Niagara Falls 6.0% 6.6% -1.5% -6.1% -7.5% -6.7% -11.9% 8.2% -2.9% 4.9% 1.9%
Charlotte-Gastonia-Rock
Hill 5.7% 8.6% -4.9% -9.1% -9.4% -5.0% -19.7% 13.9% -5.3% 0.5% -7.6%
Chicago-Joliet-Naperville 4.8% 8.7% -2.0% -7.2% -7.5% -4.9% -17.7% 8.5% -4.2% 2.2% -4.2%
Cincinnati-Middletown 5.0% 7.4% -2.3% -8.9% -8.1% -3.9% -9.0% 5.7% 3.4% 0.7% 5.5%
Cleveland-Elyria-Mentor 5.7% 8.1% -1.6% -6.2% -4.1% -3.5% -17.5% 14.7% -4.5% 4.5% -8.9%
Columbus 4.0% 11.8% -1.6% -9.8% -5.2% -4.4% -16.0% 10.6% -6.8% 3.7% -4.5%
Dallas-Fort Worth-
Arlington 6.0% 7.8% -4.4% -8.1% -8.0% -4.7% -14.1% 7.6% -6.5% 1.5% -11.1%
Denver-Aurora-Broomfield 7.1% 6.5% -5.3% -10.7% -4.7% -2.1% -9.2% 1.0% -9.1% 3.8% -11.3%
Detroit-Warren-Livonia 2.8% 6.1% -1.9% -4.1% -6.1% -5.7% -16.0% 14.3% -2.0% 4.1% -1.8%
Hartford-West Hartford-
East Hartford 7.4% 7.4% -0.7% -4.1% -7.9% -2.4% -14.1% 6.7% -7.1% 1.0% 1.2%
Houston-Sugar Land-
Baytown 5.6% 7.4% -4.5% -7.3% -8.5% -5.0% -11.7% 8.8% -7.6% 1.2% -13.6%
Indianapolis-Carmel 4.8% 5.8% -2.8% -6.3% -8.5% -4.7% -12.5% 8.3% -6.5% 1.4% -5.6%
Jacksonville 5.8% 7.9% -3.3% -8.8% -8.4% -5.9% -16.3% 14.4% -5.8% 2.1% -7.0%
Kansas City 4.1% 8.0% 1.3% -5.7% -3.0% -7.0% -11.2% 8.2% -7.8% 3.0% -5.2%
Las Vegas-Paradise 3.5% 7.4% -2.9% -6.4% -9.8% -1.3% -10.6% 4.6% -8.2% 2.5% -7.0%
Los Angeles-Long Beach-
Santa Ana 5.2% 7.8% -3.9% -6.4% -6.8% -5.1% -11.6% 1.7% -7.8% 0.9% -9.5%
Louisville/Jefferson County 4.3% 7.9% -1.5% -8.8% -7.9% -8.2% -11.3% 8.8% -0.2% 2.4% 2.3%
Memphis 5.1% 7.8% -4.5% -8.9% -9.8% -5.4% -30.9% 27.8% -9.3% 1.3% -7.9%
Miami-Fort Lauderdale-
Pompano Beach 5.1% 8.0% -1.5% -6.3% -8.3% -8.5% -8.6% 12.9% -7.8% 2.0% -7.1%
Milwaukee-Waukesha-
West Allis 4.5% 8.4% -4.2% -5.0% -4.1% -4.5% -12.5% 10.6% -3.1% 3.6% -5.7%
Minneapolis-St. Paul-
Bloomington 5.4% 8.5% -1.6% -8.1% -7.2% -3.3% -15.1% 6.5% -4.4% 1.1% -5.7%
Nashville-Davidson--
Murfreesboro —Franklin 6.1% 7.8% -4.6% -9.1% -5.6% -5.6% -12.6% 9.3% -4.4% 4.3% -8.2%
New Orleans-Metairie-
Kenner 5.8% 10.1% -3.5% -17.6% -1.5% -5.6% -26.6% 23.0% -7.6% 3.2% -14.9%
New York-Northern New
Jersey-Long Island 4.6% 8.7% -3.0% -7.0% -6.2% -4.8% -20.9% 9.8% -9.4% 2.4% -10.1%
86
Oklahoma City 8.0% 8.9% 2.9% -12.4% -8.6% -2.7% -8.7% 5.0% -6.1% 6.7% 2.6%
Orlando-Kissimmee-
Sanford 7.4% 8.6% -3.0% -6.9% -9.2% -7.0% -11.1% 9.9% -6.1% 0.7% -6.8%
Philadelphia-Camden-
Wilmington 5.1% 7.5% -4.6% -7.4% -8.3% -4.7% -14.9% 13.7% -8.1% 2.8% -7.8%
Phoenix-Mesa-Glendale 6.0% 7.2% -2.9% -7.6% -7.8% -5.7% -6.7% 2.8% -7.0% 3.4% -3.9%
Pittsburgh 4.9% 5.0% -1.6% -3.6% -7.2% -6.2% -7.6% 6.4% -2.9% 3.1% -3.4%
Portland-Vancouver-
Hillsboro 5.7% 9.1% -4.2% -4.5% -4.4% -2.4% -11.3% 1.5% -9.8% 1.1% -8.0%
Providence-New Bedford-
Fall River 6.2% 7.3% -1.4% -7.9% -6.5% -7.2% -10.1% 5.2% -8.4% 4.4% -3.9%
Raleigh-Cary 6.5% 7.5% -4.1% -9.8% -5.1% -2.3% -15.0% 11.9% -9.2% 0.5% -9.9%
Richmond 6.5% 7.5% -3.0% -7.9% -7.4% -5.2% -19.0% 17.5% -7.0% 1.7% -5.0%
Riverside-San Bernardino-
Ontario 4.1% 6.4% -3.4% -7.7% -5.8% -5.1% -5.9% 2.8% -4.7% 0.6% -0.7%
Rochester 6.4% 7.1% -1.7% -9.2% -6.1% -4.5% -10.1% 7.5% -4.5% 1.7% -1.9%
Sacramento--Arden-
Arcade--Roseville 5.9% 6.7% -4.3% -8.1% -6.7% -4.6% -15.3% 4.4% -7.9% 1.3% -5.8%
St. Louis 5.8% 6.0% 1.4% -7.0% -6.8% -5.4% -13.4% 11.5% -6.8% 2.8% -6.3%
Salt Lake City 3.4% 6.6% -2.7% -8.6% -4.0% -5.0% -7.5% 1.5% -1.6% 3.4% -5.8%
San Antonio-New
Braunfels 6.4% 7.0% -2.3% -8.7% -8.8% -6.5% -0.4% 1.9% -7.7% 3.7% -5.4%
San Diego-Carlsbad-San
Marcos 6.9% 8.4% -3.8% -7.0% -7.1% -5.2% -5.5% 2.3% -6.4% 1.8% -6.4%
San Francisco-Oakland-
Fremont 4.8% 9.0% -3.1% -7.7% -6.1% -5.5% -16.1% 3.6% -10.2% 1.5% -11.6%
San Jose-Sunnyvale-Santa
Clara 4.9% 8.8% -3.7% -7.6% -7.6% -4.4% -14.7% 1.2% -9.9% 1.0% -15.3%
Seattle-Tacoma-Bellevue 6.3% 9.4% -4.8% -3.8% -3.9% -3.1% -19.4% 4.4% -8.0% 1.8% -3.7%
Tampa-St. Petersburg-
Clearwater 6.0% 6.9% -0.3% -7.6% -8.5% -9.7% -9.5% 6.9% -5.4% 1.8% -3.2%
Virginia Beach-Norfolk-
Newport News 7.9% 7.6% -3.0% -7.9% -8.8% -5.1% -19.2% 16.6% -6.1% 1.1% -4.0%
Washington-Arlington-
Alexandria 4.8% 9.3% -5.7% -5.3% -5.1% -3.0% -25.2% 14.0% -3.8% 2.1% -10.0%
87
Paper 2: The un-fixed workplace: Interactions between the workplace and space-time
constraints on daily activities
Abstract Paper 2
This paper looks at how activity patterns are affected when the need to travel to the
workplace during traditional hours is reduced. Information communication technologies (ICTs)
have changed the nature of the travel game. Individuals are no longer faced with the same
constraints of having to travel places during their opening hours to access goods and services and
socialize. ICTs have also made working remotely from the workplace more possible. The
relaxation of these constraints means that individuals can multitask with greater ease and seek
out a greater set of activities resulting in patterns of increased fragmentation.
Using negative binomial and ordinary least squares regressions, as well as descriptive
statistics on the National Household Travel Survey for 2009. This paper shows that individuals
with flexible workplace options (those who have the option to work from home, the self-
employed and those who can adjust their work start times) fragment their work and non-work
activities more than individuals without the option. For the most part, they make more work trips
and trips outside of work and home (with the exception of those who have the option to work at
home). Those with flexible workplace options also spend less time engaged in work-related
activities and activities outside of work and home. The ‘freed’ up time within the traditional
workday is used to spend more time at home (for those who have the option to work at home)
and more time engaged in more non-work related activities (for the self-employed and those with
flexible start times).
88
Introduction Paper 2
Modern information and communication technologies (ICTs) are reinventing access
through transportation, proximity and connectivity- an accessibility paradigm promoted to better
suit modern cities.
1
Mobile applications are allowing for the proliferation and ease of car sharing
services such as Zip Car and Car2Go, peer-to-peer applications such as RelayRides where
individuals can rent out their personal cars to others, and i-sharing such as Lyft and Uber, where
individuals lend themselves as taxi drivers. The ability to obtain real-time information on various
modes of transportation and their schedules directly through one’s phone allow individuals to
better coordinate the timing of their activities. Other shared resource applications such as room
rentals, parking spaces, and personal equipment have occurred due to the networking and
logistical sophistication of mobile applications. ICTs have also enabled the virtual transportation
of communities through social applications such as mobile e-mail, texting, twitter, Facebook and
videoconferencing, to name a few.
These advances have brought access, transportation and proximity to the tips of our
fingers, putting into question traditional notions of space, place and time. They have decreased
the fixity (fixed nature) of activities occurring in traditional times and spaces. The increasing
sophistication of technology is rendering our home a more feasible place in which to conduct
work, shop and play. Technology in general has brought people closer to activities and services,
eliminating the need for certain kinds of trips and increasing and expanding our need for travel as
well.
1
http://www.um-smart.org/about/special_focus.php#2
89
Activity analysis, as affected by the recent radical developments in mobile ICTs, has
engendered renewed interest in geographers, planners and transportation scholars. ICT has been
theorized to lift some of the constraints associated with traditional activity behaviors such as
where and at what time the activities take place (Couclelis, 2003; Hubers, Dijst, & Schwanen,
2008) resulting in the fragmentation of activities and decreased ‘fixed’ levels of time and space.
Fragmentation, as advanced by Couclelis 2003, occurs when an unfinished activity is replaced by
another and then returned to it, or when several activities occur at the same time such as
multitasking. Space-time fixity analyses consider the binding nature of time and space on
activities and how this binding nature affects future activities (Schwanen, Kwan, & Ren, 2008).
While arising in the 1960s and 1970s, space-time analyses have resurged recently (Schwanen,
Kwan, & Ren, 2008).
These concepts are increasingly important to understand, as rapidly evolving ICTs
continue to influence urban transportation. How people are demanding transportation given the
tools available has implications for how we plan and manage urban spaces. The range and
availability of transportation options increase an individual’s choices and allows her the ability to
better balance work/life demands, reduce emissions related to transportation, be more efficient in
how she allocates time and money, and expands her work possibilities. Understanding these
phenomena in the context of ICT, allows for the facilitation of appropriate urban transportation
initiatives.
Few empirical studies on fragmentation and fixity exist mainly due to the limited
availability of data and the nascence of theory. Moreover, data are limited to surveys of sub-
samples of the population. Existing studies have examined how ICT, personal characteristics and
90
work-related variables influence the fragmentation of work (Alexander, Ettema, & Dijst, 2010;
Lenz & Nobis, 2007), how the use of ICT affects spatial flexibility (Schwanen & Kwan, 2008;
Line, Jain, & Lyons, 2011), and how fixity levels play out differently across genders (Schwanen,
Pokwan, & Ren, 2008). However, these studies assume that, apart from context such as socio-
demographics, gender and occupation, individuals face the same constraints with respect to key
anchors such as the workplace
Home and work are pivotal domains around which activities revolve. Modern ICTs
transform the home environment into a space where activities such as work, shopping,
networking and socializing can now take place in unison and without a need to travel outside the
home. ICTs have also made working at home more conducive. Through PCs, tablets, phones and
internet connections, many can undertake work at home or elsewhere with greater ease. This
eliminates, to some extent, the need to travel to the workplace, minimizes the social and resource
advantage of the workplace, and allows people to conduct work any place and any time.
This study considers the resulting activity patterns of individuals when constraints such
as the need to travel to the workplace (e.g. the ability to work at home) are lifted. Using a
national dataset, this paper conducts an in-depth analysis of the activity differences between
people who have the option to work at home, are self-employed, or can adjust their working start
times, versus those who do not have these options, considering personal factors, gender and
socio-demographics.
91
Theoretical Discussion
Hägerstrand’s Time-Geography
Actions and activities are constrained in both time and space. Early theories of how space
and time influence the way people travel arose in the 1970s and one of the most widely used
concepts is Hägerstrand’s time-geography (1970). Although criticized, the framework is useful
for understanding the interactions between individuals, society and technology (Schwanen &
Kwan, The Internet, mobile phone and space-time constraints, 2008).
Under Hägerstrand’s time-space framework, every individual faces a life path which is
inevitably constrained by the society in which she lives. Within her life path are shorter-term
paths such as day paths. These paths move through both space and time constrained by
physiological and physical necessities, as well as by private and social decisions. Hägerstrand
groups these constraints into three categories: capability constraints (biological, cognitive and
instrumental), coupling constraints (paths that are joined through individuals, materials and
tools), and authority constraints (laws, rules, power and norms).
