The fragmented worker? ICTs,coping strategies and gender differences in the temporal and spatial fragmentation of paid labour

Abstract

Information and Communication Technologies (ICTs) are often proclaimed to facilitate the fragmentation of activities, a process whereby a certain activity is divided into several smaller pieces, which are performed at different times and/or locations. This study analyzes two-day combined activity, travel and communication diaries collected among Dutch households and presents quantitative findings of the associations between ICTs and the spatiotemporal fragmentation of paid labour. Controlling for various coping strategies, employment and commute factors, household characteristics, lifestyle orientation, time personality and spatial context, statistically significant relations were found between ICTs and the spatiotemporal fragmentation of paid labour for both men and women. The fact that both positive and negative associations were found suggests that ICTs can be adopted to make use of opportunities to arrange paid labour in a flexible way, or as a compensation when such opportunities are lacking. The results also indicate that up to a certain degree these associations are gender specific.

Keywords

Spatiotemporal activity fragmentation information and communication technologies gender differences the Netherlands paid labour

Introduction

The reorganisation of human activities across space and time facilitated by Information and Communication Technologies (ICTs) has emerged as a key theme in the social sciences (Castells, 2000; Urry, 2004). Regarding paid work, for example, many scholars have hypothesized that the dominance of the traditional office as the main employment location and the 9-to-5 workday are in decline (Gershuny, 2000; Kaufman-Scarborough, 2006; Perrons, 2003). Instead, ICTs are expected to provide employees with more choice and flexibility in determining where and when to work. The type of spatio-temporal ‘reorganisation’ that has received most attention from scholars to date is teleworking from home, where employees perform part or all of their paid work from their own homes (Bailey and Kurland, 2002; Choo et al., 2005; Felstead and Jewson, 2000; Jackson and Van der Wielen, 1998; Ory and Mokhtarian, 2006; Tietze and Musson, 2002). A considerably smaller number of empirical studies have examined whether employees are also taking advantage of the fact that ICTs enable them to work from various other locations beyond the home and traditional office environment including teleworking while on the move (Axtell et al., 2008; Hislop and Axtell, 2007; Laurier, 2002; Letherby and Reynolds, 2003; Lyons et al., 2007).

More recently, research inspired by social constructivist approaches and Science and Technology Studies (STS) has underlined the importance of human agency in how ICTs come to be used and how this then might be of consequence for the spatial and temporal reorganisation of paid labour (Nansen et al., 2010; Wajcman et al., 2010). For Wajcman et al. (2010: 271) “[i]t is clear that some employees are not simply responding to the Internet in accordance with its technical capabilities to overcome temporal and spatial boundaries. Instead, they are actively making decisions about how they incorporate the technology into their lives in ways that are beneficial to them.” Similarly, Leonardi et al. (2010) found that employees working from home actively manage the connectivity afforded by ICTs by alternately using them to increase the distance between themselves and the office (for example, when they need to work on a specific task without being disturbed), and at other times to decrease this distance (for example, when consultation with colleagues at a different location was needed). They conclude that “[t]heories informing distributed work should consider not simply the material features inscribed in the design of technology, but also the motivations of individuals to enact a variety of social practices through technology use” (Leonardi et al., 2010: 99). The reorganisation of paid labour across space and time through the use of ICTs therefore depends on both individual motivations and circumstances, as well as technical capabilities.

As noted by Fawcett and Song (2009), although there seems to be general consensus that ICTs are likely to be related to a reorganisation of paid labour across space and time, to date little quantitative empirical evidence has been provided for this notion. Furthermore, they point out that most quantitative studies focus on aggregate time budget analyses of total working hours, rather than “day-to-day time management” (Fawcett and Song, 2009: 313). This paper aims to fill these gaps by presenting quantitative findings of the role played by ICTs in the distribution of paid labour across time and space for highly educated people in the Netherlands, accounting for the role played by individual motivations and circumstances. It focuses on the level of spatial and temporal fragmentation of paid labour, rather than total working hours. The relaxation of the associations between activities, times and places through ICTs might, according to Couclelis (2000, 2004), cause human activities to fragment. As a concept developed in human geography, activity fragmentation implies that a certain activity becomes divided into several smaller components, which can each be performed at different times and/or locations. Since it addresses both the spatial and temporal side of activities, activity fragmentation was judged to be a particularly suitable concept to study the relation between ICTs and the organisation of paid labour. In doing so, it will provide insight into current paid labour patterns that form a benchmark to which future assessments of the spatiotemporal organisation of paid labour can be compared to determine whether reorganisation is occurring. Throughout the paper particular attention will be paid to possible gender differences in the relation between ICTs and the organisation of paid labour. This is because men and women tend to differ regarding the spatial and temporal flexibility of their various activities and ICT usage (Kwan, 2000; Schwanen and Kwan, 2008; Tivers, 1985).

The following section discusses the concept of activity fragmentation and the individual motivations and circumstances expected to be related to the use of ICTs and the spatiotemporal organisation of paid labour. The Research design section presents the data consisting of unique two-day combined activity, travel and communication diaries and a questionnaire, administered among over 500 mainly highly-educated Dutch couple households. The Operationalisation of variables section describes the operationalisation of the fragmentation of paid labour, associated variables and the selected methods of multivariate analysis, the results of which are presented in the Results section.

Theoretical background

The temporal and spatial fragmentation of paid labour

By addressing both the spatial and temporal aspects of activities, the activity fragmentation concept offers a comprehensive view on the spatiotemporal diversification of employment activities in relation to ICT use (Alexander et al., 2011; Couclelis, 2000, 2009; Hubers et al., 2008; Lenz and Nobis, 2007). The concept of fragmentation of paid labour that underlies the current paper is inspired by Couclelis (2003: 11) and is defined as follows:

Fragmentation is a process whereby the paid labour activity is divided into several smaller components, which are performed at different times and/or locations.

In this paper, two types of activity fragmentation are distinguished: temporal fragmentation – different times at which the smaller components are performed – and spatial fragmentation – different locations at which the components are performed. Taking a typical working ‘day in the life of’ the first author as an example, her paid labour activities usually commence by quickly checking her work e-mail account in the early morning at home. During her 40-min train commute, she usually reads some scientific articles. After arriving at her university office, she first answers some e-mails, then works on her research, during which she regularly checks her e-mail. The afternoon might include a meeting with colleagues or grading students’ papers, and before heading back home she checks her e-mail for the last time. Some more work-related reading is done on the way home where she checks her work e-mail at least once during the evening. This example demonstrates how the paid labour activity comprises such components as e-mailing, reading articles or having meetings. These components in turn consist of multiple fragments or activity episodes – uninterrupted stretches of time devoted to a particular component. In this example, the component of e-mailing consists of at least five activity episodes. With the term activity location, we denote the unique locations where the activity episodes are performed. In the above example, the component of e-mailing is spatially fragmented across two activity locations: the home and the employment location. For the first author at least, the notion that ICT usage fragments activities holds true. Not only is the component of e-mailing itself fragmented both temporally and spatially, it also fragments other components temporally.

