Hubers, C., Schwanen, T. And Dijst, M. (2008). ICT and temporal fragmentation of activities: An analytical framework and initial empirical findings. Jtijdschrift voor economische en sociale geografie, 99(5), 528-546. /
ICT and Temporal Fragmentation of Activities:
An analytical framework and initial empirical findings
Christa Hubers*
Tim Schwanen
Martin Dijst
UtrechtUniversity
Faculty of Geosciences
Department of Human Geography and Planning
PO Box 80.115
3508 TC Utrecht
The Netherlands
* Corresponding author
Phone: +31-30-2532407
Fax: +31-30-2532037
E-mail:
Abstract
It is commonly believed that the widespread use of Information and Communication Technologies (ICTs) facilitate the fragmentation of daily activities across times and spaces. However, a clear conceptualization of what fragmentation is and how it can be measured empirically has been lacking. As a consequence, hardly any empirical evidence has been provided for these notions. The goal of this paper is twofold: (1) to propose a theoretical and methodological framework for identifying and measuring activity fragmentation; and (2) to assess temporal fragmentation empirically and consider its associations with ICT usage while controlling for sociodemographic variables, residential context, day of the week, activity pattern characteristics and some attitudinal variables. Activity fragmentation is defined as a process whereby a certain activity is divided into several smaller pieces, which are performed at different times and/or locations. The proposed theoretical and methodological framework covers three main dimensions of fragmentation: the number of fragments; the distribution of the sizes of fragments; and the temporal configuration of fragments. Based on travel diary data from The Netherlands the analytical results are insightful and promising. The framework is not only capable of detecting temporal activity fragmentation for various trip purposes, but there are also indications of a positive relation between ICT-usage and temporal fragmentation.
Keywords:
Activity fragmentation; Information and Communication Technologies; The Netherlands
1
/ This is a pre-publication version of the following article:Hubers, C., Schwanen, T. And Dijst, M. (2008). ICT and temporal fragmentation of activities: An analytical framework and initial empirical findings. Jtijdschrift voor economische en sociale geografie, 99(5), 528-546. /
1. INTRODUCTION
It is commonly believed that, due to developments in Information and Communication Technologies (ICTs), “professional and social relations can be established and maintained almost equally easily over any distance across the globe” (Couclelis, 1996, p. 388). As a consequence, activities seem to be getting less firmly linked to fixed spatial locations and times which might be manifested in the fragmentation of activities into tasks that are widely distributed over space and across time (Couclelis, 2000; Dijst, 2004). This so-called ‘activity fragmentation’ is foreseen to have considerable impacts on the daily life of individuals. The fragmentation of daily activities across times and spaces facilitates the blurring of the boundaries between previously separated life domains of work, care and leisure and may offer new opportunities as well as challenges to people juggling paid labor and care giving responsibilities. It is furthermore foreseen to have considerable impacts on transportation flows since the predicted increases in travel demand that may result from activity fragmentation may increase road congestion across time (especially during what are now considered non-peak hours) and space (new bottlenecks in addition to existing ones). The demand for certain facilities and services may also decrease, manifest itself at other times, or facilities may experience alterations in their functions. E-shopping, for example, may ultimately reduce brick-and-mortar stores to showrooms for products that are than purchased on the Internet. Likewise, telecommuting may reduce the relevance of the physical nearness to the employment location when searching for a new residence (Ory and Moktharian, 2006).
However, there is almost no empirical support for the propositions about activity fragmentation. Therefore, it is not known to what extent such fragmentation is indeed occurring, how it takes place and what the role of ICT in this process is. This is among others due to the fact that activity fragmentation as a concept is intuitively sensible but difficult to grasp theoretically, methodologically and empirically. There is not only a lack of a clear framework for analysing and measuring fragmentation but also of appropriate data. A first attempt to measure fragmentation empirically has been made by Lenz and Nobis (2007). Although these authors find some evidence of the occurrence of activity fragmentation, their research does not provide a detailed insight into the ways in which activities are fragmented. It furthermore remains unclear what the relative strength of the relation is between ICT and activity fragmentation. There are reasons to believe that the extent of activity fragmentation is also related to sociodemographic factors or the residential and temporal context (cf. Hanson, 1982; Yamamoto and Kitamura, 1999). We expect ICTs to function as facilitators of fragmentation because they create new choice sets for the performance of activities in space and time, rather than them being determinants of fragmentation. We hypothesize that characteristics of both the activity and the individual determine whether or not these new options will actually be chosen (cf. Mokhtarian et al. 2006).