ICTs inevitably relax some of these constraints through the creation of virtual space.
Virtual space (Janelle & Hodge, Information, Place and Cyberspace Issues in Accessibility,
2000) behaves differently than physical space by allowing individuals to connect and
information to be transferred without the need for travel, or the need for a physical place. In
order to move in physical space from one point to another, we must use transportation, in turn,
compromising time and future activities (Miller & Shaw, 2001; Yu & Shaw, 2007). When
navigating in virtual space, the need for transportation is reduced, if not eliminated, and we have
92
a wider array of activities between which to move back and forth. This phenomenon has
drastically been enhanced with modern ICTs.
Capability Constraints
Capability constraints limit activities due to the biological needs of the individual. These
constraints can be thought of as time-oriented, such as the need to sleep and eat, or distance-
oriented, which configure a set of ‘accessibility rings’ around an individual. The radii of the
rings are determined by the individual’s ability to move and communicate. The smallest ring is
the space around which an individual can move her arms. An individual usually has a home base
and rings are ultimately determined by how far and for how long she can leave the home base in
order to seek things such as work, socializing or obtaining goods.
Hägerstrand conceptualizes a time component to an individual’s space (see Figure 1).
Space and time form a prism representing the individual, where distance is on the x axis and time
is on the y axis. If an individual is stationary, say at work, the prism shrinks in proportion to the
length of the stay, as the individual cannot access other activities. When an individual is not
stationary, the prism expands, as the possibilities of when (to engage in an activity), where (the
range of accessible activities) and how (what mode of transportation) of the activity expand.
The figure below maps out such a path. The bold line represents a path taken. When the
individual is stationary, the prism shrinks. When the individual begins to travel, the prism
expands to the set of possibilities the individual can travel to. Hägerstrand visualized the
common Western prism as breaking into the portions before work, during lunch and in the
evening.
93
Figure 4:
Hägerstrand Space-Time Geography
Modern ICTs profoundly modify such capability constraints. In the figure above, the red
boundaries now become geographies and prisms of their own. Wireless networks, tablets, PCs
and smart phones, used in conjunction; affect both what can take place within the rings and at
what time. These technologies can bring many activities closer to the individual (Schwanen,
Kwan, & Ren, 2008) and in the space of the home. Now an individual may shop, work and
socialize at her fingertips, wherever she has a device and a connection. Whereas Hägerstrand
envisioned discreet sets of time revolving around the work day, many previously ‘stationary’
activities can be settings for multiple activities rendering a series of prisms possible. One can
now shop, while working, or work from the aisles of a grocery store. Absence is permitted in
many activities through advents, such as the ubiquitous and accessible voice-mail, e-mail,
texting, and various other forms of instant communication and messaging (Schwanen, 2007).
94
The breaking up of once discreet sets of time for activities into smaller sets that are traded
back and forth is coined ‘fragmentation’ as advanced by Helen Couclelis (2003). The
combination of the ability to partake in remote activities, the ability to more easily switch back
and forth between activities, and the ability to engage in several activities at once, as permitted
by ICTs, leads to fragmentation. Activities, with the aid of ICT, can be divided into their sub-
tasks and executed from different places, times, physically or virtually and in different orders.
ICTs have enabled waiting and travel time to be used for other activities (Lyons & Urry, 2005),
thereby, facilitating multitasking, which relaxes the zero-sum attribute of time (Kenyon &
Lyons, 2007). In other words, one activity is no longer completely sacrificed, in time, through
another. Involvement in one activity does not preclude the engagement in another.
Couclelis (2003) argues that the house is the most immediate and profoundly affected
space. The home, through ICT, now can become a place of working and shopping, as well as
socializing, education and entertainment, participation in community affairs, amongst others. The
Hägerstrand prism becomes increasingly more complex, as both individuals and households have
new flexibility and can distribute tasks and their components, via their home, or other locations
at varying time intervals. Work, once beholden to the tune of the factory whistle, can be
conducted throughout the day and throughout the week, including weekends and at home. In a
sense, this is somewhat a return to the pre-industrial era, where many worked from home and
multi-tasked to some extent, likely resulting in more episodes of fragmentation throughout the
day, different than during the industrial era. Because people worked more at home, they had
more control over the spacing of their activities and could combine non-work activities with
work-activities.
95
ICTs and their resulting applications have morphed and made fluid our rings and
resulting patterns. The range of possibilities for activities and transportation increases with the
more information we have available resulting in constantly expanding, contracting and shifting
rings. For example, applications that alert us to a range of restaurants while seeking a place to eat
can cause us to re-route ourselves, or make our travel more efficient. Applications that inform us
of the transportation options and schedules can allow us to more efficiently adjust our schedules
and plan accordingly (Bonsall, Traveller behavior: Decision making in and unpredictable world,
2004). Couclelis (2000) posits that fragmentation leads to an increase in travel demand as the use
of ICTs can increase our awareness of our surroundings and lead us to travel longer distances or
travel more in general. However, the increasing awareness of our environment may also lead to
decreased travel, as individuals can better select activities nearer to the home and places of work
and because individuals may spend less travel time ‘getting lost’.
Flexible work settings, such as the ability to work at home result in both a more sedentary
and mobile lifestyle, opening up the option to engage in discretionary trips, previously not
considered (Couclelis, 2003), resulting in a higher demand for travel (Lenz & Nobis, 2007).
Although it is likely that those who fragment their work patterns also tend to live in more urban
settings, they may live more on the peripheries of those areas and may be more disposed to
longer commutes (Alexander, Ettema, & Dijst, 2010; Hjorthol, 2002).
Coupling Constraints
Hägerstrand’s second set of constraints involves ‘others’. Coupling constraints also
define paths by time, place and extent of the activity. These constraints surface when an
individual must ‘join’ another person’s materials or tools in order to consume, make transactions,
96
and produce. When two or more such entities intersect, they are grouped as bundles. Bundles can
be spatially thought of as shops, offices, factories, classrooms, etc. An individual’s freedom is
limited in bundling. Although she may be able to choose her type and place of work, she must
adhere to the principles and rules of others, such as supervisors. Individuals are also constrained
by what is permissible in time and space. Historically, shops, banks and doctors had discreet
accessible times and activities had to be engineered to meet such time frames (the individual
must take off from work to visit the doctor).
Modern day ICTs loosen coupling constraints. Individuals can be joined from virtually
anywhere and materials can be brought to the person. Not only can bundles be brought to the
person without the need to travel, travel itself can engender bundling. This affects both physical
and virtual materials. ICTs can decrease one’s need to travel to obtain material goods yet the
material goods still need to be brought to them. ICT can also decrease one’s need to travel to
seek services, as these can be brought to them in a virtual domain. Individuals with networked
smart phones and tablets can work and communicate while in transit. Even the traditional
concept of the doctor’s or bank office can be adjusted within one’s personal space in their home
base. Patients can seek the advice of their doctor through the web, or over the phone. Most bank
services and transactions can also take place within one’s inner ring.
ICTs have modified the way in which we plan and coordinate social and professional
interactions. Instant communication allows us to keep each other better informed when we are
running late, so that others can adjust their own schedules (Bonsall, Traveller behavior: Decision
making in and unpredictable world, 2004). ICT allows for increased flexibility in activity
arrangements as it allows the working out of schedules, details of meetings and work, and
97
planned activities through repeated texts and e-mails. Social networks can be accessed and
mobilized quickly (Rheingold, 2003). This redefines the importance of space in terms of social
connectivity, shifting the nodes to individuals (Kwan, 2007).
Authority Constraints
Authority constraints are the final grouping of constraints in Hägerstrand’s activity
analysis. Authority constraints form ‘domains’, which have both a time and space component
and the domain itself is in the control of a person, or persons. Authority constraints can be
thought of as legal rules, the domain of an institute, or at a large scale, such as that of a nation.
Domains are influenced by a hierarchy; those in more power can restrict the possibilities of
others. The more powerful may also be immune to such restrictions.
Authority constraints influence how and where one can use ICTs to modify activities.
How much authority or status in the workplace, influences her ability to work at home when she
chooses. People can command more authority if their skill set is highly valuable. ICTs also
modify the authority constraints of many institutions. Universities, for example, are providing
more on-line content, which can be accessed at times other than when the courses are taught.
Information itself, such as provided by a library, can now be accessed at virtually any time.
Some argue that ICTs can reduce the hierarchies between men and women (Madge & O'Connor,
2006).
How and when one can adjust their times of work depends on the management models of
employers and also on the particular situations of the employees and how they perceive their
relationships with work and organizations (Garhammer, 1995; Breedveld, 1998; Brannen, 2005).
The worker’s ability to negotiate and articulate her working demands, along with her social
98
skills, determines her levels of flexibility (Garhammer, 1995). The more educated tend to have
better control over their work time (Breedveld, 1998). The household’s ability to absorb the
flexibility (Garhammer, 1995) and individual’s ‘ethics’ related to work and family determines
their flexibilization (Brannen, 2005).
Constraint Moderators
Ultimately, the degree to which constraints are fixed or flexible and the resulting activity
patterns are contingent upon context, such as: the characteristics of the person, technologies
available and used, types of activities, the timing of activities and social environments
(Schwanen, Kwan, & Ren, 2008; Schwanen & Kwan, 2008; Alexander, Ettema, & Dijst, 2010;
Lenz & Nobis, 2007). Activity ‘stickiness’ is due to the level of personal commitment, as well as
external factors, such as social norms, physical and economic circumstances (Cullen & Godson,
1975; Cullen, Godson, & Mayor, 1972). The individual’s background, such as her age, job and
the composition of her household is important in understanding the fixed levels of her activities
(Cullen & Godson, 1975). Higher-level fragmenters tend to be highly educated and higher wage
earners, than groups who fragment less (Hjorthol & Gripsrud, 2009; Alexander, Ettema, & Dijst,
2010; Lenz & Nobis, 2007). One’s perception and ability to navigate both space and time
evolves and is modified by the particularities of the individual, the types of activities and modes
of access such as ICT (Schwanen, Kwan, & Ren, 2008).
Moreover, ICTs, which relax constraints, do “not have inherent properties or universal
impacts” (Valentine & Holloway, 2002) pg. 316. The availability of ICT does not immediately
or automatically lead to changes in behavior, but rather behavioral changes are a consequence of
the individual within a societal framework (Lenz & Nobis, 2007). The use of time is moderated
99
by ICT, but heavily contingent on the individual and how much her social circles and work
environment enable the shift or dissolution of strict time schedules and places (Lenz & Nobis,
2007). Innovations, especially with technology, may be more facilitative, that is, facilitate
existing behaviors, needs and personalities, rather than determine them.
Use of ICT and age
Use of ICT has been linked to levels of fragmentation. Heavy users of the internet and
mobile devices fragment more (Lenz & Nobis, 2007) and engage in more spatially and
temporally fragmented work patterns (Alexander, Ettema, & Dijst, 2010; Hjorthol & Gripsrud,
2009; Hubers, Dijst, & Schwanen, 2008).
Yet, use of ICT is contingent upon many factors, such as, age, education, income and
gender. Younger generations express better cognitive ability to adapt to newer technologies
(Shepard, 1999; Skirbekk, 2003; Castells et al., 2007), are more eager internet consumers
(Hjorthol & Gripsrud, 2009) and are more willing to adopt progressive technologies in the
workplace (Bailey and Kurland, 2002). Members of generation X and Y, those aged roughly
from 18 to 46 in 2012, demand greater work-life balance than older generations and are more
self-reliant and individualistic (Eisner, 2005). They are independent, enjoy challenge, and prefer
flexibility in the workplace (Crampton & Hodge, 2009; Tulgan & Martin, 2001). Gen Y is seen
as transformational (Haynes, 2011). Because this generation is constantly connected through
their mobile devices and various social media applications, they are transforming both social and
business norms. Many organizations now use social media applications as their method of
outreach, engagement and advertising. Younger generations are also more likely to fragment
100
(Lenz & Nobis, 2007; Alexander, Ettema, & Dijst, 2010; Hjorthol & Gripsrud, 2009), as
activities become more fixed in time the older one becomes (Schwanen, Pokwan, & Ren, 2008).
Gender
Sex and gender are ‘major axes’ in space-time constraint differentiation (Schwanen &
Kwan, 2008). Some contemporary feminist theories view cyberspace as liberating, transforming
gender relations and equalizing hierarchies (Wajcman, 2010). However, studies have found that
women partake in more fixed activities (activities which can’t be modified in space or time) than
men (Pickup, 1988; Hanson & Pratt, 1995; Schwanen & Dijst, 2003).
Gender tends to determine levels of flexibility, partly, because women partake in more
activities related to the household and child rearing, which tend to be more fixed in both space
and time and also, because gender interacts with other activity factors such as time of day and
type of activity (Schwanen, Kwan, & Ren, 2008). Women’s time slots are more compromised
than those of men (Schwanen, On 'arriving on time', but what is 'on time'?, 2006). Kwan (2000)
finds that women face more constraints on time and space than men, even when taking into
account employment status and commute times due to household and child-care activities.
Schwanen, Kwan, & Ren (2008) find that the Internet does not make traditional female and male
domains more equal and balnced, rather, it tends to confirm previously exisiting biases and
making space more constrained for females. Other studies find that women’s spaces and time
frames are restricted due to levels of fear (Valentine, The geography of women's fear, 1989;
Pain, 1997; Koskela, 1999). Women also make choices related to child-rearing, which tend to
decrease their levels of flexibility. For example, women more often hold part-time jobs and,
101
because these jobs tend to be lower in status, decrease their ability to have control over their
work times (Odih, 1999).