In previous work, Hubers et al. (2008) developed basic guidelines for assessing the extent and form of activity fragmentation proposing three relevant dimensions (Figures 1 and 2): the number of fragments; the distribution of the sizes of the fragments; and the configuration of the fragments. The first two dimensions describe how much a certain activity is fragmented; the third explains in what way the activity is fragmented and provides valuable insights into the spatial and temporal patterns the different activity locations and episodes form. Obviously, an activity such as paid labour can be considered more fragmented if it consists of a greater number of episodes and/or is performed at more different locations. The second dimension emphasises the time spent per episode. As the durations of individual episodes resemble one another more strongly, the pattern is considered more fragmented than if there is one large paid labour episode, combined with several much smaller episodes. Spatially, the paid labour pattern is more fragmented if, for example, equal amounts of paid labour are done at the workplace and the home, instead of most paid labour being performed at the workplace. In addition to the number of episodes, their respective sizes and the locations at which they are performed, their timing and spatial configuration may also be affected by the possibilities ICTs offer. The third dimension is therefore concerned with the temporal and spatial distances between the different activity episodes and locations. Suppose a person has three work meetings of 45 min scheduled on a certain day. When these are all executed in the morning, the paid labour pattern is different to when they are scheduled in the morning, afternoon, and evening. Likewise, if the locations of these meetings are all in different offices in the same city, the spatial pattern of paid labour is different to when all episodes take place in offices in three different cities. The growing use of ICTs can also affect the spatial distances between the work locations. For example, if ICTs increase the utility of travel time, people might be less daunted by distant work locations, possibly leading to more spread out work location patterns (Lyons and Urry, 2005).

Figure 1.

Dimensions of temporal fragmentation (source: Alexander et al. 2011).

Figure 2.

Dimensions of spatial fragmentation (source: Alexander et al. 2011).

The spatiotemporal organisation of paid labour: The role of ICT strategies, use, affordability and opinions

Time geography has drawn attention to the fact that activities are subject to space–time constraints, meaning that their execution is often tied to specific times and locations (Hägerstrand, 1970). The degree to which the timing, duration or the place where activities are undertaken can be altered, determines the level of space–time flexibility of activities (cf. Kwan 2000). A higher level of space–time flexibility enables more fragmentation. ICTs can enhance the space–time flexibility of activities by reducing the rigidity of their space–time constraints (Schwanen and Kwan, 2008 for an extensive discussion). Therefore, ICTs open up more possibilities for activities to become fragmented in both space and time. Telecommuting is a commonly used example. However, ICTs themselves can cause new and enhanced constraints; although a person may no longer be dependent on the office to work, he or she instead depends on the presence of a laptop or a fully charged cellular to telecommute (Dijst, 2004; Schwanen and Kwan 2008). Likewise, people have to acquire and maintain a certain knowledge level and experience in order to use ICTs effectively and be willing and able to pay for their use. And although ICT usage can erase certain space–time constraints, others may surface (see also Leonardi et al., 2010). The complexity of simply making an appointment for joint activity participation may increase when people’s activity schedules become more diversified. More time may then be spent coordinating and reconciling these various activity schedules, possibly increasing the time pressure people experience instead of reducing it.

Men and women often use ICTs in different ways (Lemish and Cohen, 2005). Men are more likely to bring the employment sphere into the domestic sphere through ICTs, whereas for women it is commonly the other way around (Chesley, 2005). Furthermore, given that women still tend to carry the primary responsibility for domestic activities which are often highly fixed in space and time (Tivers, 1985; Kwan 2000), they might be more inclined to employ ICTs to overcome these spatiotemporal constraints. The relation between the adoption of space–time flexibility-enhancing strategies and the structuring of paid labour will therefore be examined for men and women separately.

Non-ICT-related coping strategies

Previous geographical studies, especially those by feminist geographers, have demonstrated the role of human agency in dealing with space–time constraints (see, for example, England, 1996; Hanson and Pratt, 1995; Jarvis, 2005; Schwanen, 2008). People actively (try to) manage the space–time constraints they encounter and develop various strategies for doing so. People actively manage constraints related to paid labour not only by using ICTs, but also by adopting non-ICT-related coping strategies. Moreover, paid labour is strongly interlinked with people’s unpaid labour activities. Therefore, strategies adopted to deal with the space–time constraints of unpaid labour could also affect the spatiotemporal fragmentation of paid labour in an indirect way. These strategies can be categorized as follows:

Strategies aimed at a reorganisation of employment activities. Examples include decreasing the number of hours of paid labour, or finding a job in an organisation with extensive work/life policy.

Strategies aimed at a reorganisation of care-giving activities. For example, bringing children to a professional day-care centre.

Strategies aimed at a reorganisation of housework activities. Employing a cleaning lady, for instance, or performing domestic tasks faster.

The relations between these classes of strategies and the fragmentation of paid labour patterns are likely to differ. Strategies aimed at a reorganisation of employment activities by definition will directly alter the spatiotemporal arrangement of employment activities. Those aimed at the reorganisation of care-giving or housework activities can be related to either more or less fragmented paid labour patterns. The nature of some paid jobs makes it impossible to perform them at another time or place. In these cases, strategies might be called upon that adjust unpaid to paid labour responsibilities. On the other hand, strategies that decrease the amount of time people spend on domestic responsibilities enable them to spend more time on paid employment, which is likely to increase its fragmentation.

Other factors influencing the fragmentation of paid labour

There are therefore several other factors which may influence the spatiotemporal fragmentation of paid labour. Employment and commute factors may strongly influence the extent and form of fragmentation. People with more autonomy over employment times, for example, may be more likely to fragment them. The commute mode and time can also be of relevance, as a 60-min commute by train is more suitable to perform paid labour en route than a 5-min commute by bicycle. The employment sector may be related to the structure of paid labour activities with, for example, the business sector having a reputation of long continuous working hours.

The presence and age of children in the household is also known to be related to work practices (Vlasblom and Schippers, 2006). Women with young children are likely to have a larger amount of domestic responsibilities with more rigid space–time constraints and therefore might display different work practices than those without. Besides these household characteristics, a person’s lifestyle orientation and time personality may also affect the extent and form of activity fragmentation. The term time personality refers to individual differences in behaviour, cognitions and affect related to time (Francis-Smythe and Robertson, 1999). As for the importance of personal motivations, people who value family life very highly might be more inclined to adjust their paid labour activities to care-giving responsibilities. People who get more satisfaction from paid labour may be more inclined to do the opposite. Regarding time personality, we expect people who tend to plan activities very accurately and those who hold the view that work and home spheres should be separated as much as possible to portray less fragmented paid labour patterns (Brannen, 2005). Moreover, the opportunities to engage in non-employment activities offered by the physical surroundings of the home and work locations may affect whether paid labour is interrupted by non-work activities (Jarvis, 2005). Shopping during lunch break, for example, is more likely if there are stores located close to the employment location.