In order to address these issues we aim, firstly, to develop a theoretical and methodological framework for measuring fragmentation of activities. Our second aim is to apply this framework on travel diary data in order to assess the relevance of ICT-usage for temporal activity fragmentation.
For the development of the theoretical and methodological framework, an interdisciplinary approach is employed. This framework will be presented in Sections 2 and 3. Sections 4-5 introduce the empirical analysis in which we describe the fragmentation for paid labor, shopping for daily and non-daily goods, and leisure activities, and assess whether fragmentation varies systematically with ICT usage, and possible other factors that as we will argue below might have an influence on the level of fragmentation. Note that the exploratory analysis in the current paper will be confined to the fragmentation of activities across time. Although activity fragmentation is usually discussed as a phenomenon in both space and time, the unravelling of temporal patterns is so complex that a separate paper is warranted. The paper ends with conclusions and a discussion.
2. AN INTERDISCIPLINARY APPROACH TO FRAGMENTATION
2.1. Defining Fragmentation
To avoid re-inventing the wheel, we have conducted an interdisciplinary literature search about the nature of fragmentation and measurement approaches.Fragmentation is a notion used in social and natural sciences and applied on various divisible phenomena or objects. Sociologists have extensively studied the temporal fragmentation of (leisure) activities, often in relation to time pressure (e.g, Mattingly and Bianchi, 2003; Sullivan, 1997). In human geography, economic geography and spatial planning spatial fragmentation refers often to the development of socio-spatial specialised zones in relation with segmentation of infrastructures (Graham and Marvin, 2001), segmentation and relocation of economic activities (Arndt and Kierzowski, 2001) or decentralised land-use governance (Ulfarsson and Carruthers, 2006).Computer scientists Mark et al. (2005) have investigated the extent of fragmentation of work activities to find out whether the development of computer software programs assisting workers in picking up their work after interruptions is warranted. But probably the most well known application of the fragmentation concept in the world of computers is concerned with hard disk fragmentation (Diskeeper Corporation Europe, 2006).This fragmentation process implies that accessing fragmented computer files consumes more time and necessitates users to de-fragment their hard disks. Ecology is nonetheless the discipline contributing the most relevant insights for our study: a vast literature exists on forest and ecosystem fragmentation, on different dimensions of fragmentation, and on measurement approaches(Rutledge, 2003).
While fragmentation has been studied in many research areas, each discipline employs its own specific definition to the concept. As a consequence, there is no unequivocal definition of fragmentation. Inspired by Couclelis (2003, page 11), we define fragmentation as:
a process whereby a certain activity is divided into several smaller pieces, which are performed at different times and/or locations
Generally speaking, two types of activity fragmentation can be discerned: temporal fragmentation– different times at which the smaller sub-tasks are performed – and spatial fragmentation – different locations at which the sub-tasks are performed. The current paper focuses on the temporal fragmentation.
An example helps to clarify the above definitions. Suppose a person wants to purchase a Flatscreen TV. She could start browsing the Internet for some general information on the types of Flat TVs available. She may then go to a brick-and-mortar electronics store to get a better grasp of the difference between Plasma and LCD technology and see with her own eyes the differences in the picture quality, and see which features she likes best. She may also read some independent product reviews on the Web or talk to Flat TV-owning friends, relatives or colleagues about their experiences. After having decided what TV to buy, she has to decide where to purchase it. Comparison sites on the Internet might be used to get the best bargain. Having chosen the dealer and consequently purchased the new Flat TV online, she finally has to determine how to have the product delivered, where and when.
Not all purchases will be made in this or similar ways. But the example shows clearly that the activity of shopping for a certain product comprises several sub-tasks (e.g. searching and evaluating product information; purchasing the product; and transporting the product or having it delivered) and this probably holds for the majority of shopping activities (see also Salomon and Koppelman, 1988). These sub-tasks can exist of smaller fragments. We use the term ‘activity episode’ to denote the different times at which these smaller fragments a sub-task consists of are performed. If on a given day the individual in the above example in the morning talks to some colleagues at the office about their Flat TVs, stops by an electronics store to view some possible TVs after work and later at night browses the Internet for some more product information from home, the sub-task of searching and evaluating product information consists of three activity episodes throughout that single day. The term ‘activity location’ represents the different locations at which the smaller fragments of the sub-task are performed. Again taking the sub-task of searching and evaluating product information as an example, we may say this sub-task is spatially fragmented across three activity locations: the employment location, the electronics store and the home. The example also indicates the significance of ICTs in this process. Due to ICTs, the number of times when and locations where activity episodes can be performed has increased dramatically now that they are no longer exclusively dependent on shop opening hours and locations.