Since men and women face different constraints (for example, women are more oriented
towards family responsibilities), most studies on work fragmentation and telework consider
gender as an explanatory variable. Evidence shows that women fragment less than men,
especially in terms of work (Hubers, Dijst, & Schwanen, 2008; Alexander, Ettema, & Dijst,
2010) and they also partake in less internet use and shop on-line less frequently than men (Farag,
Dijst, & Lazendorf, 2003).
From the discussion above, the increasing sophistication of and access to modern
information communication technologies, likely modifies modern day travel by reducing the
fixed nature of time and space and by providing travelers with more information and therefore,
more opportunities. ICTs bring access within the reach of our personal space changing the zero-
sum attribute of activities by making once stationary activities such as work or shopping
available to us at all times. ICTs also bring people and other domains into the reach of our
innermost ring rendering services like banking, medicine, universities and libraries available at
most times throughout the day.
The reduced fixity of activities, as engendered by ICT, is likely to result in a greater
fragmentation of time. That is, both trips and activities will be dispersed into their sub-
components throughout the day. Instead of traditional work, shopping, doctors’ visits and so
forth being conducted during discrete sets of time in tune with traditional working hours,
activities will be dispersed throughout the day (chopped-up), resulting in both an increase of
102
activities and of trips made to such activities. As activities and trips increase, activity durations
should decrease.
Yet, technology, only theoretically, can modify travel behavior. The way in which
individuals use technology to engage in travel and activities is contingent upon a variety of
social, personal and demographic factors. The higher educated and income earners reflect a
group of people who have skill sets that demand more authority. Greater authority leads to more
control over one’s time. People also carry different attitudes toward their use of personal and
family time. Young adults are likely to have better adaptive skills to technology, further
expanding their possibilities. They are likely to be unencumbered by family responsibility.
Gender also affects the fixity of one’s activities, as women tend to be more responsible for
childcare and other household matters, binding a female’s domain of space and time more so
than that of males.
Apart from the personal, social and demographic factors mentioned above, which result
in different forms of constraints, lies a major constraint obstruction to the timing and spacing of
activities throughout the day: work and the workplace. The ability to work at home and other
work aspects, such as self-employment status, fundamentally change the nature of home as a
pivot around which activities revolve. They also provide more flexibility in how other activities
are conducted throughout the day. Individuals who practice flexible workplace, such as
conducting work remote from the workplace (home or coffee shop) during traditional working
hours, or arriving at work later in the day, can better determine the use of their time, dispersing
and modifying the shrinking prism with respect to stationary work throughout the day and
throughout the week. They are likely to engage in the timing of trips, differently than individuals
103
who are required to be physically present at work during set times of the day. As a consequence,
the home will be a key place of activity consumption and because the physical constraint of the
workplace is removed, individuals will be able to seek out activities throughout the workday.
ICTs affect how all of us operate with regard to capability constraints. They minimize
these constraints for everyone who has access to mobile devices and networks. Most of us now
can summon goods and services and socialize at the tips of our fingers, leading us to increasingly
fragment our activities. So, although, most individuals will experience modification of traditional
activity behaviors as a result of diminished capability constraints, the manner in which ICTs
modify coupling and authority constraints, will depend to a large extent on an individual’s
context.
This paper hypothesizes that those who engage in flexible workplace practices will also
engage in more fragmentation, than the population who faces a fixed workplace, controlling for
socio-demographic and personal factors as their coupling and authority constraints with respects
to work are hypothetically reduced. Therefore, those with flexible workplace options are likely to
fragment more their work episodes, spacing work activities over the day, resulting in more
frequent work episodes of shorter duration. The freeing up of work time allows for the
introduction of more discretionary activities and trips. Finally, those who engage in flexible
workplace options are likely to use their home as a purposeful activity hub, thereby spending
more time, on average, in the home.
Examining a subset of people who are allowed workplace flexibility and their travel
patterns, allows us to better explore the workplace in the realm of activity analysis and shed light
on how coupling and authority constraints may be different for these types of workers.
104
Workplace practices are likely to continue changing as both technology and society evolve.
There is likely to be more and more part-time, contract and self-employed work available.
Therefore, more general and traditional travel patterns are also likely to shift. This subset of
people allow us to understand how the future may look.
Data
This study employs the 2009 National Household Transportation Survey (NHTS). The
NHTS was conducted by the US Department of Transportation from March 2008 to May 2009.
The survey was completed for 150,147 US households and all individuals over the age of 5
within the household. The survey collected information on travel patterns, work-related
information, and household demographics. Datasets are available at the household level, person
level, vehicle level, and daily trip level.
The sample is intended to be nationally representative through the random sampling of
list-assisted dialing banks, so that each household with a landline telephone had equal probability
of being selected. The sampling frame includes 100-banks with sets of 100 land-line telephone
numbers with the same first eight digits. Telephone numbers were then sorted by the ten Census
Divisions and then by metropolitan/non-metropolitan areas. Metropolitan areas were then sorted
by population of the Metropolitan Statistical Area (MSA) and primary metropolitan statistical
area (PMSA). Within these, telephone numbers were sorted by those in the central city, then
those in the non-central cities.
Information such as household income, education, gender, age, travel behaviors and
workplace characteristics was gathered along with information from a daily trip diary for each
household member over the age of 5. Daily trips were recorded through a 24-hour interval,
105
starting at 4:00 am to 3:59 am the following day. The diary asked at what time, where and by
what mode the first trip was made and all subsequent trips in the 24-hour period. Imputations
such as travel times and dwell times (the time spent at an activity, derived as the time a person
stops from the time they begin travelling again) were calculated by the NHTS. A computer
assisted telephone interview (CATI) program was employed to automatically assign a travel date
to each household based on a 1/7
th
weight for each day of the week, so that variation in travel by
day of the week was balanced. Travel days were assigned from March 28, 2008 to April 30,
2009.
The unit of analysis is the individual. Individuals had to meet several criteria in order to
be used in the analysis, to compare individuals who face similar constraints. Firstly, only
employed individuals were used. Secondly, only individuals who were surveyed during the
weekday and worked full-time were used in order to be able to accurately compare like periods
of time for activity analysis. Individuals face different constraints on their activities during the
weekday, than they do on the weekend. Also, people who work part time (less than 35 hours per
week as defined by the NHTS) will necessarily face fewer constraints than those who must work
at least 35 hours per week. With respect to socio-demographics, only individuals with a valid
entry for household income, age and education were used. Finally, in order to examine
workplace flexibility, only individuals with a valid response to the following questions were
used:
Ability to work at home: “Do you have the option of working at home, instead of going
into your primary workplace?”
106
Ability to adjust work start times: “Do you have the ability to set or change your own
start work time?”
Self-Employed: “Are you self-employed?”
This data set does not allow me to examine how modern day ICTs affect capability
constraints in general and the resulting fragmentation of activities as a cause of interruptions and
multi-tasking. For instance, while at work, an individual may socialize by calling her friends, or
she may partake in some on-line banking, or on-line shopping- the result of the modification of
capability constraints through ICTs. Likewise, any individual may engage in work activity while
at home, or may shop on-line. Information on how individuals interrupt their activities or
multitask is not available in the NHTS dataset. Trips and activities are discrete.
Therefore, this study will focus on what it can measure: the effects of diminished
coupling and authority constraints on populations who are allowed flexible workplace options,
versus those who are not, through trips made out of both the workplace and the home. These are
the activities that ‘fit around’ work and home for a specific purpose (coded as personal, social,
shopping etc.) and serve as a proxy to how much overall fragmentation may be occurring. This
study will also look at how work episodes themselves are fragmented and how much time is
spent at home.
Because we have a limited amount of time during the day to travel and engage in
activities due to capability constraints (biological constraints), and coupling constraints
(household needs and service availability) understanding travel in the form of both quantity of
trips taken and time spent engaging in an activity (dwell time) in necessary. Shorter dwell times
107
outside of the home allow for more travel to occur. Longer dwell times in the home allow for
more activities to occur in the home ameliorating coupling and capability constraints.
Analysis Strategy
Overview
Fragmentation is assessed by looking at three dimensions: work-related travel (number of
trips, duration of a work activity and work travel time); non-home-and-work related travel
(number of trips, duration out of home and work and travel time to destinations not related to the
work or home); and total time spent in the home.
The hypothesis is that those who are allowed the flexibility to adjust work start times and
work at home as well, as the self-employed, can organize their time according to their needs and
not necessarily according to traditional working hours. This organization, or use of time, will
result in more discretionary trips and activities of shorter duration, particularly with regards to
work.
The effect on travel behavior as a result of reduced workplace constraints is examined
through controlling for key constraint moderators such as socio-demographics, life cycle
circumstances, ICT use, and work related variables on trip variables. The dependent trip
variables are measured by:
1. Total number of trips taken throughout the day (out of work and home)
2. Average time spent for an activity throughout the day (out of work and home)
3. Total number of work related trips throughout the day
4. Average time spent for a work trip throughout the day
5. Total time spent at home throughout the 24 hour period
108
6. Total number of trips made to the home throughout the day
7.
Descriptive analyses are used to gain general insight on the sample. Ordinary least square
regressions are used to analyze the three average time variables and negative binomial
regressions are used to analyze the three count variables. The following relationships with trip
characteristics are examined:
(a) Personal attributes: education, age, gender
(b) Household factors: household income, household size, age of smallest child and
whether or not a one, or a two-parent household, household vehicle count
(c) Geographical context: population density of census tract
(d) ICT use: web use, number of internet purchases made in a month
(e) Work factors: Ability to work at home, self-employment status, ability to adjust
work start-time, part-time status
The differences between individuals who are allowed the option to work at home and
those who do not, is examined by testing the significance of the dummy variable (1=ability to
work at home) and interactions with this variable on other work factors and gender. This enables
the analysis of how the ability to work at home interacts with various income levels on trip
patterns.
109
Variable Definitions and Measurement
The following provides a brief list of the variables used. For a more in-depth description
of how the variables were measured and scales changed from the survey, see Appendix for paper
2 (pg. 146).
Household Income
Education
Age
Household Size
Square mile Population Density
Web Use
On-line Purchases
Walking, Biking and Public Transit Trips
Household Vehicle Count
Average Dwell Time
Number of Trips:
Calculated as the total number of trips an individual made during the survey day. These
are separated by different types of activities:
1. Home: Return home
2. Work: Go or return to work; Attend Meetings; Other work-related affairs
3. School/Daycare/Religious Activities
4. Medical/Dental Services
5. Shopping: Buying of goods or services such as dry-cleaning
6. Social/Recreational: Exercise; Play sports; Visit friends or family; Entertainment
7. Personal: Use of professional services such as an attorney; Grooming (hair and nail);
Pet care and walking; Attend PTA meetings or local government meetings
8. Transport Someone
9. Meals (including coffee)
10. Other
Time spent engaged in activity (dwell time) is examined through ordinary least squares
regressions. Because all activities in the dataset have positive values (there are no zeros or
negative values) and because the variable is continuous, ordinary least squares regression is
appropriate for examining activity time (dwell time). There are zero values for many of the trip
110
categories and because these data are count data, negative binomial analysis was used for trip
counts.
Negative Binomial Analysis
Count data require models other than linear regression models, as these models assume a
continuous distribution from negative infinity to positive infinity. Count data are censored and
cannot have negative values. They are generally in the integer form of 0,1,2,3…., rendering them
non-continuous, which result in biased or inconsistent parameters (Jang, 2005). Count models
offer other benefits: 1. Linear-regression models can predict negative values, impossible for trip
frequencies and the resulting predictions may be over or under-inflated and 2. They produce
discrete probability distributions of trip frequencies, rather than forecasted frequencies (Jang,
2005).
Commonly used count models are the Poisson and the Negative Binomial. Issues arise
when using the Poisson model, as it assumes equality of mean and variance in the dependent
variable (equi-dispersion) and it is more common to have over-dispersion (Jang, 2005) and the
standard errors may be biased.
Poisson regressions ultimately assume that populations with the same values for the
outcomes have the same underlying rates of the outcome, which is not possible, unless every
explanatory variable is accounted for and residual variation omitted. The negative binomial
model accounts for the residual variation by the addition of an error term to capture the
heterogeneity, which has not been measured:
var(Yi)= exp(Yi) + øexp(Yi)
2
111
Differences in socio-demographics and travel behaviors by Flexible Work Options and
Gender
This section discusses the descriptive statistics between groups that have flexible work
options and those who don’t. The total sample size used is 53,612 individuals. Around 47% of
the population is female and 53% male. Around 16% of the sample is allowed the option to work
at home. Amongst this group, 30% are female and 70% male. A larger percentage of the sample
population can adjust their working start time (46%) and females represent a larger share of this
group at (40%). Finally, 9% of the sample population is self-employed and 30% of these are
female. 57% of individuals who are allowed the option to work at home are also self-employed.
Table 10:
Sample Population Characteristics
Option to Work at Home Option to Adjust Start Times Self-Employed
Male No Option 22,803 13,468 24,804
Female No Option 22,131 15,411 23,887
Male w Option 5,448 14,783 3,447
Female w Option 3,230 9,950 1,474
Total 53,612
Total Females 25,361
Total Males 28,251
In general, those who have the option to work at home, the self-employed and those who
can adjust their work start time, compared to those who do not, on average:
Have higher incomes
Have higher levels of education
Are older
Live in denser census tracts in terms of population (with the exception of the self-
employed)
Use the web more (with the exception of the self-employed)
112
Table 11 lists the mean values for socio-demographic and web use variables. The table
separates individuals by three groups, those with the ability to work from home, the self-
employed and those with the ability to adjust work start time. Within these three groups,
individuals are separated by whether or not they have the option or status.
In terms of income, individuals with flexible workplace options earn more than
individuals without flexible workplace options. The highest earners of the three groups are
individuals who are allowed the option to work at home, earning on average $77K per year. The
lowest earners are those who do not have the option to adjust their working start hours, earning
on average, $60K per year. The largest discrepancy lies in the income differences between those
who are able to work at home and those who are not (14K difference). Those who are self-
employed, versus those who are not, earn similar incomes. There is a 12K income difference
between those who can work flexible hours, versus those who can’t.