Research design

The data were gathered through a survey among one-earner and dual-earner households residing in the Utrecht-Amersfoort-Hilversum area, the Netherlands in February–July 2007. Different neighborhoods in the study area were selected based on the average personal income, urban density, and the availability of a train station within walking distance. Addresses within each neighborhood were selected randomly using digital files containing all street addresses. In total, 13,500 selection questionnaires were distributed by surface mail. Respondents willing to participate in the main survey were sent a questionnaire and two-day combined activity, travel and communication diary. The activity diaries were all in paper format and distributed by surface mail, whereas the questionnaire could be filled in online or on paper (and returned via surface mail). In the questionnaire, respondents were asked to report which of a wide range of work-life coping strategies in the paid labour, care-giving and leisure domains they had adopted to balance their various responsibilities.

In the diaries people logged where they were, what activities they performed at that place, for how long, and with whom. They were also asked to provide details about their electronic communications (with whom they communicated, for how long, for what purpose, and by what means) at stationary locations and when traveling. For travel episodes, information was collected about the transport mode used, total trip duration, with whom they traveled, and whether other activities were performed while traveling (e.g. sleeping, working/studying, reading, listening to music, etc.). Two consecutive days were randomly allocated to each household to guarantee an even distribution across all weekdays (only the combination Saturday/Sunday was omitted because we were especially interested in workdays). Eventually, we collected 681 usable diaries. For information on the response rate, the reader is referred to Hubers et al. (2011).

In the current study, we use data for 557 person-days for respondents in couple households. This sample is not representative of the Dutch population. Higher-educated people are over-represented; the share of people with a bachelors or masters degree is 71% of the sample against 32% in 2005 for the total Dutch population (CBS Statline, 2011). Highly educated individuals were deliberately oversampled as they are more likely to make extensive use of ICTs for paid labour and other purposes. Additionally, because highly educated women tend to perform more hours of paid labour (Beckers et al., 2009), we expect work–life balance issues to loom larger on their agendas than on those of women with a lower educational attainment. The share of male and female participants is about equal.

At the time of the data collection, 2007, smartphones were used by only a very small number of people. As smartphones connected to the Internet greatly increase the opportunities to reorganise activities in time and space, the data presented in this study possibly underestimate present day levels of fragmentation. The data do, however, offer valuable insights into factors that cause or are at least related to activity fragmentation given the wide range of information on ICT-related variables, coping strategies, employment and commute factors, household characteristics, lifestyle orientation, time personality, and spatial context.

Operationalisation of variables

Defining paid labour activities

A crucial step in studying activity fragmentation is defining the activity and its constituting components. In the current study, the following five components of paid labour have been discerned: work meetings; work-related emailing; work-related Internet-browsing; working while travelling; and other work (a residual category which could be used to report working on one’s actual task and encompasses acts, such as teaching a class, writing a consultancy report, or giving medication to a patient; the exact content therefore depends on the respondents’ profession). Working while travelling includes, for example, the reading of reports or preparing for meetings, but also electronic communications with colleagues or customers and clients. Please note that people with a mobile job, such as taxi or delivery van drivers, were not included in the analysis. Admittedly, the selection of components is an arbitrary process and these are definitely not the only components that could have been discerned. However, they apply rather well to the paid labour practices of the respondents targeted in the data collection. As indicated earlier, each component can be made up of multiple activity episodes conducted at a single or several locations.

Fragmentation measures

Together, the episodes of the five components of paid labour discerned above form the basis of the fragmentation measures shown in Table 1. A detailed description of these measures is given in Alexander et al. (2011). The measures were calculated on the level of the day. The number of episodes and the number of locations represent the number of fragments dimension. The distribution of the sizes of fragments dimension is reflected by the time index measure, which is based on the number of minutes allocated to the specific components. For the spatial dimension, the spatial index measures how the time spent on these different components is distributed across the different stationary and mobile employment locations. As most people engage in paid labour only at a very limited number of locations, the configuration dimension could only be measured meaningfully for the time dimension. The standard inter-episode duration measures the ‘global’ dispersion of episodes for the day as a whole (see Figure 1). It provides an overall indication of whether these episodes are performed with much or little time spent on non-work episodes in between. However, as Figure 1 indicates, various possibilities of ‘local’ clustering are identified. To distinguish between the five timelines presented there, we also need an indication of whether the intervals in-between episodes have similar durations as in timeline A, or whether these are more varied as in the right timelines of examples B, C and D. The inter-episode duration index provides information on such local clustering. Values close to zero on this index indicate a paid labour pattern consisting of several successive paid labour episodes and a single outlier (timeline C); values close to one indicate multiple paid labour episodes with intervals of equal durations (timeline A); and values around 0.5 indicate multiple intervals with differing durations suggesting the presence of several local clusters (timeline B). When interpreted simultaneously, these last two measures provide a sense of how one’s paid labour activities are spread across the day.

Table 1.

Overview of fragmentation of paid labour measures.^a

	Dimension of fragmentation	Fragmentation measure	Description	Interpretation	Value
Temporal fragmentation	Number of fragments	Number of episodes	Simple count of the number of paid labour episodes on a given day	Higher number indicating more temporal fragmentation	min = 1, max = ∞
	Distribution of sizes of fragments	Time index	Divides the duration of each paid labour episode by the total time spent on paid labour on a given day and sums all the square roots of the outcomes of these divisions. Standardized to control for differences in number of episodes	Transformed in such a way that high values indicate more evenly distributed episode durations and thus more fragmentation	min = 0, max = 1
	Configuration of fragments	Standard inter- episode duration	Measures the standard duration in-between paid labour episodes (time spent on non-workrelated activities) by relating all episodes to the mean starting time and subtracting the time spent on other activities	Higher value indicating more time in-between paid labour episodes, and thus a temporally more spread out paid labour pattern	min = 0, max = 1440
	Configuration of fragments	Inter-episode duration index	Identical to distribution of the episode durations, but now based on the in-between episodes durations	Transformed in such a way that high values indicate that the time in-between paid labour episodes is more evenly spread, while low values indicate the presence of an outlier	min = 0, max = 1
Spatial fragmentation	Number of fragments	Number of locations	Simple count of the number of unique locations where paid labour is conducted on a given day	Higher number indicating more spatial fragmentation	min = 1, max = ∞
Spatial fragmentation	Distribution of sizes of fragments	Spatial index	Measures how the total time spent on paid labour on a given day is fragmented across different locations. Standardized to control for differences in number of episodes	Transformed in such a way that high values indicate that time spent on paid labour is more evenly distributed across the different work locations and is thus more spatially fragmented	min = 0, max = 1

Equations and a detailed description of the calculation of the fragmentation measures are provided in Alexander et al. (2011).

Independent variables and method of analysis

A total of 56 independent variables, classified into seven distinctive blocks, were tested (see Table 2). Various ICT-related variables were included, ranging from ICT-enabled coping strategies, overall ICT usage and experience indicators, affordability and general interest in and opinions regarding ICTs. Some variables such as the adventurous personality and negative attitude towards ICT variables were construed using factor analysis. Another set of 15 statements included in the questionnaire to measure the time personality (Francis-Smythe and Robertson, 1999) of respondents resulted in five separate components interpreted and labeled as follows:

tendency to keep a strict schedule

avoidance of home-to-work and work-to-home spillover

time pressure

appreciation of connectedness through ICTs and enjoyment of multitasking

level of autonomy over time use in general.

Table 2.

Overview of independent variables.^a

	Women	Men	Total
ICT strategies, use, affordability and opinion
Work from home [0 = never, 2 = once a week or more]	.64	.74	.69
Buy laptop to enable working anywhere, anytime	10%	13%	11%
Regularly make private calls/send private e-mails during working hours	28%	40%	34%
Regularly make work calls/send work e-mail outside working hours	42%	49%	46%
Number of hours internet use per week	5.57	6.79	6.22
Frequency of internet use [0 = hardly ever, 6 = more than 5 times a day]	4.28	4.79	4.55
Years of internet experience	8.7	9.6	9.18
Mobile phone plan^b: Unlimited number of minutes	21%	47%	35%
Limited number of minutes	29%	22%	25%
Prepaid	47%	27%	37%
Receive (partial) reimbursement from employer for: Mobile phone costs	17%	42%	30%
Internet expenses	5%	20%	13%
Be highly interested in ICTs	−.197	.332	.085
Negative attitude towards ICTs	−.082	.047	−.013
Coping strategies
Decrease number of hours paid labor	43%	17%	29%
Get job closer to the home	22%	11%	16%
Get job in an organisation with extensive work/life policy	19%	12%	15%
Regularly use commute time to perform or arrange/organise personal matters	17%	20%	19%
Regularly use commute time to perform work-related tasks	13%	20%	17%
Adjust one's working hours to working hours of partner	31%	31%	31%
Limit number of extracurricular activities of children	21%	19%	20%
Teach children to do things independently	61%	62%	61%
Perform domestic tasks faster	41%	29%	35%
Lower standards for cleaning/tidying up	65%	42%	53%
Save up domestic tasks to perform them all at once	43%	35%	39%
Perform domestic tasks on fixed days	44%	38%	41%
Do groceries for multiple days at once	81%	78%	79%
Do groceries by car	32%	43%	38%
Combine daily shopping trips with other activities	49%	28%	38%
Division of domestic tasks beteen partners: shop for groceries [1 = mostly ♀, 7 = mostly ♂]	3.09	3.10	3.09
Number of timesaving domestic appliances [freezer, dishwasher, tumble drier, microwave oven]	3.25	3.37	3.31
Use ready-made meals [1 = hardly ever, 5 = once a week or more]	2.04	1.93	1.98
Use dry cleaner’s and/or ironing service	1.33	1.30	1.31
Own second car	43%	45%	44%
Employment and commute factors
Number of hours of paid labor	28	37	33
Number of hours of paid labor partner	34	21	27
Employment time sovereignty [1 = no, 3 = total control over working hours]	.92	1.23	1.08
Net household income [1 = 1000 euro or less, 5 = 4000 euro or more]	3.79	3.87	3.83
Respondent holds management position [expressed in number of employees one supervises]	6.5	10.1	8.5
Partner holds management position [expressed in number of employees partner supervises]	8.5	3.4	6.1
Sector of employment: Business	9%	21%	15%
Healthcare	36%	8%	21%
Education	22%	13%	17%
Governmental	9%	16%	13%
Commute by^c: Car	43%	51%	47%
Train	8%	8%	8%
Commute time in minutes	29	33	31
Work on weekend day	9%	8%	8%
Household characteristics
Child aged: 5–12	25%	31%	28%
13–16	21%	26%	24%
Chauffeuring responsibilities [number of chauffeuring episodes on working day]	0.33	0.15	0.24
Partner has higher vocational or university degree	59%	58%	58%
Partner unemployed	12%	16%	14%
Lifestyle orientation
Importance attached to paid labor^d [1 = totally disagree, 7 = totally agree]	3.11	3.38	3.26
Importance attached to leisure^d [1 = totally disagree, 7 = totally agree]	3.57	3.54	3.55
Adventurous personality	−.029	.073	.025
Time personality
Tendency to keep a strict schedule	.223	−.187	.004
Avoidance of home-to-work and work-to-home spillover	.021	−.235	−.115
Level of autonomy over time use in general	−.308	.091	−.095
Time pressure	.029	−.030	−.002
Appreciates connectedness through ICTs and enjoys multitasking	.110	−.004	.049
Spatial context
Density residents of employment location context	3.30	3.04	3.16
Density of shops and restaurants of employment location context [number of employees per m² of 4-digit zip code area]	0.73	.81	.77
Relative proportion of employees and residents of employment location context [higher score indicating relatively more employees per m² of 4-digit zip code area]	12.88	16.24	14.66

Presented as either a mean score, or percentage of number of persons with a value of 1 in case of a dummy variable.

Reference category is formed by people who have a subscription with limited number of minutes.

Reference category is formed by other commute modes, e.g. walking, bus, tram.

Mean score on five statements about the importance ascribed to these life domains.

All analyses were performed for men and women separately using linear modelling. As ordinary least squares (OLS) regression is inappropriate for non-negative or otherwise censored variables (Greene, 2002), we used tobit regression models for the fragmentation measures consisting of non-negative ratio variables and Poisson regression models for the non-negative integer variables. The standard Poisson model assumes the mean and variance to be equal. When this assumption is violated and the variance is larger than the mean (overdispersion) or lower (underdispersion), specific models that address this issue are preferred. For this reason, our fragmentation measure for the number of episodes was regressed using a negative binomial model for overdispersed variables, and a gamma model was selected for the underdispersed fragmentation measure for the number of locations. Since the pool of potential explanatory variables was very large, we reduced this set in several steps during the model construction phase until we arrived at model specifications with only variables significant at the p < .10 level for each fragmentation measure.