The same pattern holds for education. Flexible workplace groups are more educated than
non-flexible groups. The most highly educated are those with the ability to work from home,
followed by those with flexible work start time. The mean-level education for all groups is some
college. The educational average for those who can work at home is right at the level of a
bachelor’s degree.
In general, individuals with flexible workplace options live in slightly denser population
census tracts, with the exception of the self-employed. The self-employed live in the least dense
census tracts of all groups. The self-employed are also the oldest of the groups, at an average of
113
53 years. Those with flexible work start and place are slightly older on average than those
without flexibility. Average household size is similar for all groups ranging from 2.83 persons
per household, to 2.94 persons per household.
ICT usage holds strong patterns. Those who have flexible options, on average, use the
web more than groups without flexible options, with the exception of the self-employed, who use
the web less on average, than those who are not self-employed.
In summary, those with flexible workplace options earn higher incomes, are more highly
educated and are older than groups who do not have flexible options. For the most part, they tend
to live in denser areas and use the web more. The self-employed behave slightly differently,
living in the least dense census tracts of all groups and taking likewise the least amount of transit
trips.
Table 11:
Socio-demographics and Self-Reported Trips by Mode by Flexible Workplace Options T-tests
Work at Home Self Employed Adjust Work Start Time
Units
No Yes No Yes No Yes
Number
45,034 8,578 48,787 4,825 28,950 24,662
Household
Income
1= < 5K, Increments of 5 K until
18 which is >= 100K
13.69* 16.11* 14.06 14.35 13.12* 15.21*
Education
1 =< HS; 2=HS; 3=Some
College; 4=Bachelors;
5=Graduate
3.36* 3.99* 3.45* 3.55* 3.24* 3.72*
Age Years 47.64* 48.62* 47.32* 52.51* 46.91* 48.83*
Density of
Census Tract
Population per sq. mile
3,120 3,385 3,194* 2,850* 3,135* 3,196*
Household
Vehicle Count
Number of cars
2.56* 2.49* 2.53* 2.73* 2.55* 2.55*
Web Use
1=Never; 2=1/month;
3=1/week; 4=Several times a
week; 5=Daily
4.37* 4.86* 4.46* 4.33* 4.25* 4.68*
*Statistically significant at the .05 level
114
Differences in trip behaviors by Option to Work at Home and Gender
This section discusses daily average trips and dwell time of groups that have the option
and those who do not (Table 12). These differences are explored as well through gender (Table
13).
The dimensions of trips examined are all trips made throughout the travel day and the
dwell time of activities accessed during these trips. Work trips look at the number of work
related trips (includes meetings, travel to the workplace and other work related trips). Total trips,
not to the home or the workplace, include all out of home and out of work trips. Finally, total
time spent at home is examined. This includes the time spent at home from 4 in the morning of
the travel day, until the individual leaves home for the first trip, any time spent at home
throughout the travel day and finally the remaining time spent at home from the moment the
individual returns to the home and makes no further trips, until 4 in the morning the following
day of the 24 hour period.
The t-tests in table 12 indicate that individuals with flexible workplace options fragment
their work more than those without the option. Although work episodes at the home, or places
other than the workplace, cannot be measured. Individuals with flexible workplace options are
making more work-related trips outside of the home (an average of .30 more trips for all three
groups). The self-employed make the most work-related trips, followed by those with the option
to work at home and those who can adjust start-working times. These out of the home work-
related activities are also shorter in duration (an average of 90 minutes shorter for all three
groups).
115
More work-related trips of shorter durations could be explained by the nature of the work
in which these individuals are engaged (e.g. more meetings are required, or the self-employed
work in service-related jobs), or that individuals are simply engaging in more fragmented work,
because they are spacing out work throughout the day in blocks of time, rather than continuous
episodes. Those with the ability to work from home may be less constrained by coupling
constraints (i.e. the need to meet others) and therefore spend more time at home working.
Evidence of fragmentation for flexible workers also occurs in trips not to the work or to
home (e.g. personal trips, social trips, school trips and transporting others). Individuals with
flexible workplace options make more trips out of home and work (an average of .30 more trips
for all three groups). However, they are not spending less time engaging in out of home and
work activities, than groups who are not allowed flexible workplace options. This is an
indication that relaxed constraints of the workplace and ICTs are inducing more trips. Those with
the ability to work at home, make the most amount of out-of-home and work trips, followed by
the self-employed and those who can adjust their working start time.
When it comes to the home, those with the ability to work at home spend the most total
time at home over a 24-hour period. This means, they are either leaving the home later than the
general population and returning at roughly the same time, or they are returning home earlier and
leaving at the same time. Those with the option to work at home spend around 32.4 minutes
more at home, than those without the ability and about 12 minutes more than the self-employed
and those with the ability to adjust work start time. More time spent at home may be a result of
loosened coupling constraints.
116
Lastly, when separated by gender, those with the ability to work at home make the most
trips, but females within this category make more trips than males. Males with the ability make
the most work-related trips and females make the most non-home, or work related trips. Males
with the option to work at home spend the most total amount at home, followed by females with
the option. So, workplace options matter, but gender also plays a significant role in what types of
trips are made and time spent at home.
In summary, the descriptives here show that those with flexible workplace options are
substituting out of home-work during a traditional travel day, with more trips for other activities
and with time spent at home. Those who have the option to work at home spend the most amount
of overall time at home and they also make the most trips not related to work or home. The self-
employed make the most amount of work-related trips and spend the least amount of time
engaged in a work-related trip. Those with flexible work start times follow the behavior of those
with the ability to work at home, but to a much less extent.
Table 12:
T-tests for Trips by Trip Type and Flexible Workplace Option
Option to WAH Self-Employed Flex Start Times
No Yes No Yes No Yes
Work Trips 1.27 1.42 1.25 1.73 1.17 1.45
Average Work Dwell Time (min) 421.06 336.00 418.69 301.30 439.79 371.39
Average Work Travel Time (min) 23.27 26.32 24.13 19.90 23.21 24.35
Total Trips not to Home or Work 1.71 2.12 1.76 2.02 1.67 1.91
Average Dwell Time for Non-
Home and Work Activities (min)
47.66 48.28 47.69 48.50 47.58 47.98
Total Time Spent at Home
(hours)
14.50 15.04 14.56 14.84 14.56 14.62
Not Statistically Significant
117
Table 13:
ANOVA tests for Trips by Trip Type and Flexible Workplace Option
Male No
Option
Female No
Option
Male w
Option
Female w
Option
F-Test
Total Number of Trips 3.51 3.68 4.21 4.31 39.1*
Total Work Trips 1.36 1.18 1.50 1.29 88.7*
Total Trips not Home or Work 1.54 1.90 2.00 2.33 14.2*
Total Time Spent at Home 14.55 14.45 15.08 14.98 51.9*
*Groups significantly different at .05 level
Relative Distribution of Trips Made by Time of Day for Ability to Work at Home
In order to visualize how coupling and authority constraints affect flexible workers
differently, the following figures graph the distribution of trips made throughout the day by time
of day for individuals who have the option to work at home and individuals without the option.
Total trips are divided by the two populations for each hour of the day. In each graph, the solid
line represents individuals who are not allowed the option to work at home and the bars represent
individuals who are allowed the option. Several interesting patterns emerge from the following
charts.
In terms of work trips, two general peaks are seen, one in the AM and one in the
afternoon around traditional lunch hours. Those who have the ability to work at home start their
work trips, on average, later than those who are not allowed the option. They also make on
average more trips during traditional lunch hours.
118
Figure 5:
Average Distribution of Work Trips
For shopping trips, there is one large peak for those without the option to work at home in
the early evening, after traditional working hours. Here, they make more trips than those with the
option. Those with the option to work at home have two peaks, one around traditional lunch
hours and one in the early evening. They make more trips throughout the day, than those without
the option, until the early evening.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
12-1 AM
1-2 AM
2-3 AM
3-4 AM
4-5 AM
5-6 AM
6-7 AM
7-8 AM
8-9 AM
9-10 AM
10-11 AM
11-12PM
12-1 PM
1-2 PM
2-3 PM
3-4 PM
4-5 PM
5-6 PM
6-7 PM
7-8 PM
8-9 PM
9-10 PM
10-11 PM
11PM-12AM
Average Distribution of Work Trips
Can work at Home
119
Figure 6:
Average Distribution of Shopping Trips
Personal trips show even different patterns between the two groups. These are trips
related to personal affairs such as hair and nail care, visiting an attorney and so forth. For those
without the option, there is one large peak in the early evening, yet, there is no real peak for
those with the option to work at home. Those with the option make the same consistent amount
of trips throughout the day, from morning until evening. They make almost double the amount of
trips throughout the day than those without the option.
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
12-1 AM
1-2 AM
2-3 AM
3-4 AM
4-5 AM
5-6 AM
6-7 AM
7-8 AM
8-9 AM
9-10 AM
10-11 AM
11-12PM
12-1 PM
1-2 PM
2-3 PM
3-4 PM
4-5 PM
5-6 PM
6-7 PM
7-8 PM
8-9 PM
9-10 PM
10-11 PM
11PM-12AM
Average Distribution of Shopping Trips
WAH NWAH
120
Figure 7:
Average Distribution of Personal Trips
Medical trips and ‘other’ trips follow the same distributions. These trips are bunched in
the morning, before traditional working hours and in the early evening, after traditional working
hours. Those who have the option to work at home make more trips than those without the option
and most of their trips are before traditional working hours. In terms of other trips, those with the
ability to work at home make many more trips throughout the day than those without the option.
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
12-1 AM
1-2 AM
2-3 AM
3-4 AM
4-5 AM
5-6 AM
6-7 AM
7-8 AM
8-9 AM
9-10 AM
10-11 AM
11-12PM
12-1 PM
1-2 PM
2-3 PM
3-4 PM
4-5 PM
5-6 PM
6-7 PM
7-8 PM
8-9 PM
9-10 PM
10-11 PM
11PM-12AM
Average Distribution of Personal Trips
Can work at Home
121
Figure 8:
Average Distribution of Medical Trips
0
0.002
0.004
0.006
0.008
0.01
0.012
12-1 AM
1-2 AM
2-3 AM
3-4 AM
4-5 AM
5-6 AM
6-7 AM
7-8 AM
8-9 AM
9-10 AM
10-11 AM
11-12PM
12-1 PM
1-2 PM
2-3 PM
3-4 PM
4-5 PM
5-6 PM
6-7 PM
7-8 PM
8-9 PM
9-10 PM
10-11 PM
11PM-12AM
Average Distribution of Medical Trips
WAH NWAH
122
Figure 9:
Average Distribution of Other Trips
The graphs reveal insights with respect to how authority and coupling restraints are
slightly lifted for those who have the option to work at home. Because those with the option to
work at home are not bound to the workplace, time is freed up during the day to engage in other
types of activities. These individuals inherently have more authority in the workplace, as they are
granted permission to work remote from the workplace and don’t have to respond to authority, if
there is any, in a physical realm. The relaxation of these two constraints is seen in the above
graphs, where activities other than work, personal, shopping, medical and other are spaced and
0
0.0005
0.001
0.0015
0.002
0.0025
0.003
0.0035
0.004
0.0045
0.005
12-1 AM
1-2 AM
2-3 AM
3-4 AM
4-5 AM
5-6 AM
6-7 AM
7-8 AM
8-9 AM
9-10 AM
10-11 AM
11-12PM
12-1 PM
1-2 PM
2-3 PM
3-4 PM
4-5 PM
5-6 PM
6-7 PM
7-8 PM
8-9 PM
9-10 PM
10-11 PM
11PM-12AM
Average Distribution of Other Trips
WAH NWAH
123
accessed differently for those who have the option to work at home, versus the rest of the
working population not allowed the option. Those who have the option to work at home
consistently make more trips than those without the option. Workers without the option to work
at home, on average, bunch their trips either before the start of a typical workday, or at the end of
the typical work day. Workers with the option are not as tied to saturating these types of trips
during these times and access non-work activities more consistently throughout the day.
Grouping does occur for this group around traditional working hours, but to a far less extent.
Regression Results
As mentioned in the introduction of this paper, the degree to which constraints are fixed
or flexible, are contingent upon an individual’s context (household characteristics,
demographics, gender) as well as her workplace constraints. In order to properly examine how
workplace trends affect travel behaviors by modifying coupling and authoritarian constraints,
aspects of individual’s characteristics need to be controlled.
Therefore, OLS and negative binomial regressions were run for Number of Trips (Out of
Home and Work, Work and Home Trips) and Average Dwell Time (Out of Home and Work,
Work and Home activities) on the following dependent variables: Work Factors, Personal
Attributes, Household Factors, Geographical Context and ICT use. Results are presented in
124
Table 14 and Table 15. Because education was largely insignificant for duration of
activities, it was removed from the results. Likewise, income, which was largely insignificant for
number of trips taken, was also removed from the results.
The OLS models all have low Adjusted R-squared values. Low r-squared values are a
point of major contention amongst scholars of all fields. Some scholars argue that the R-squared
value is of little importance, as it is not an actual statistical test and should not be used as a
descriptive statistic (Cameron, 1993). Moreover, the purpose of regression modeling is not to
reach a high r-squared, but “rather to obtain dependable estimates of the true population
regression coefficients and draw statistical inferences about them” (Gujarati, 2004, p. 222).
Modeling behaviors tend to fall constantly into the problem of regressions, with resulting
low r-squared values. This is because behavior is highly variable and the larger a sample size, the
more variation introduced and the more explanatory variables needed to explain the variation.
Accounting for all possible explanatory variables becomes difficult and it is usually impossible
to know how much variation in the dependent variable is systematic.
Modeling travel behaviors from the NHTS data are likely to produce low r-squared
values, because the behavior surveyed was over a one-day period in the individual’s life and all
the various variables that could account for such travel are not present in the survey. The sample
size used is also quite large (greater than 50,000). Therefore, although the models reveal low r-
squared values, the significance of the coefficients on the variables included provides valuable
insight as to whether or not workplace factors are correlated with constraints holding all other
theoretical variables constant.
125
Work Relationships
Controlling for personal attributes, household factors, geographical context and ICT use,
the coefficients on flexible workplace options are mainly statistically significant in terms of
dwell time and number of trips with a few exceptions.
The option to work at home is not significantly correlated to the number of trips one
makes throughout the day, regardless of the type of trip. That is, whether the trip is a work trip, a
trip back to the home, or all other trips combined. Individuals with the option to work at home do
not make more or less trips than individuals who don’t have the option. The regression results
do show that individuals who have the option to work at home engage in activities of shorter
duration. Compared to individuals who don’t have the option, those with the option spend around
51 minutes less engaged in work related trips. They spend around 13 minutes less at the home
during the travel day, but only about half a minute less engaged in other activities such as
socializing or personal activities.
The extra time resulting from decreased activity time during the travel portion of the day,
namely in the form of work outside of the home, is being spent at home. Total time spent at
home is significant and positive for the ability to work at home and the only variable in this
model whose coefficient is significant. Those with the ability to work at home spend the most
amount of time at home overall (around 21 minutes more than individuals not allowed to work at
home and 13 minutes more than the self-employed).
126
Table 14:
Dwell Time (OLS): Gray color indicates coefficient is significant at the .05 level
Out of Home and Work Work Activities Total Home Time
B Beta Sig B Beta Sig B Beta Sig
(Constant) 65.48 0.000 505.69 0.000 13.74 0.00
Work Dummies
Option to Work at Home -0.55 -.003 0.002 -50.80 -.098 0.000 0.35 0.03 0.02
Self-Employed -1.13 -.005 0.044 -76.62 -.119 0.000 0.13 0.01 0.09
Flexible Start Times 0.43 .003 0.000 -36.08 -.096 0.000 0.01 0.00 0.78
Work and WAH Interactions
WAHbySelf-Employed 2.62 0.01 0.26 -35.01 -.033 .000 0.03 0.00 0.84
WAHbyFlextime 2.05 0.01 0.47 4.32 0.01 0.58 0.05 0.00 0.75
Gender
Female -4.34 -.034 .000 -0.41 0.00 0.83 -0.01 0.00 0.71
WAHbyFemale -2.16 -0.01 0.21 15.16 .019 .002 -0.03 0.00 0.73
Household
Household Size 1.66 .033 .000 4.02 .027 .000 0.02 0.01 0.32
Population of Census Tract 0.00 .013 .011 0.00 .012 .011 0.00 0.00 0.69
Household Vehicle Count 0.92 .017 .005 -2.34 -.015 .005 -0.02 -0.01 0.31
Infant One Adult -23.95 -.023 .000 -51.16 -.015 .001 0.03 0.00 0.94
Infant Two Adults -14.93 -.083 .000 -10.84 -.020 .003 -0.09 -0.01 0.23
Small Child One Adult -15.15 -.029 .000 -23.58 -.014 .002 -0.30 -0.01 0.06
Small Child Two Adults -11.25 -.072 .000 -7.97 -.017 .010 -0.08 -0.01 0.19
Teenage One Adult 2.90 0.00 0.43 1.92 0.00 0.85 -0.32 -0.01 0.11
Teenager Two Adults -0.84 0.00 0.53 0.41 0.00 0.91 -0.01 0.00 0.84
Income Dummies
<=10K 26.57 0.02 0.000 8.63 0.00 0.34 -0.13 0.00 0.46
10-20K 25.65 0.04 0.000 -12.55 -.011 .030 -0.04 0.00 0.73
20-30K 15.79 0.03 0.000 -2.03 0.00 0.64 0.13 0.01 0.14
30-40K 11.28 0.02 0.002 -6.35 -0.01 0.09 0.06 0.00 0.42
40-50K 10.28 0.02 0.004 -5.66 -0.01 0.10 0.00 0.00 0.96
50-60K 3.83 0.01 0.27 -5.44 -0.01 0.10 0.05 0.00 0.46
60-70K 13.68 0.03 0.00 -3.66 -0.01 0.26 0.01 0.00 0.86
70-80K 2.82 0.01 0.42 -5.09 -0.01 0.10 -0.06 0.00 0.34
80-100K 0.52 0.00 0.86 -5.57 -.010 .039 -0.02 0.00 0.75
Age -2.92 -.056 .000 -5.92 -.039 0.000 0.00 -0.01 0.28
Webuse -0.58 -0.01 0.08 -9.30 -.059 0.000 0.01 0.00 0.40
Rsquared 0.01
0.069
0.002
Adjusted Rsquared 0.01 0.068 0.001
127
Table 15:
Number of Trips (Negative Binomial): Gray color indicates coefficient is significant at the .05 level
Out of Home and Work Work Activities Home Activities
B Exp(B) Sig B Exp(B) Sig B Exp(B) Sig
(Constant) 0.22 1.25 0.00 -0.04 0.96 0.39 -0.98 0.37 0.00
Work Dummies
Option to Work at Home 0.06 1.06 0.20 0.06 1.07 0.21 -0.04 0.96 0.52
Self-Employed 0.10 1.10 0.00 0.22 1.25 0.00 0.09 1.09 0.01
Flexible Start Times 0.06 1.06 0.00 0.16 1.17 0.00 -0.05 0.95 0.01
Work and WAH Interactions
WAHbySelf-Employed 0.09 1.10 0.02 0.01 1.01 0.90 0.07 1.07 0.20
WAHbyFlextime 0.06 1.06 0.22 -0.10 0.91 0.07 0.14 1.15 0.03
Gender
Female 0.22 1.24 0.00 -0.12 0.89 0.00 -0.01 0.99 0.38
WAHbyFemale -0.03 0.97 0.30 -0.02 0.98 0.53 0.03 1.03 0.48
Household Characteristics
Household Size 0.00 1.00 0.68 -0.02 0.98 0.03 0.05 1.05 0.00
Population of Census Tract 0.00 1.00 0.12 0.00 1.00 0.08 0.00 1.00 0.00
Household Vehicle Count -0.02 0.98 0.00 0.02 1.02 0.00 -0.01 0.99 0.11
Infant One Adult 0.65 1.91 0.00 0.09 1.09 0.42 0.02 1.02 0.89
Infant Two Adults 0.22 1.25 0.00 0.00 1.00 0.93 0.00 1.00 0.88
Small Child One Adult 0.45 1.57 0.00 0.00 1.00 0.93 0.45 1.56 0.00
Small Child Two Adults 0.25 1.29 0.00 0.00 1.00 0.95 0.26 1.30 0.00
Teenage One Adult 0.19 1.21 0.00 -0.01 0.99 0.87 0.31 1.36 0.00
Teenager Two Adults 0.07 1.08 0.00 -0.03 0.97 0.20 0.14 1.15 0.00
Education Dummies
Less than HS -0.25 0.78 0.00 0.04 1.04 0.31 -0.21 0.81 0.00
HS Degree -0.20 0.82 0.00 0.00 1.00 0.84 -0.21 0.81 0.00
Some College -0.10 0.91 0.00 0.03 1.03 0.12 -0.15 0.86 0.00
Bachelors -0.04 0.96 0.01 0.02 1.02 0.20 -0.08 0.92 0.00
Age 0.02 1.02 0.00 0.02 1.02 0.00 0.02 1.02 0.02
Webuse 0.03 1.03 0.00 0.03 1.03 0.00 0.03 1.03 0.00
Dispersion Coefficient 1
a
1
a
1
a
Log Likelihood -96630
-84012
-53320
Likelihood Ratio Chi-Square 1470.17 0.00 690.08 0.00 678.36 0.00
a. Fixed at the displayed value.
128
Behavior of the self-employed is slightly different. Here the regression results
demonstrate that the self-employed make more trips than individuals who are not self-employed,
regardless of activity (coefficients on both work trips and non-work or home trips are positively
significant). The self-employed are making over one more work-related trip than the non-self-
employed and around one more non-work or home trip, as well as one more trip back to the
home during the day.
The self-employed are also engaging in activities of shorter duration, including work
activities, but not activities in the home. They are spending around 77 minutes less for work
related trips than those who are not self-employed, although they are not spending more time at
home during the travel day (home dwell time is not significant).
Those with the ability to adjust work start time make more trips than individuals who
cannot adjust their start time, with the exception of trips back to the home (flexible start time is
significant across all the trip models). Those with flexible start time engage in activities of
shorter duration when it comes to work and the home (36 minutes less spent on work trips and 17
minutes less on trips back to the home). Those with flexible work start time spend a little more
time engaged in non-work and home activities (half a minute more). Although the coefficients on
travel time are significant for work and home activities, they are small indicating a negligible
effect.
These results indicate that controlling for household and personal factors, flexible
workplace options, reduce both coupling and authority constraints resulting in increased
fragmentation of activities, particularly with work. Interestingly, if an individual is both self-
employed and has the ability to work at home, her work dwell time is further reduced by another
129
35 minutes and she makes an additional trip outside of home and work. Individuals who are self-
employed and have the ability to work at home, express the largest amount of authority over the
way they run their day. These individuals are likely in service or contracting professions and
while the combination does not result in more work-related trips, the combination results in more
out of work and home trips and shorter dwell times for work-related activities.
Gender
Women make more trips outside of work and home than men. This is likely due to the
fact that women partake in more childcare and household related activities than men (Schwanen,
Kwan, & Ren, 2008). The coefficient on the female dummy is significant and positive for trips
outside of the home and work and negative for work-related trips. Women in general spend less
time engaged in activities than men, both in trips to the home during the day and in all other
activities, with the exception of work. Women are spending less time traveling for activities than
men in terms of work and non-work activities. Females who have the option to work at home
spend more time engaged in work activities, than males who have the option to work at home.
Household Characteristics
Whether or not there is a child present in the household, the age of the child influences
the number of trips an individual makes. Households with children make more trips outside of
home and work than households without children. Single parents with small children make the
most amounts of these types of trips, followed by two-parent households with small children.
Presence of children in the household is not significantly correlated with work trips, however, if
the child in the household is a toddler (not an infant), to a teenager, individuals make more trips
back to the home during the travel day.
130
Individuals in households with small children (infants and toddlers) engage in activities
of shorter duration, both at work and outside of work, although, they spend the same amount of
time in the house during the travel day and before and after the travel day, than households
without children.
Income
Income is generally insignificant across the models. With regards to number of trips
made, income is insignificant for all types of trips, except for trips made back to the home during
the day. All income categories (except 10 to 20K) make more trips back to the home, than high
earning individuals. In terms of dwell time, income is only significant for non-home and work
trips. Individuals earning under 70K engage in activities of longer duration, than those earning
over 70K.
The results for income negate evidence found in previous studies, showing that higher
earners fragment more (make more trips) (Hjorthol & Gripsrud, 2009; Alexander, Ettema, &
Dijst, 2010; Lenz & Nobis, 2007).
Education, Age and Web Use
With the exception of work trips, individuals with a graduate degree or higher, make
more trips than less educated groups. Duration of activities is not significantly correlated with
education, with the exception that the higher educated populations spends less time in work
related activities than those with a high school degree or less.
Age is positively correlated and significant with the number of trips one makes for work,
back to the home and for all other activities. Similarly, it is negatively correlated with dwell time
131
for these same types of activities. These results may indicate that authority constraints play a
large role in travel and fragmentation throughout the day. The significant and positive correlation
on the coefficient of age is contrary to the notion that younger generations are also more likely to
fragment (Lenz & Nobis, 2007; Alexander, Ettema, & Dijst, 2010; Hjorthol & Gripsrud, 2009),
as activities become more fixed in time the older one becomes (Schwanen, Pokwan, & Ren,
2008). Younger generations may be desiring to fragment and carry out activities throughout the
day differently, yet, may still be facing higher levels of authority constraints than older workers.
As with age, webuse is positively and significantly correlated with the number of trips a
person makes for work, back to the home and for all other activities. The more one uses the web,
the shorter the duration of their-out of home and work and activities related to work. This shows
that more ICT use may decrease capability constraints, encouraging a person to seek out more
activities, spending less time engaged in the activities.
Conclusion Paper 2:
This paper sets to explore how the workplace affects one’s level of flexibility to engage
in activities. It hypothesizes that individuals with flexible workplace options experience more
flexibility and fragmentation than folks without flexible workplace options, as certain coupling
and authority constraints are attenuated.
Although we are all likely to fragment more with the current state of technology, while at
a physical place like work or home, we can e-bank, e-shop and e-socialize, The reduction of the
physical constraints of the workplace are likely to further fragment activities traveled to and
engaged in, as travel and activities are not forced to fit around the workplace and the workday.
Given the data, this paper was not able to test for the first types of activities (those that involve
132
multitasking, or interruption without travel). It examined the latter form of fragmentation,
measuring the number of work and non-work and home trips and the time spent engaging in
these behaviors.
Apart from the obvious factors that affect travel, such as gender, household and
demographic constraints, my results reveal that the workplace is an important aspect to consider.
The ability to work at home, adjust working start time and self-employment lead to the reduction
of the coupling and authority constraints associated with a traditional workplace. Individuals
with these options are not bound to ‘join’ others and materials in the traditional bundles of the
workplace. The relaxation of coupling constraints with respect to the workplace, frees up an
individual’s physical space and the spacing of her activities. She can therefore seek out more
activities throughout the day and at times of her choosing. Likewise, flexible workplace practices
reduce authority constraints associated with the physical workplace. Because an individual is not
being watched or managed in a physical realm, she has more authority as to how her time can be
spent. The relaxation of these constraints should result in travel patterns and activity
consumption that is more fragmented, compared to individuals who don’t have flexible
workplace options. Those with flexible workplace options are likely to engage in more activities
of shorter durations staggered differently throughout the day.