To determine which individual or categories of explanatory variables are most capable of explaining differences in the fragmentation of women’s and men’s paid labour patterns, the contributions to the model fit of each separate variable and group of variables (e.g. ICT variables, coping strategies, etc.) were calculated. This is a heuristic approach to make comparisons between women and men for one and the same fragmentation measure, and between the six fragmentation measures for each gender separately. To this end, the improvement in model fit was estimated by calculating the difference between the log likelihood (which is the statistic of the degree of misfit of the model to the data) of the end model (LL(F)) and the log likelihood of a model from which the respective (category of) explanatory variable(s) was omitted (LL(SV)). This difference was then standardized by dividing it by the log likelihood of a constants-only model (LL(C)). The larger the number, the larger the increase in goodness of fit of the model when a certain (group of) explanatory variable(s) is included.

Results

Descriptive analyses of the fragmentation measures

Before presenting the results of the multivariate analyses, we discuss some descriptive statistical measures of possible differences between men and women regarding the fragmentation measures. Table 3 indicates that men’s paid labour activities tend to be both temporally and spatially more fragmented than women’s. Men on average have more paid labour episodes and work at more different locations than women (dimension 1). The durations of men’s paid labour episodes are also more equal to one another than those of women and on average last shorter (89 versus 102 min, respectively), although this difference is not statistically significant. Furthermore, as the time men spend on paid labour is more evenly distributed across locations than women’s, their paid labour is considered more spatially fragmented (dimension 2).

Table 3.

Descriptive statistics fragmentation measures.

		N	Minimum	Maximum	Mean	Std. deviation
Temporal fragmentation
Dimension 1
Number of episodes	♂	297	1	28	7.18^a	5.074
Number of episodes	♀	260	1	25	5.53^a	4.449
Dimension 2
Time index	♂	297	0	1	0.73	0.300
Time index	♀	260	0	1	0.66	0.371
Mean episode duration	♂	297	1	668	89.36	102.885
Mean episode duration	♀	260	2	610	101.64	109.376
Dimension 3
Standard inter-episode duration	♂	267	0	388	39.70	53.234
Standard inter-episode duration	♀	208	0	368	36.86	55.661
Inter-episode duration index	♂	218	0	1	0.44^a	0.317
Inter-episode duration index	♀	152	0	1	0.33^a	0.315
Mean inter-episode duration	♂	267	0	776	38.20	85.264
Mean inter-episode duration	♀	208	0	735	40.14	94.060
Spatial fragmentation
Dimension 1
Number of locations	♂	297	1	5	1.85^a	1.205
Number of locations	♀	260	1	5	1.46^a	0.876
Dimension 2
Spatial index	♂	297	0	1	0.22^a	0.330
Spatial index	♀	260	0	1	0.15^a	0.289

Difference between men and women statistically significant at p < 0.01.

The higher mean scores on the two temporal measures for the configuration of the fragments (dimension 3) indicate that for men the durations between paid labour episodes are both longer and more equal in size than for women. Men’s paid labour episodes thus appear to be dispersed more strongly and evenly throughout all parts of the day, whereas women’s paid labour episodes are from a temporal perspective more clustered both globally and locally, although only the difference in local clustering is statistically significant. However, since for both men and women the average score on the inter-episode duration index is closer to zero than to one, the temporal pattern of their paid labour episodes is often characterized by an outlier or multiple local clusters rather than by evenly spread out paid labour episodes.

Results of multivariate analyses

The model fit improvements presented in Table 4 show that coping strategies are the only group of variables contributing to all 12 of the estimated models. Employment and commute factors are also related to most of the fragmentation measures, as are ICT strategies, use, affordability, and opinion and time personality. However, when we look at the actual size of the improvements to the model fit, a different picture emerges. Not coping strategies, but employment and commute factors make the biggest contributions to the models as they are responsible for the largest model fit improvements in half of the models. Employment and commute factors explain most of the differences in the temporal fragmentation (number of fragments and distribution of sizes dimension) of paid labour patterns for both women and men. For the other fragmentation measures, the results are less similar between men and women. Coping strategies are mainly related to how men’s paid labour is distributed across the day (temporal configuration dimension), whereas for women, spatial context and ICT strategies, use, affordability and opinion make the largest contributions to these models. And while for men the spatial fragmentation of their paid labour is most strongly related to employment and commute factors, for women it is most strongly related to the coping strategies they have adopted. Overall, the number of locations where men perform paid labour is related to a much greater variety of explanatory variables than women’s. On the whole, ICT strategies, use, affordability and opinion are the second most important group of variables when it comes to explaining the spatial fragmentation of paid labour patterns for both sexes.

Table 4.

Overview of standardized model log likelihood improvements per group of independent variables, largest improvements in bold.^a

	ICT strategies, use, affordability and opinion	Coping strategies	Employment and commute factors	Household characteristics	Lifestyle orientation	Time personality	Spatial context
Women
Number of episodes	3.1	22.9	75.1	–	–	7.2	–
Time index	–	36.2	56.9	5.2	–	11.4	–
Standard inter-episode duration	27.0	19.8	–	–	–	13.5	33.4
Inter-episode duration index	78.8	27.4	10.8	–	–	16.3	–
Number of locations	22.1	52.3	20.4	–	–	–	–
Spatial index	26.3	33.6	5.9	–	–	6.6	4.9
Men
Number of episodes	13.7	6.6	74.1	18.3	–	–	–
Time index	31.8	8.3	40.8	18.6	–	–	–
Standard inter-episode duration	–	46.8	41.8	–	11.9	24.2	–
Inter-episode duration index	41.6	49.8	43.3	–	–	–	–
Number of locations	21.3	14.9	48.2	12.6	6.9	16.6	–
Spatial index	23.5	20.9	26.9	20.5	–	11.5	–

Contributions of each set of variables was calculated using the following equation: $\frac{(LL (SV) - LL (F))}{LL (F) - LL (C)} * 100$ .

ICT strategies, use, affordability and opinions

Table 5 presents the independent variables with the largest contributions to the models and which are therefore the most strongly related to the fragmentation of paid labour. Due to space restrictions, the discussion of the results in this and the following sections will largely be limited to these variables. Nevertheless, Tables 4 and 5 do not tell us whether the explanatory variables are related to more or less fragmented paid labour patterns, nor whether this differs between the various variables belonging to a certain category. This information is presented in Table 6 in Appendix 1 which contains the final specifications of the models for women and men separately. Statistically insignificant variables have been omitted. For the most part, these relations suggest more fragmentation, supporting our expectations. Yet, there are also instances where ICTs may be used to avoid (spatial) fragmentation as we have defined it. This situation is exemplified by the negative relation between regularly making private calls or sending private e-mails during employment hours and the spatial fragmentation of both women and men’s paid labour activities. Perhaps the inability to perform their paid labour at other locations is compensated by arranging domestic issues from the employment location by means of the mobile phone or e-mail. Comparing the results for men and women also shows some interesting differences, suggesting that both adopt ICTs in different ways and possibly with different aims. While working from home, for example, is related to a larger number of paid labour episodes for men, it does not affect the number of episodes but only the distribution of these episodes across the whole of the day and across their various employment locations for women.