The descriptives show that all flexible workplace groups make more trips (work, home
and non-work related), than groups who don’t have these options. They also engage in activities
outside of the home for shorter durations. When further dissected, the shorter durations mainly
come in the form of work activities and this ‘saved’ time is spent in more time at home and more
trips for non-work activities.
133
Activities are also accessed differently for groups that have flexible workplace options.
Not only are more trips made for personal, shopping, medical and other activities, they are made
more evenly throughout the day. Individuals who don’t have flexible workplace options group
these type of activities before and after traditional working hours, whereas, groups with flexible
workplace options are not as prone to this grouping.
The regression- analysis controls for likely variables that should influence travel, such as:
gender, household factors, income, education, and age. Indeed, the results demonstrate that
women make more trips out of home and work than men, reflecting higher household demands.
They spend less time engaged in activities as well. Households with children make more trips out
of home and work and spend less time engaged in activities. Older individuals (who are likely
less bound by authority constraints), make more trips and of shorter duration, as well as the more
highly educated. Income does not really play a role in explaining fragmentation in this context,
though. The results also indicate that individuals who use the web more are experiencing
diminished capability constraints; that is, ICTs allow for more possibilities to seek out activities,
services and social experiences. Higher levels of web use lead to more and shorter duration
activities out of the home and work.
Controlling for the above factors, workplace considerations do matter when it comes to
fragmentation. Also, how the fragmentation is expressed, differs by the three forms of workplace
flexibility addressed here. Those who have the ability to work at home do not make more trips
per day, than the rest of the working population, yet, they engage in activities of shorter duration,
particularly work, outside of the home. This extra time is spent at home, before and after the
travel portion of the day. The self-employed are making more trips than the rest of the
134
population, and they also engage in activities of shorter duration, particularly in work activities
outside of the home. They use this extra time for making more trips for non-home-and-work
activities. Those with flexible start time follow similar patterns to the self-employed, though, to a
less extent.
These results indicate that those with the option to work at home are bound slightly more
to the home than the self-employed and those with flexible start time, fragmenting work episodes
outside of the home more, but spending more time at home. The self-employed and those with
flexible start time fragment all activities more, but are less bound to the home, engaging in more
trips of all types throughout the day.
When someone is both self-employed and has the ability to work at home, they
experience even shorter times engaged in activities outside of the home and make more of these
types of trips. Those who are self-employed and have the ability to work at home express the
highest level of authority over their work styles and time, indicating that the workplace still
orients trips, but to a lesser extent; time.
One can argue that the above patterns (higher levels of fragmentation) observed through
workplace factors may have more to do with one’s type of work more, so than the decreased
coupling and authority constraints un-fixed workplaces engender. For example, the self-
employed may make more work-related trips, because they are more heavily employed in
services, such as sales and real-estate, in which many trips need to be made to clients and so
forth. Although this may be true, the nature of one’s work still does not explain why, for
instance, the self-employed make more trips for non-work related purposes. The authority and
coupling constraints still seem to explain these behaviors.
135
Furthermore, it is not clear, whether, or not, the self-employed are those more present in
sales and service occupations. According to a Bureau of Labor report on self-employment in
2009, the self-employed are mainly in industries of agriculture, construction, small business
services such as retail and specialized design and consulting industries (Hipple, 2010). Also, the
self-employed are represented in sales and related services (around 7%), but much more in
management (11%) and construction (11%). This is an avenue that merits further research and
this study was limited on this front as occupational and industry data are not available in the
dataset.
Discussion Paper 2
Planners have been interested in how those with flexible workplace options (namely the
ability to telework) travel in general (see Mokhtarian, Handy and Salomon 1995, Mokhtarian
2004, Pendyala, Goulias and Kitamura 1991, Henderson and Mokhtarian 1996, Choo,
Mokhtarian and Salomon 2005, Ory and Mokhtarian 2006). Because the constraints of
teleworkers are lessened, this research has examined if this group of people travel more in
general and/or live further away from urban centers experiencing longer commutes. In general
the studies have found that telecommuters exhibit reductions in overall VMT and PMT. Planners
have also been interested in how flexible workplace practices can alleviate peak-hour travel.
Evidence in this paper shows that people with flexible workplace options (the type of
option matters and the literature to date has not dealt with this) spend less time engaged in
activities out of the home which frees up time for more trips and thus more activities
(fragmentation) outside of the home. They engage in these activities at different times of the day
than the general working population, likely adjusting their travel to avoid congested periods of.
136
These findings promote the assumption and the argument for utilizing flexible workplace
practices as a strategy to alleviate peak-hour travel.
Those with flexible workplace options engage in more activities throughout the day than
the general working population, which has important implications for neighborhood community
development and neighborhood transportation. Workplace flexibility increases activities sought
throughout the day making neighborhood businesses more productive and active. There is also
preliminary evidence that those with flexible workplace options utilize more varied modes of
transportation than the general working population. That is, they make more walking, biking and
transit trips and use more of these options in unison throughout the day than the general working
population. This increased use of modes is likely a result of greater flexibility of time.
Encouraging flexible workplace practices can influence neighborhood vitality. A major
component of economic success within a neighborhood has to do with the employment
population in a neighborhood that activates a community during the day. It is necessary, though,
to understand the types and frequencies of activities consumed by the flexible workforce so that
planning that can accommodate these businesses at the neighborhood level. Since flexibility in
the workplace engenders more trips throughout the day and via different modes this has
important consequences on travel planning especially since flexible workplace practices and
modified work days are likely to increase. Transportation planning, in this context, should focus
on provision throughout the day at the neighborhood level, between neighborhoods and in a
variety that accommodates and connects the different modes.
Understanding how the workplace fits into the scheme of daily travel is an important
undertaking, particularly the study of flexibility in the workplace. We are already beginning to
see different models of work in the form of more contract, distributed, part-time and self-
137
employed work and are slowly beginning to see more work at home. As cities face burgeoning
congestion, more expensive housing and as work becomes ever more globally connected,
information communication technologies will be increasingly used to adjust to urban and modern
living and it behooves planners to understand how these changes will play out in our urban
landscapes.
Much more work can be done on this front. Firstly, travel distances and how they relate
to types of activities and the home should be explored alongside the types of modes accessed by
flexible workplace individuals. Other types of workplace practices, such as part-time work and
the unemployed can be investigated. This study only looked at weekday travel, however, the line
between weekday and weekends is blurring and understanding how work and activities are
occurring on the weekend is as important. It would be useful to investigate how the younger
‘flexers’ are behaving, compared to older populations, but as of now, more in-depth data do not
exist. Because behaviors are likely to express themselves differently across different spaces, an
examination of place is necessary. Finally, the interactions between gender and workplace
flexibility and the resulting travel patterns is an avenue that merits further consideration. Here,
preliminary results suggest that women’s travel patterns are arranged differently than men’s and
female ‘flexers’ have the most distinct travel patterns, but women are still not offered flexible
workplace options as much as men.
There are new and exciting data out there in terms of innovative media platforms, such as
Twitter and Facebook, as well as other passive data collection occurring through cell-phones and
mapping applications that hold much promise in the investigation of fragmentation and activity
fixity and the resulting travel patterns.
138
References Paper 2
Alexander, B., Ettema, D., & Dijst, M. (2010). Fragmentation of work activity as a multi-
dimensional construct and its association with ICT, employement and sociodemographic
characteristics. Journal of Transport Geography , 18, 55.
Asensio, J. (2002). Transport mode choice by commuters to Barcelona's CBD. Urban Studies ,
39, 1881-1895.
Bagley, M., & Mokhtarian, P. (1997). Analyzing the preference for non-exclusive forms of
telecommuting: Modeling and policy implications. Transportation , 24, 203-226.
Bailey, D., & Kurland, N. (2002). A review of telework research: Findings, new directions, and
lessons for the study of modern work. Journal of Organizational Behavior , 23, 383-400.
Balkin, D., & Gomez-Mejia, L. (1987). Toward a contingency theory of compensation strategy.
Strategic Management Journal , 8, 169-182.
Bernardino, A., Ben-Akiva, M., & Salomon, I. (1993). A stated preference approach to modeling
the adoption of telecommuting. Transportation Reseach Record: Journal of the Transportation
Research Board, No. 1413 , 22-30.
Bonsall, P. (2004). Traveller behavior: Decision making in and unpredictable world. Journal of
Intelligent Transport Systems , 8 (1), 45-60.
Bonsall, P., & Shires, J. (2006). Employer expectations for commuting and business-related
travel in an environment rich in information and communication technologies. Transportation
Research Record: Journal of the Transportation Research Board, No. 1977 , 268-276.
Brannen, J. (2005). Time and negotiation of work-family boundaries. Time and Society , 14, 113-
131.
Breedveld, K. (1998). The double myth of flexibilization: trends in scattered work hours, and
differences in time-sovereignty. Time and Society , 7, 129-143.
Cameron, S. (1993). Why is the R squared adjusted reported? Journal of Quantitative Economics
, 183-186.
Castells, M. (2003). The Internet Galaxy: Refections on the Internet, Business and Society.
Oxford: Oxford University Press.
Chatman, J., & Jehn, K. (1994). Assessing the relationship between industry characteristics and
organizational culture: How different can you be? The Academy of Management Journal , 37,
522-553.
139
Choo, S., Mokhtarian, P., & Salomon, I. (2005). Does telecommuting reduce vehicle-miles
traveled? An aggregate time series analysis for the US. Transportation, 32, 37-64.
Couclelis, H. (2003). Housing and the new geography of accessibility in the information age.
Open House International , 28, 7-13.
Covin, J., Slevin, D., & Covin, T. (1990). Content and preformance of growht-seeking strategies:
A comparison of small firms in high and low-technology industries. Journal of Business
Venturing , 5, 391-412.
Crampton, S., & Hodge, J. (2009). Generation Y: Unchartered territory. Journal of Business and
Economic Research , 7.
Cullen, I., & Godson, V. (1975). Urban networks: The structure of activity patterns. Progress in
Planning , 4, 1-96.
Cullen, I., Godson, V., & Mayor, S. (1972). The structure of activity patterns. In A. Wilson,
Patterns and Processes in Urban and Regional Systems (p. 286). London: Pion.
DiMaggio, P., & Powell, W. (1983). The iron cage revisited: Institutional isomporhism and
collective reationality in organizational fields. American Sociological Review , 48, 147-160.
Drucker, J., & Khattak, A. (2000). Propensity to work from home: Modeling results from the
1995 Nationwide Personal Transportation Survey. Transportation Research Record: Journal of
the Transportation Research Board, No. 1706 , 108-117.
Ellen, I., & Hempstead, K. (2002). Telecommuting and the demand for urban living: a
preliminary look at white-collar workers. Urban Studies , 39, 749.
Farag, S., Dijst, M., & Lazendorf, M. (2003). Exploring the use of e-shopping and its impact on
personal travel behavior in the Netherlands. Transportation Research Record 1858 , Paper No.
03-3058, 47-58.
Faulconbridge, J., & Beaverstock, J. (2008). Geographies of international business travel in the
professional service economy. In D. Hislop, Mobility and technology in the workplace (pp. 87-
102). Abingdon: Routledge.
Florida, R., Gulden, T., & Mellander, C. (2088). The rise of the mega-region. Cambridge
Journal of Regions, Economy and Society , 1, 459-76.
Frank, L. (2000). Land use and transportation interaction: Implications on public health and
quality of life. Journal of Planning Education and Research , 20 (6).
140
Frank, L., Bradley, M., Kavage, S., Chapman, J., & Lawton, T. (2008). Urban form, travel time,
and cost relationships with tour complexity and mode choice. Transportation , 35, 37-54.
Gao, S., Mokhtarian, P., & Johnston, R. (2008). Exploring connections among job accessibility,
employment, income and auto ownership using structural equation modeling. Annals of Regional
Science , 42, 341-356.
Garhammer, M. (1995). Changes in working hours in Germany: The resulting impact of daily
life. Time and Society , 4, 167-203.
Giuliano, G. (2003). Travel, location and race/ethnicity. Transportation Research Part A: Policy
and Practice , 37, 351-372.
Gujarati, D. (2004). Basic Econometrics 4th Edition. The McGraw-Hill Companies.
Handy, S. (2005). Smart growth and the transportation land-use connection: What does the
research tell us? International Regional Science Review , 28, 146-167.
Handy, S., & Mokhtarian, P. (1995). Planning for telecommuting: Measurement and policy
issues. Journal of the American Planning Association , 61, 99-111.
Handy, S., & Mokhtarian, P. (1996). The future of telecommuting. Futures , 28, 227-240.
Handy, S., & Mokhtarian, P. (1996). The future of telecommuting. Futures , 28 (3), 227-240.
Hanson, S., & Pratt, G. (1995). Gender, work and space. London: Routledge.
Haynes, B. (2011). The impact of generational differences on the workplace. Journal of
Corporate Real Estate , 13 (2), 98.
Henderson, D., & Mokhtarian, P. (1996). Impacts of center-based telecommuting on travel and
emissions: Analysis of the Puget Sound Demonstration Project. Transportation Research Part D,
1 (1), 29-45.
Hipple, S. (2010). Self-Employment in the United States. Monthly Labor Review. Bureau of
Labor Statistics.
Hjorthol, R. (2002). The relation between daily travel and use of the home computer.
Transportation Research PArt A , 36, 437-452.
Hjorthol, R., & Gripsrud, M. (2009). Home as a communication hub: The domestic use of ICT.
Journal of Transport Geography , 17, 115-123.
Holtshouse, D. (2010). Knowledge work 2020: Thinking ahead about knowledge work. On the
Horizon , 18 (3), 193-203.