Table 5.

Count of three highest scoring individual independent variables based on their standardized model log likelihood improvements.^a

	Temporal fragmentation			Spatial fragmentation			Total
	Women	Men	Overall	Women	Men	Overall	Women	Men	Grand
ICT strategies, use, affordability and opinion
Work from home	2	1	3	1		1	3	1	4
Years of internet experience	1		1		0	0	1		1
Mobile phone plan: Prepaid				0	1	1		1	1
Receive (partial) reimbursement from employer for mobile phone costs		1	1	0		0		1	1
Coping strategies
Get job closer to the home	1		1				1		1
Regularly use commute time to perform or arrange/organise personal matters	1		1				1		1
Regularly use commute time to perform work-related tasks				1		1	1		1
Save up domestic tasks to perform them all at once				1	0	1	1		1
Perform domestic tasks on fixed days		0	0	1		1	1		1
Own second car		0	0	1		1	1		1
Employment and commute factors
Number of hours of paid labour	2	2	4				2	2	4
Employment time sovereignty	0		0	1	0	1	1		1
Net household income	2	1	3				2	1	3
Respondent holds management position		1	1		1	1		2	2
Partner holds management position	0		0		1	1		1	1
Work on weekend day	2	3	5	0	1	1	2	4	6
Household characteristics
Child aged: 5–12	0	1	1					1	1
13–16		0	0		1	1		1	1
Partner has higher vocational or university degree		1	1					1	1
Time personality
Tendency to keep a strict schedule	1		1				1		1
Avoidance of home-to-work and work-to-home spillover		1	1		2	2		3	3
Spatial context
Log for relative proportion of employees and residents of employment location context	1		1				1		1

Contributions of each individual variable was calculated using the following equation: $\frac{(LL (SV) - LL (F))}{LL (F) - LL (C)} * 100$ .

For women who have bought a laptop with the goal to (in principle) be able to work anywhere and anytime, paid labour episodes are more evenly spread across the day. These results could indicate that women alternate paid labour with housework and care-giving responsibilities. Purchasing a laptop, however, is not associated with the spatiotemporal organisation of men’s paid labour.

For both men and women, years of Internet experience is related to how they arrange paid labour activities in space and time. The intervals between the paid labour episodes of women who have been using the Internet for more years are less similar, suggesting that their paid labour patterns tend to take the shape of different local clusters, or as a large successive block of paid labour episodes combined with an outlying episode. The more years a man has been using the Internet, the less evenly spread across his different employment locations his time spent on paid labour is.

The financial aspects of mobile phone use are also relevant as those using a prepaid phone experience less spatial fragmentation of their paid labour activities, presumably because they use the phone less for paid labour purposes than those with some sort of subscription. This claim is further substantiated by the greater spatial fragmentation among women who receive a financial reimbursement from their employer for mobile phone use and are therefore perhaps expected to use it more for work-related phone calls. In contrast, women who get their Internet expenses reimbursed by their employer show less spatial fragmentation. Comparable to women, men who use a prepaid phone have fewer employment locations. In addition, the time spent on paid labour is also less evenly distributed across their different employment locations. Men who are financially reimbursed for mobile phone use by their employer do not experience more spatial fragmentation as women do but more temporal fragmentation; the durations of their paid labour episodes tend to be more equal.

Coping strategies

The relationships between the adoption of strategies to ease the juggling of employment and domestic responsibilities and the extent of fragmentation of paid labour activities are complex and differ between men and women. For women, adoption of coping strategies, especially those aimed at a reorganisation of domestic tasks, is mainly related to the spatial fragmentation of their paid labour: both the number of employment locations but especially the distribution of their working time across these locations. In contrast, men’s coping strategy adoption is mainly related to the temporal configuration of their paid labour patterns. As we expect men with a sizeable domestic workload to adopt coping strategies more frequently, their unpaid workload appears to influence how they schedule their paid workload.

With regard to the coping strategies aimed at a reorganisation of employment activities, the results indicate that, for women, purposively finding a job in an organisation with an extensive work/life policy is related to a spatially more fragmented paid labour pattern, suggesting that women actually use the spatial flexibility such policies offer. In contrast, the paid labour episodes of men who have adopted this strategy are less spread across the working day. This could indicate that these men prefer to protect their evening hours at home from being interfered by paid labour. As expected, the paid labour episodes of women who regularly use their commute time to perform work-related tasks are spatially more fragmented as indicated by their higher number of employment locations. Additionally, the intervals between the paid labour episodes of those who frequently perform or arrange personal matters while commuting, tend to be longer. Perhaps these women have many responsibilities outside of paid labour that are alternated with paid labour episodes, as a result of which paid labour gets more spread out across the day.

Strategies aimed at a reorganisation of housework activities show mixed results and clearly differ between men and women. Time-saving strategies like, for example, performing domestic tasks faster, appear to have saved up time for engaging in more paid labour activities and are therefore related to more fragmentation for women, both temporally and spatially. At the same time, other strategies such as performing domestic tasks on fixed days are mainly associated with less fragmentation of women’s paid labour, especially in a spatial sense. For men, these strategies are more frequently related to the temporal fragmentation of paid labour, particularly regarding the configuration dimension. Presumably, these men have taken on a larger share of domestic responsibilities, which increases their need for coping strategies. Focusing on the results for the configuration measures, it is difficult to find a clear pattern in their relations with the coping strategies as some are related to paid labour episodes that are distributed more evenly across the day, but others to more uneven distributions.

Other factors associated with the fragmentation of paid labour

Employment and commute factors are mainly associated with the temporal fragmentation of paid labour for women, whereas these are more related to spatial fragmentation for men. Regarding the number of hours of paid labour, as expected, men and women who work more hours tend to have more paid labour episodes. Additionally, for women working more hours, their different paid labour episodes have more equal durations, whereas for men working more hours, the nonwork episodes in between their paid labour episodes are slightly shorter. Employment time sovereignty also appears to be related to the fragmentation of both women’s and men’s paid labour patterns. For women, however, it is associated with both temporally and spatially more fragmented paid labour patterns, whereas for men it is only related to the spatial fragmentation of their paid labour. While a higher net household income comes with a more spatially and temporally fragmented paid labour pattern for women, for men it is only related to less temporal fragmentation. Perhaps in these households the women are more likely to be employed full time instead of part-time. Whether the male partner holds a management position is related to both women’s and men’s paid labour patterns. Managers tend to have a more spatially fragmented paid labour pattern, while female partners have temporally less fragmented paid labour patterns. Finally, for both genders, working on a weekend day is associated with both spatially and temporally, less-fragmented paid labour patterns.