141
Hubers, C., Dijst, M., & Schwanen, T. (2008). ICTs coping strategies and gender differences in
the temporal and spatial fragmentation of paid labor. In: Paper presented at the Fourth
International Specialist Meeting on ICT, Everyday Life and Urban Change, Bristol, UK.
Janelle, D. (1986). Metropolitan expanision and the communications-transportation tradeoff. In
S. Hanson, The geography of urban transportation (pp. 357-385). New York: The Guilford
Press.
Janelle, D., & Hodge, D. (2000). Information, Place and Cyberspace Issues in Accessibility.
Berlin: Springer.
Jang, T. (2005). Cout data models for trip generation. Journal of transportation engineering ,
131, 444-450.
Karnowski, S., & White, B. (2002). The role of facility managers in the diffusion of
organizational telecommuting. Environment and Behavior , 34, 322-24.
Kenyon, S., & Lyons, G. (2007). Introducing multitasking to the study of travel and ICT:
Examining its extent and assessing its potential importance. Transportation Research A , 41,
161.
Khandwalla, P. (1976). Some top management styles, their context and performance.
Organization and Administrative Sciences , 7 (4), 21-51.
Kitamura, R., Mokhtarian, P., & Laidet, L. (1997). A micro-analysis of land use and travel in
five neighborhoods in the San Francisco Bay Area. Transportation , 24, 125-158.
Koskela, H. (1999). "Gendered exclusions':Women's fear of violence and changing relations to
space. Geografiska Annaler B, Human Geography , 81, 111.
Kwan, M.-P. (2007). Mobile communications, social networks, and urban travel: New topoligies
of spatial interaction. The Professional Geographer , 59 (4), 434-446.
Learner, E., & Storper, M. (2001). The economic geography of the internet age. Journal of
International Business Studies , 32, 641-65.
Lenz, B., & Nobis, C. (2007). The changing allocation of activities in space and time by the use
of ICT- "Fragmentation" as a new concpet and empirical results. Transportation Research Part A
, 41 (2), 190-204.
Leung, A. (2003). Different ties for different needs- recruitment practices of entrepreneurial
firms at different development phases. Human Resources Management , 42 (4), 303-320.
142
Line, T., Jain, J., & Lyons, G. (2011). The role of ICTs in everyday mobile lives. Journal of
Transport Geography , 19, 1490-1499.
Littlejohn, S., & Foss, K. (2005). Theories of human communciation 7th ed. Belmont, CA:
Wadsworth.
Lyons, G., & Urry, J. (2005). Travel time use in the Information Age. Transportation Research A
, 39, 257-276.
Madge, C., & O'Connor, H. (2006). Parenting gone wired: empowerment of new mothers on the
Internet? Social and Cultural Geography , 7 (2), 199-220.
Mahmassani, H., Yen, J., Herman, R., & Sullivan, M. (1993). Employee attitudes and stated
preferences towards telecommuting: an exploratory analysis. Transportation Research Record:
Journal of the Transportation Research Board, No. 1413 , 31-44.
Mannering, J., & Mokhtarian, P. (1995). Modeling the choice of telecommuting frequency in
California: An exploratory analysis. Technological Forecasting and Social Change , 49, 49-73.
Martins, L., Gibson, L., & Maynard, M. (2004). Vitual teams: What do we know and where do
we go from here? Journal of Management , 30, 805-35.
Mayo, M., Pastor, J., Gomez-Mejia, L., & Cruz, C. (2009). Why some firms adopt
telecommuting while others do not: A contingency perpective. Human Resource Management ,
48 (6), 917-939.
McQuail, D. (2000). Mass communication theory. London: Sage.
Mehrabian, A. (1981). Silent messages: Implicit communication of emotions and attitudes.
Belmont, CA: Wadsworth.
Miller, H., & Shaw, S.-L. (2001). Geographic Information Systems for Transportation:
Principles and Applications. New York: Oxford University Press.
Mokhtarian, P. (2004). Telecommuting, residential location, and commute-distance travelled:
evidence from State of California employees. Environment and Planning A , 36, 1877-1897.
Mokhtarian, P., & Salomon, I. (1996). Modeling the choice of telecommuting: 3. Identifying the
choice set and estimating binary choice models for technology-based alternatives. Environment
and Planning A , 28, 1877-1894.
Mokhtarian, P., & Salomon, I. (1994). Modeling the choice of telecommuting: Setting the
context. Environment and Planning A , 28, 1859-1876.
143
Mokhtarian, P., & Salomon, I. (1994). Modeling the choice of telecommuting: Setting the
context. Environmental Planning A , 26, 749-766.
Mokhtarian, P., & Salomon, I. (1997). Modeling the desire to telecommute: the importance of
attitudinal factors in behavorial models. Transportation Research A. , 30 (1), 35-50.
Mokhtarian, P., Handy, S., & Salomon, I. (1995). Methodological issues in the estimation of
travel, energy and air quality impacts of telecommuting. Transportation Research A , 29, 283-
302.
Mokhtarian, P., Salomon, I., Saxena, S., Sampath, S., Cheung, P., Le, K., et al. (1996). Adoption
of Telecommuting in Two California State Agencies. Research Report UCD-ITS-RR-96-5,
Institute of Transportation Studies, University of California, Davis, September.
Moon, N. (2007). Private sector telework and its implications for economic development:
Results of a case survey and policy assessment. Atlanta, Georgia: Enterprise Innovation Institute
and School of Public Policy's Science, Technlology, and Innovation Policy (STIP) Program
Georgia Institute of Technology.
Newman, D. (2012, May 22). 6 Technology trends transforming the workplace. Retrieved June
28, 2013, from millenialceo.com: http://millennialceo.com/leadership/6-technology-trends-
transforming-workplace/
Odih, P. (1999). Gendered time in the age of deconstruction. Time and Society , 8 (1), 9-38.
Offstein, E., Morwick, J., & Koskinen, L. (2010). Making telework work: Leading people and
leveraging technology for competitive advantage. Strategic HR Review , 9 (2), 32-37.
Olson, G., & Olson, J. (2000). Distance matters. Human Computer Interaction , 15, 130-79.
Olszewski, P., & Mokhtarain, P. (1994). Telecommuting frequency and impacts for State of
California employees. Technological forecasting and social change , 45, 275-286.
Ory, D., & Mokhtarian, P. (2006). Which came first, the telecommuting or the residential
relocation? An empirical analysis of causality. Urban Geography, 27 (7), 590-609.
Pain, R. (1997). Social geogrphies of women's fear. Transactions of the Institute of British
Geographers , 22, 231.
Pendyala, R., Goulias, K., & Kitamura, R. (1991). Impact of telecommuting on spatial and
temporal patterns of household travel. Transportation, 18 (4), 383-409.
Pennings, J., & Gresov, C. (1986). Technological networking and innovation implementation.
Organization Science , 3, 317-334.
144
Peters, P., & Heusinkveld, S. (2010). Institutional explanations for managers' attitudes towards
telehomeworking. Human Relations , 63, 107.
Peters, P., Tijdens, K., & Wetzels, C. (2004). Employees' opportunities, preferences, and
practices in telecommuting adoption. Information and Management , 41 (2), 469-482.
Piccoli, G., & Ives, B. (2003). Trust and the unintended effects of behavior control in virtual
teams. MIS Quarterly , 27, 365-95.
Pickup, L. (1988). Hard to get around: a study of women's travel mobility. In J. Little, L. Peake,
& P. Richardson, Gender and the Urban Environment (pp. 98-116). New York: New York
University Press.
Pinjari, A., Pendyala, R., Bhat, C., & Waddell, P. (2007). Modeling residential sorting effects to
understand the impact of the built environment on commute mode choice. Transportation , 34,
557-573.
Popuri, Y., & Bhat, C. (2003). On modeling choice and frequency of home-based
telecommuting. Transportation Research Record: Journal of the Transportation Research
Board, No. 1858 , 55-60.
Rheingold, H. (2003). Smart Mobs: The Next Social Revolution. New York: The Perseus Book
Group.
Rhoads, M. (2010). Face-to-face and computer-mediated communication: What does theory tell
us and what have we learned so far? Journal of Planning Literature , 25 (2), 111-122.
Sánchez, A., Pérez, M., de Luis Carnicer, P., & Jiménez, M. (2007). Teleworking and workplace
felxibility: A study of impact on firm performance. Personnl Review , 36 (1), 42-64.
Schrank, D., Lomax, T., & Eisele, B. (2011). 2011 Urban Mobility Report. College Station, TX:
Texas Transportation Institute.
Schwanen, T. (2007). Matter(s) of interest: Artefacts, spacing, timing. GeografiskaAnnaler,
Series B: Human Geography , 89 (1), 9-22.
Schwanen, T. (2006). On 'arriving on time', but what is 'on time'? Geoforum , 37 (6), 882-894.
Schwanen, T., & Dijst, M. (2003). Time windows in workers' activity patterns: Empirical
Evidence from the Netherlands. Tranpsortation , 30 (3), 261-283.
Schwanen, T., & Kwan, M.-P. (2008). The Internet, mobile phone and space-time constraints.
Geoforum , 1362-1377.
145
Schwanen, T., & Mokhtarian, P. (2005). What affects commute choice: Neighborhood physical
structure or preferences towards neighborhoods? Journal of Transportation Geography , 13, 83-
99.
Schwanen, T., Dielema, F., & Dijst, M. (2004). Th impact of metropolitan structure on commute
behavior in the Nethwrlands: a multilevel approach. Growth Change , 35, 304-333.
Schwanen, T., Kwan, M.-P., & Ren, F. (2008). How fixed is fixed? Gendered rigidity of space-
time constraints and geographies of everyday activities. Geoforum , 39, 2109-2121.
Schwanen, T., Pokwan, M., & Ren, F. (2008). How fixed is fixed? Gendered rigidity of space-
time constraints and geographies of every day activities. Geoforum , 39, 2109-2121.
Sener, I., & Bhat, C. (2010). A copula basedsample selection model of telecommuting choice
and frequency. Environment and Planning A , 43 (1), 126-145.
Senju, A., & Johnson, M. (2009). The eye contact effect: Mechanisms and development. Trends
in cognitive sciences , 13, 127-34.
Sia, C., Teo, H., Tan, B., & Wei, K. (2004). Effects of environmental uncertainty on
organziational intention to adopt distributed work arrangements. IEEE Transactions on
Engineering Management , 51, 256-67.
Slattery, M. (2013, March 7). How consumer technology is remaking the workplace. Retrieved
June 28, 2013, from wired.com: http://www.wired.com/insights/2013/03/how-consumer-
technology-is-remaking-the-workplace/
Susilo, Y., & Maat, K. (2007). The influence of the built environment to the trends in commuting
journeys in the Netherlands. Transportation , 34, 589-609.
Swanberg, J., Pitt-Catsouphes, M., & Drescher-Burke, K. (2005). A question of justice:
disparities in employees' access to flexible schedule arrangements. Journal of Family Issues , 26,
866.
Tang, W., Mokhtarian, P., & Handy, S. (2008). The role of neighborhood characteristics in the
adoption and frequency of working at home: Empirical evidence from Northern California.
(Paper, UCD-ITS-RR-08-21). UC Davis: Institute of Transportation Studies.
Tomaskovic-Devey, D., & Risman, B. (1993). Telecommuting innovation and organization: a
contingency theory of labor process change. Social Science Quarterly , 74, 367-385.
Tulgan, B., & Martin, C. (2001, October 3rd). Managing generation Y- part 2. Business Week
Online. Retrieved June 13th, 2012, from http://www.businessweek.com/stories/2001-10-
03/managing-generation-y-part-2
146
Valentine, G. (1989). The geography of women's fear. Area , 21, 385.
Valentine, G., & Holloway, S. (2002). Cyberkids? Exploring children's identities and social
networks in on-line and off-line worlds. Annals of the Association of American Geographers , 92
(2), 302-319.
Vana, P., Bhat, C., & Mokhtarian, P. (2008). On modeling the choice of work-hour arrangement,
location and frequency of telecommuting.
Wajcman, J. (2010). Feminist theories of technology. Cambridge Journal of Economics , 34,
143-152.
Walls, M., Safirova, E., & Jiang, Y. (2007). What drives telecommuting? The relative impact of
worker demographics. employer characteristics, and job types. Transportation Research Record
Journal of the Transportation Research Board, No. 2010 , 111-120.
Wiggins, S. (1995). Entrpreneurial enterprises, endogenous ownership, and the limits to firm
size. Economic Inquiry , 33, 54-69.
Yeraguntla, A., & Bhat, C. (2005). A classification taxonomy and empirical analysis of work
arrangements. Transportation Research Record: Journal of the Transportation Research Board,
No. 1926 , 233-241.
Yu, H., & Shaw, S.-L. (2007). Revisiting Hägerstrand's Time-Geographic Framework for
Individual Activities in the Age of Instant Access. In H. Miller, Societies and Cities in the Age of
Instant Access (pp. 103-118). Dordrecht, Netherlands: Springer.
147
Appendix Paper 2
Description of Variables and Measurements
Household Income:
Survey Scale (1 to 18):
1= Less than $5K; 2=$5K-$9.9K; 3=$10K-$14.9K; 4=$15K-$19.9K; 5=$20K-$24.9K; 6=$25K-
$29.9K; 7=$30K-$34.9K; 8=$35K-$39.9K; 9=$40K-$44.9K; 10=$45K-$49.9K; 11=$50K-
$54.9K; 12=$55K-$59.9K; 13=$60K-$64.9K; 14=$65K-$69.9K; 15=$70K-$74.9K; 16=$75K-
$79.9K; 17=$80K-$99.9K; 18=Greater than or equal to $100K
Consolidated Scale (1 to 10):
1= Less than $10K; 2=$10K-$19.9K; 3=$20K-$29.9K; 4=$30K-$39.9K; 5=$40K-$49.9K;
6=$50K-$59.9K; 7=$60K-$69.9K; 8=$70K-$79.9K; 9=$80K-$99.9K; 10= Greater than or equal
to $100K
Education:
Survey Scale (1-5):
1=Less than High-school Degree; 2=High-school Degree; 3=Some College; 4=Bachelors;
5=Graduate
Age:
Survey Scale (5-88; 92):
Age runs on a continuous scale ranging from 5 to 88. A score of 92 indicates that the individual
is 89 years or older at the time of the survey.