The time personality variables differ considerably between men and women. The avoidance of home-to-work and work-to-home spillover is the only dimension of time personality that is related to the paid labour activities of men. Those who endorse this viewpoint both have a temporally more clustered and spatially less fragmented work schedule. All in all, these personality variables suggest that men make more efforts to separate paid labour and domestic activities in a spatial sense, whereas women are more aimed at keeping the two separated temporally. The paid work schedules of women who appreciate the connectedness enabled by ICTs and multitasking are spatially less fragmented, possibly because they use ICTs to compensate for a lack of physical mobility (for example, by mothering remotely. See also Rakow and Navarro, 1993; Schwanen, 2007).

Household characteristics are mainly related to the fragmentation of men’s paid labour patterns. Although the number of hours men perform paid labour might not differ between those with or without school age children, the spatiotemporal complexity of their employment patterns does. Lifestyle orientation is hardly related to the fragmentation of paid labour once other factors are taken into account and only so for men. In keeping with previous geographical studies, the spatial configuration of land uses surrounding the employment location is related to the extent of paid labour fragmentation, but only for women. For example, if their main employment location is situated within an area with a higher concentration of shops, restaurants and employment (as indicated by the relatively high proportion of employees in that area), they have longer intervals between their paid labour episodes. Apparently, the facilities such environments offer stimulate the performance of non-work-related activities in between paid labour episodes.

Conclusion and discussion

In this paper, we have examined the suggested relation between ICTs and the spatial and temporal (re)organisation of paid labour, using the concept of activity fragmentation. Two main findings can be identified. First, the results indicate that ICTs are indeed associated with the temporal and spatial fragmentation of men and women’s paid labour. The fact that we found ICT strategies, use, skills, affordability and opinions to be associated with both more and less fragmentation suggests that ICTs offer different opportunities to their users. Firstly, to (re)organise paid labour in both space and time. Secondly, as a compensation when such opportunities are lacking. And thirdly, for consolidation of spatially or temporally concentrated patterns (one location, or consecutive episodes) when fragmentation is seen to be undesirable (Leonardi et al., 2010). Examples of ICTs increasing the fragmentation of paid labour were working from home and, for women, buying a laptop enabling them to work anywhere and anytime. On the other hand, regularly making private calls or sending private emails during working hours serves as an example where ICTs may be used to compensate for a spatially fixed paid labour situation. In this case, resulting in less spatial fragmentation (fewer paid work locations), but possibly more temporal fragmentation if paid labour episodes are alternated with private phone calls or emails. Overall though, as the results showed more instances where ICTs were related to more fragmentation, it appears that they are mainly associated to the reorganisation of paid labour, rather than compensation or consolidation. Where ICTs were associated with less fragmentation, this mainly concerned spatial fragmentation measures, showing that they are more often used for compensation or consolidation of spatially concentrated paid labour patterns than temporally concentrated ones.

Second, gender differences exist, which indicates that men and women adopt ICTs in different ways and possibly with different aims. The strategy of buying a laptop to enable working anywhere and anytime is only related to the fragmentation of women’s paid labour episodes. And while working from home is related to the temporal fragmentation of both men’s and women’s paid labour, only for women is it related to its spatial fragmentation too. These findings demonstrate that ICTs do not have a one-way and straightforward effect on social processes regarding the organisation of paid labour within households. Our findings therefore support social constructivist arguments that the impact of ICTs depends on both technical capabilities as well as personal circumstances and motivations.

The insights developed in this paper into how ICTs are related to the reorganisation of paid labour in both space and time could refine the ongoing academic discussion of this relationship described in the introduction in several ways. Consideration of the fact that ICTs are not only related to the spatiotemporal fragmentation of the paid work activity but also of non-work activities, for example, opens our eyes to alternative ways of dealing with the space–time constraints attached to paid labour in cases where these themselves cannot be mitigated via ICTs. Arranging personal issues from the workplace using ICTs forms a case in point. Furthermore, the fact that besides ICTs various other variables were related to the spatial and temporal fragmentation of paid labour has implications if one aims to increase the spatiotemporal reorganisation of paid work. Of all explanatory variables, employment and commute factors were most strongly related to the fragmentation of paid labour. Simply providing employees with the technical equipment to work remotely without due consideration of their personal circumstances and motivations could offset the potential benefits of ICTs. This could explain why most studies so far have found little evidence of radical changes in the spatiotemporal patterns of paid work that were predicted to accompany increased ICT use (see also Fawcett and Song, 2009; Halford, 2005). In this paper too, we found only marginal levels of spatial fragmentation of paid labour as the average number of employment locations in our sample was less than two. Whether in the future personal circumstances and motivations will ever come to match the technical capabilities offered by ICTs (who themselves keep evolving as witnessed by the widespread adoption of smartphones since the time our data have been collected) to work whenever and wherever, only time will tell.

Several issues have arisen from this study that warrant additional future research. Probably the most pressing one concerns the issue of causality. While this study employed cross-sectional data, longitudinal data are needed to make any statements regarding issues of causality in the relation between ICTs and the reorganisation of paid labour in time and space. There is also the question of how the growth in smartphone use and increasing supply of free wi-fi and the opportunities these offer for adjusting the space–time constraints of activities, relate to the fragmentation of activities. The data used in this paper did, however, enable comparisons between men and women, revealing some interesting gender differences. Future research could further deepen our understanding of such gender differences by examining the spatiotemporal fragmentation of paid labour on the level of the household rather than the individual level. In how far does the level of spatial and temporal fragmentation of paid work differ between the partners? Is it sufficient if only one partner can perform his or her paid work in a flexible way? Finally, future research should shed more light on how ICTs are related to the fragmentation of the paid labour activities of people with a lower educational level to determine in how far the results found in the current study can be generalized to other population groups.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research reported on is part of the Innovative Land Use programme, financed within the context of the Investments in Knowledge Infrastructure Directive (BSIK) of the Dutch government.

References

Alexander

Hubers

Schwanen

et al. (2011) Anything, anywhere, anytime? Developing indicators to assess the spatial and temporal fragmentation of activities. Environment and Planning B: Planning and Design 38(4): 678–705.

Axtell

Hislop

Whittaker

(2008) Mobile technologies in mobile spaces: Findings from the context of train travel. International Journal of Human-Computer Studies 66(12): 902–915.

Bailey

Kurland

(2002) A review of telework research: Findings, new directions, and lessons from the study of modern work. Journal of Organizational Behaviour 23: 383–400.

Beckers

Hermans

Portegijs

(2009) Betaalde arbeid. In: Merens

Hermans

(eds) Emancipatiemonitor 2008, The Hague: The Netherlands Institute for Social Research SCP, pp. 77–115.

Brannen

(2005) Time and negotiation of work-life boundaries. Time and Society 14(1): 113–131.

Castells

(2000) The Rise of the Network Society, 2nd ed. Oxford: Blackwell.