Household Size:
Survey Scale (1-14)
Count of household members
Square mile Population Density
These variables reflect the resident’s census tract density
Survey Scale of Population:
148
50=0-99
300=100-499
750=500-999
1500=1,000-1,999
3000=2,000-3,999
7000=4,000-9,999
17000=10,000-24,999
30000=25,000-999,999
Webuse
Amount of webuse in one month
Survey Scale (1-5):
1= Daily usage; 2=Several times a week; 3=Once a week; 4=Once a month; 5= Never
Reversed Coded (1-5):
1=Never; 2=Once a month; 3=Once a week; 4=Several time a week; 5=Daily usage
On-line Purchases
Number of times something was purchased via the internet in the last month
Survey Scale (0-200)
Walking, Biking and Public Transit Trips
How many times the modes were accessed in the last month
Walking Trips Survey Scale (0-99)
Biking Trips Survey Scale (0-99)
Public Transit Trips Survey Scale (0-180)
Household Vehicle Count
Number of vehicles available in the household
Survey Scale (0-27)
Average Dwell Time
Calculated as an individual’s average time (in minutes) at their various destinations throughout
the day
149
Number of Trips
Calculated as the total number of trips an individual made during the survey day. These are
separated by different types of activities:
1. Home: Return home
2. Work: Go or return to work; Attend Meetings; Other work-related affairs
3. School/Daycare/Religious Activities
4. Medical/Dental Services
5. Shopping: Buying of goods or services such as dry-cleaning
6. Social/Recreational: Exercise; Play sports; Visit friends or family; Entertainment
7. Personal: Use of professional services such as an attorney; Grooming (hair and nail); Pet
care and walking; Attend PTA meetings or local government meetings
8. Transport Someone
9. Meals (including coffee)
10. Other
150
Conclusion
This work has added to the literature on flexible workplace practices by showing that
industry matters in the provision of flexible workplace practices and flexible workplace practices
matter in daily travel.
The literature review has showed inconsistent results with respect to what characterizes
the typical teleworker, or typical organization that allows for telework. This is mainly due to the
fact that most of the past studies have relied on small sample sizes, derived from single
institutions or a single region. Results rendered from one area or one institution are not
necessarily transferable to another. Methodological lacunae exist with both the literature
examining individuals and organizations. The literature focusing on the characteristics of
individual tele/non-tele-workers has not sufficiently examined how occupation, industry or the
organization plays a role in the provision of telework. These studies have used dissimilar and
overly generalized categories for occupation, when including the variable. The literature
focusing on organizations has not sufficiently examined how larger competitive forces play a
role in whether an organization provides flexible workplace options. Workplace flexibility has
been used in these studies as a way to explore how organizations respond to competition; yet the
‘competition’ and how this ‘competition’ might vary amongst firms in different industries has
not been brought into the organization-based models.
How flexible workplace users travel is an important consideration in understanding urban
landscapes, or when encouraging these practices to alleviate peak-hour commute congestion,
especially since these forms of work are expected to burgeon in our increasing digital
environments. Flexible workplace practices are inherently dependent upon the use of information
communication technologies (ICTs) and it is unclear as to whether the use of ICTs stimulates
151
more travel or substitutes for travel. The literature on this front is new and underdeveloped.
Studies employing the concepts of fragmentation and fixity have largely left out how ‘fixed’ or
‘unfixed’ one’s place of work interacts with how she travels.
This work addresses the above-mentioned issues by performing two large-scale (at the
level of the nation) analyses of industry and travel patterns of flexible workplace users.
The first quantitative paper analyzed proportions of individuals who have the option to
work at home and adjust their start times at the level of the region. This unit of analysis allows
for the inclusion of level of industry mix. No national dataset exists that ties an individual’s
industry or occupation to her flexible workplace options. Regional mix of industry is a relatively
good indication of how saturated regions are in terms of certain occupations and therefore
industry.
The results from the first paper reveal that, indeed, industry and levels of education (skill)
matter when it comes to how much flexible workplace is available in that region. Regions that
have higher proportions of high-tech and managerial occupations see more rates of telework
provision. The argument made is that the high-tech industry faces heavier competition in the
form of talented labor that can help organizations create the newest and greatest high tech
products (e.g. mobile operating systems, mobile applications or hybrid-battery technology, or
design). Higher concentrations of managers in a region also support this statement as the high-
tech industry spins-off many start-ups that are run and managed by the owners and creators.
Regions with large shares of high-tech occupations also house the most educated, as these
occupations require substantial skill.
The story of who actually engages in the flexible workplace is slightly different. In terms
of regions with higher proportions of individuals who work at home at least once a month, high-
152
tech industry mix ceases to be significant, but share of managerial occupations and higher
educated remain significant. In terms of the weighted average of work at home by region,
median gross rent and a region’s size is positively correlated with higher averages. These results
indicate that the provision of flexible workplace practices is correlated with industry, while those
who actually use flexible workplace practices is contingent on one’s status in the workplace and
the specific conditions of a person, such as status in the workplace, rents and education levels.
Because individuals with flexible workplace practices face different coupling and
authority constraints (the constraint of the workplace is removed) and their use of flexible
workplace practices is contingent upon a level of ICT use, these groups of individuals become an
interesting venue for studying fragmentation. Results from my second paper show that the
workplace is an important consideration in how individuals space time and allot their travel.
Those who have flexible workplace options intuitively partake in more episodes of work at
shorter durations. They use this extra time at home and in traveling (for those who have the
ability to work at home) and in other activities (the self-employed and those with the ability to
adjust work start times). Those who practice a combination of flexible workplace options make
even more work trips and general trips and engage in activities of shorter duration. Flexible
workers start their work trips later in the day and make more overall trips throughout the day,
than people without flexible workplace options.
Unearthing the larger (not simply localized) patterns of how flexible workplace practices
play out in urban environments and the reasons for these patterns can help planners and policy
makers understand how their regions are competing. High-tech organizations are undertaking
new management models and these organizations provide interesting case studies as to how
153
flexi-work is being incorporated to meet the needs of the workforce. Understanding flexible
workplace trends informs regional and neighborhood travel patterns.
How flexible workers travel and interface with their urban environments sheds light on
how ICTs are influencing urban landscapes. This research shows that those with flexible
workplace options are travelling differently than traditional workers accessing more activities
throughout the day and in using flexibility to avoid congested travel periods of the day. This has
important planning implications in understanding how flexible workplace options influence the
economic vitality of a neighborhood and how neighborhood and regional transportation should
be structured especially given that these work models are likely to increase.
Useful, quality and regular data have been one of the main impediments to research on
the above mentioned. Modern day ICTs have made possible a host of new data sources that
many companies are beginning to analyze and explore to understand consumer behavior, such as
social media platforms, ride share platforms, mapping applications and the like. These data
sources, while inevitably containing many quality issues and requiring sophisticated skills to
process, hold promise in our ability to ask new questions and enrich our existing knowledge of
how our digital environments are interweaving into our physical ones.
Abstract (if available)
Abstract
Flexible workplace options such as the ability to work remote from the workplace (telework) or to adjust working start times has interested scholars and policymakers for decades. In a transportation context, flexible workplace options are proposed as alternatives to peak-hour commuting as the commute trip can either be eliminated or adjusted. Transportation scholars and policymakers are also interested in how flexi-workers travel in general. Within human resource and management domains, flexible workplace options are of interest in understanding how organizations adjust to competitive climates. Flexible work arrangements allow organizations to hire talent from a wider geography or through offering their employees benefits in the form of flexi-work. Flexible workplace practices have also been proposed as tools to engender better work-life balance to fit the specific needs of households. ❧ Within the last decade renewed interest in flexible workplace practices has arisen due to the recent and rapid advances in information communication technologies (ICTs). ICTs have made remote work more possible than ever. The ubiquitous and cheaper availability of portable devices, cloud technologies and 4G networks allow for the portability and feasibility of one’s workplace virtually anywhere. Organizations have new labor demands for talent that has the necessary skills for the emerging digital environment. The global and networked nature of many business operations means that communication and collaboration is occurring not only within near proximity, but simultaneously around the world. Not only is a new language being spoken in the workplace, but the language is being spoken with a new set of tools that are continuously expanding, evolving and becoming obsolete, causing employers, employees and customers to communicate, transact, collaborate and coordinate in a host of virtual ways. ❧ Overall, the research on flexible workplace practices has not reached a consensus as to why people engage in flexible workplace practices and why organizations allow these practices. Studies considering why individuals adopt flexi-work have not comprehensively examined the occupations and industries of these individuals. Ultimately, one’s organization and type of work determine whether or not the individual can engage in flexi-work. Concomitantly, studies considering why organizations provide flexible workplace practices have not comprehensively considered the larger competitive forces under which the organizations operate. These studies seek to understand how organizations adjust to their competitive environments through flexible employment practices, yet have not examined the competitive variations under which different organizations operate. The literature on individuals and organizations has generally been kept at small scales of analysis at varying locations rendering results inconsistent and un-generalizable. ❧ Parallel to the study of why individuals adopt flexible workplace practices is the investigation of how those who adopt flexi-work, travel. Flexible workplace practices are inherently bound to ICT use in order to conduct remote work and are therefore a measure of ICT use. ICT use has the potential to substitute trips (e.g. replace a commute trip), compliment trips (allow one to modify their trip while en route), or incentivize trips (encourage more travel). ❧ Concepts such as fragmentation and fixity have been developed to understand such behaviors. ICTs can engender more travel by informing or requiring extra travel, or they can reduce travel by bringing more activities within the domain of the home. The literature on fragmentation and fixity has focused on individual aspects such as gender and socio-demographics, generally leaving out how one of the main pivots in an individual’s travel, the workplace, influences travel. It is still unclear as to how the use of ICTs affect travel and how people who partake in more fragmented work behaviors (e.g. work from home) as a result of ICTs travel in general. The research is new on this front and large-scale data are hard to come by. ❧ This work explores the above mentioned issues in three papers. The first paper conducts a literature review on flexible workplace practices. The literature review is followed by two quantitative papers: the first of which examines flexible workplace practices at the level of the region and the second, which examines the travel patterns of individuals who have flexible workplace options. ❧ A regional analysis of the nation has several benefits. Small-scale analyses only convey results with respect to the institution or area surveyed. A national analysis allows for larger patterns to be detected. Because there is no national dataset which simultaneously incorporates an individual’s flexible workplace options and her occupation, examining trends at the regional level allows for the consideration of industry and occupation. ❧ The second paper also uses a national dataset and looks at the travel patterns of individuals who have flexible workplace options such as self-employment status, ability to work at home and the ability to adjust working start times. These individuals serve as a group who can utilize ICTs to conduct remote work. Comparing these groups to groups who do not have flexible workplace options and measuring behavior such as number of trips they make and their duration of activities can shed light on how flexibility and ICT may be resulting in non-traditional travel patterns.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The demand for reliable travel: evidence from Los Angeles, and implications for public transit policy
PDF
The long-term impact of COVID-19 on commute, employment, housing, and environment in the post-pandemic era
PDF
Active travel, outdoor leisure, and neighborhood environment: path analysis, Los Angeles County
PDF
The built environment, tour complexity, and active travel
PDF
Children’s travel behavior in journeys to school
PDF
Healthy mobility: untangling the relationships between the built environment, travel behavior, and environmental health
PDF
Urban air pollution and environmental justice: three essays
PDF
Essays on congestion, agglomeration, and urban spatial structure
PDF
Spatial and temporal expenditure-pricing equity of rail transit fare policies
PDF
Environmental justice in real estate, public services, and policy
PDF
An examination of soft skills in the virtual workplace
PDF
Household mobility and neighborhood impacts
PDF
Black women in tech: examining experiences in tech industry workplaces
PDF
Three papers on dynamic externalities and regional growth
PDF
From a privilege to an option: hybrid work schedule: a gap analysis
PDF
How does collaborative governance work? The experience of collaborative community-building practices in Korea
PDF
Diverging pathways to citizenship and immigrant integration in the U.S.
PDF
Planning care with the patient in the room: a patient-focused approach to reducing heart failure readmissions
PDF
The role of social media for community-based organizations focused on environmental justice in southeast Los Angeles
PDF
A strategic talent management retention model: an effective way to shape the United States Space Force
Asset Metadata
Creator
Rhoads, Mohja Lynn
(author)
Core Title
The flexible workplace: regional tendencies and daily travel implications
School
School of Policy, Planning and Development
Degree
Doctor of Philosophy
Degree Program
Policy, Planning, and Development
Publication Date
01/29/2015
Defense Date
06/30/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
fixity,flexible start times,flexible work,flexible workplace,fragmentation,high tech industries,OAI-PMH Harvest,telecommuting,telework,virtual work,work at home
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Giuliano, Genevieve (
committee chair
), Boarnet, Marlon (
committee member
), Schweitzer, Lisa (
committee member
)
Creator Email
mlr40@columbia.edu,mrhoads@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-528426
Unique identifier
UC11298362
Identifier
etd-RhoadsMohj-3152.pdf (filename),usctheses-c3-528426 (legacy record id)
Legacy Identifier
etd-RhoadsMohj-3152.pdf
Dmrecord
528426
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Rhoads, Mohja Lynn
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
fixity
flexible start times
flexible work
flexible workplace
fragmentation
high tech industries
telecommuting
telework
virtual work
work at home