CBS Statline (2011) http://statline.cbs.nl.

Chesley

(2005) Blurring boundaries? Linking technology use, spillover, individual distress, and family satisfaction. Journal of Marriage and Family 67(5): 1237–1248.

Choo

Mokhtarian

Salomon

(2005) Does telecommuting reduce vehicle-miles traveled? An aggregate time series analysis for the U.S. Transportation 32(1): 36–64.

10.

Couclelis

(2000) From sustainable transportation to sustainable accessibility: Can we avoid a new tragedy of the commons? In: Janelle

Hodge

(eds) Information, Place and Cyberspace: Issues in Accessibility, Berlin: Springer-Verlag, pp. 341–356.

11.

Couclelis

(2003) Housing and the new geography of accessibility in the Information Age. Open House International 28(4): 7–13.

12.

Couclelis

(2004) Pizza over the internet: e-commerce, the fragmentation of activity and the tyranny of the region. Entrepreneurship and Regional Development 16(1): 41–54.

13.

Couclelis

(2009) Rethinking time geography in the information age. Environment and Planning A 41(7): 1556–1575.

14.

Dijst

(2004) ICT and accessibility: An action space perspective on the impact of new information and communication technologies. In: Beuthe

Himanen

Reggiani

Zamparini

(eds) Transport Developments and Innovations in an Evolving World, Berlin: Springer, pp. 27–46.

15.

England

(1996) Who will mind the baby? In: England

(ed.) Who will mind the baby?, London: Routledge, pp. 3–19.

16.

Fawcett

Song

(2009) Modelling the use of space and time in the knowledge economy. Building Research & Information 37(3): 312–324.

17.

Felstead

Jewson

(2000) In Work, at Home: Towards an Understanding of Homeworking, London: Routledge.

18.

Francis-Smythe

Robertson

(1999) Time-related individual differences. Time Society 8(2): 273–292.

19.

Gershuny

(2000) Changing Times: Work and Leisure in Post-industrial Society, Oxford: Oxford University Press.

20.

Greene

(2002) LIMDEP Version 8.0 Econometric Modeling Guide Volume 2, Bellport, NY: Econometrics Software Inc.

21.

Hägerstrand

(1970) What about people in regional science? Regional Science Association Papers 24(1): 7–21.

22.

Halford

(2005) Hybrid workspace: re-spatialisations of work, organisation and management. New Technology, Work and Employment 20(1): 19–33.

23.

Hanson

Pratt

(1995) Gender, work, and space, London: Routledge.

24.

Hislop

Axtell

(2007) The neglect of spatial mobility in contemporary studies of work: the case of telework. New Technology, Work and Employment 22(1): 34–51.

25.

Hubers

Schwanen

Dijst

(2008) ICT and temporal fragmentation of activities: an analytical framework and initial empirical findings. Tijdschrift voor Economische en Sociale Geografie 99(5): 528–546.

26.

Hubers

Schwanen

Dijst

(2011) Coordinating everyday life in the Netherlands: A holistic quantitative approach to the analysis of ICT-related and other work-life balance strategies. Geografiska Annaler B, Series B: Human Geography 93(1): 57–80.

27.

Jackson

Van der Wielen

(1998) Teleworking: International Perspectives from Telecommuting to the Virtual Organization, New York: Routledge.

28.

Jarvis

(2005) Moving to London Time: household co-ordination and the infrastructure of everyday life. Time & Society 14(1): 133–154.

29.

Kaufman-Scarborough

(2006) Time use and the impact of technology: examining work-spaces in the home. Time & Society 15(1): 57–80.

30.

Kwan

M-P

(2000) Gender differences in space–time constraints. Area 32(2): 145–156.

31.

Laurier

(2002) The region as a socio-technical accomplishment of mobile workers. In: Brown

Green

Harper

(eds) Wireless World: Social and Interactional Aspects of the Mobile Age, London: Springer-Verlag, pp. 46–61.

32.

Lemish

Cohen

(2005) On the gendered nature of mobile phone culture in Israel. Sex Roles 52(7/8): 511–521.

33.

Lenz

Nobis

(2007) The changing allocation of activities in space and time by the use of ICT – ‘Fragmentation’ as a new concept and empirical results. Transportation Research Part A: Policy and Practice 41(2): 190–240.

34.

Leonardi

Treem

Jackson

(2010) The connectivity paradox: using technology to both decrease and increase perceptions of distance in distributed work arrangements. Journal of Applied Communication Research 38(1): 85–105.

35.

Letherby G and Reynolds G (2003) Making connections: the relationship between train travel and the processes of work and leisure. Sociological Research Online 8(3). Available at: http://www.socresonline.org.uk/8/3/letherby.html (accessed 2 January, 2012).

36.

Lyons

Urry

(2005) Travel time use in the information age. Transportation Research Part A: Policy and Practice 39(2–3): 257–276.

37.

Lyons

Jain

Holley

(2007) The use of travel time by rail passengers in Great Brittain. Transportation Research Part A: Policy and Practice 41(1): 107–120.

38.

Nansen

Arnold

Gibbs

et al. (2010) Time, space and technology in the working-home: an unsettled nexus. New Technology, Work and Employment 25(2): 136–153.

39.

Ory

Mokhtarian

(2006) Which came first, the telecommuting or the residential relocation? An empirical analysis of causality. Urban Geography 27(7): 590–609.

40.

Perrons

(2003) The New Economy and the work–life balance: conceptual explorations and a case study of new media. Gender, Work and Organization 10(1): 65–93.

41.

Rakow

Navarro

(1993) Remote mothering and the parallel shift: women meet the cellular phone. Critical Studies in Mass Communication 20(3): 144–157.

42.

Schwanen

(2007) Matter(s) of interest: Artefacts, spacing and timing. Geografiska Annaler, Series B: Human Geography 89(1): 7–21.

43.

Schwanen

(2008) Managing uncertain arrival times through sociomaterial associations. Environment and Planning B: Planning and Design 35(6): 997–1011.

44.

Schwanen

Kwan

M-P

(2008) The Internet, mobile phone and space–time constraints. Geoforum 39(3): 1362–1377.

45.

Tietze

Musson

(2002) When “Work” meets “Home”: temporal flexibility as lived experience. Time & Society 11(2/3): 315–334.

46.

Tivers

(1985) Women Attached: The Daily Lives of Women with Young Children, London: Croom Helm.

47.

Urry

(2004) Connections. Environment and Planning D: Society and Space 22(1): 27–37.

48.

Vlasblom

Schippers

(2006) Changing dynamics in female employment around childbirth: evidence from Germany, the Netherlands and the UK. Work, Employment and Society 20(2): 329–347.

49.

Wajcman

Rose

Brown

et al. (2010) Enacting virtual connections between work and home. Journal of Sociology 46(3): 257–275.