The discussion so far has not touched upon two complicating factors. First, our example does not make clear when an activity is fragmented as opposed to being two separate activities. In our view, the answer to this question greatly depends on the objectives of the study. If, for example, it seeks to examine whether the paid work activity of people who work from home is more often alternated with maintenance activities (activities that are performed for the upkeep of the household, such as cooking, cleaning and shopping) than is the case for people working in an office, a more general classification of paid and maintenance activities suffices. However, if one wants to find out whether the process of shopping for shoes is more or less fragmented than the process of shopping for a Flat TV, more detailed information on the sub-tasks that constitute both shopping activities is necessary.
Second, there are several concepts, such as balkanization and contamination or multi-tasking, that are intimately associated with fragmentation. We will discuss these related concepts briefly to demarcate what topics will not be investigated but may still be relevant to (future studies of) activity fragmentation. The term balkanization refers to the division of a place or country into several small political units that are often unfriendly to one another (New Dictionary of Cultural Literacy, 2002). With respect to daily activities, balkanization is concerned with the ways people evaluate activity fragmentation. Mark et al. (2005) showed that an interruption of a certain work activity was evaluated more negatively when the interrupting task had nothing to do with the original task. The interruption was, however, evaluated less negatively and sometimes even considered beneficial, when the interrupting and disrupted tasks were associated with one another. It also mattered whether the interruption was self-induced or imposed externally with the former being less negatively evaluated than the latter. Contamination, or multitasking,is derived from the sociological literature and is concerned with the fact that several activities can be performed simultaneously (Mattingly and Bianchi, 2003; Sullivan, 1997; Felker Kaufman et al., 1991). For instance, watching the television while eating, working on the train while travelling, or calling a friend with your cellular phone while standing in line at the grocery store.With multitasking the emphasis is on how at a single moment in time multiple activities are performed, whereas with temporal activity fragmentation the emphasis is on how a single activity is performed at multiple times and locations. Several transportation studies have indicated that Internet and mobile phone use stimulate multitasking (Kenyon and Lyons, 2007; Schwanen and Kwan, 2008). Since the description of temporal fragmentation is already very complex without addressing issues of how people experience it and the performance of multiple activities simultaneously, the latter two issues are left to future studies.
2.2 Dimensioning fragmentation
According to the ecological literature in particular, fragmentation can be seen as composed of three dimensions (Figure 1). The most commonly identified dimension is the number of fragments or segments in which a given object (activity, forest or hard disk) is divided (Mattingly and Bianchi, 2003; Sullivan, 1997; Rutledge, 2003). Rutledge (2003, page 7) gives a simple but telling example: “A plate that is broken into 100 pieces is more fragmented than a plate broken into 10 pieces.” In our framework this means that when a certain sub-task is performed, for example, four times a day, this activity type counts four fragments, also called activity episodes.
The second dimension concerns the distribution of sizes of the fragments. As Rutledge (2003, page 7)continues: “Similarly, a plate broken into 10 pieces of equal size is more fragmented than a plate broken into 10 pieces, one of which is 90% of the original plate.” This is also recognized in social science for employment-related (Mark et al., 2005) and leisure activities (Sullivan, 1997).
Finally, the configuration of fragments is considered an important dimension of fragmentation in ecology (Rutledge, 2003). While Rutledge acknowledges that, in a strict sense, configuration is not related to spatial fragmentation, he argues that the survival of plant and animal species depends on the configuration of their habitat fragments. If habitats become too isolated, the survival of plants and animals is threatened. Although it is probably not a matter of life and death in social sciences, studying the configuration of activity fragments can provide valuable insights into their timing.ICT use may imply that activity episodes become more spread out across the day. This is shown clearly in the example in Section 2.1 where the person gathers product information from her colleagues in the morning during office hours, after work just before the closing time of the electronics store and again later at night on the Internet which conveniently has no closing time. Fragmentation can have a direct or indirect effect on the timing of activity episodes. As ICTs reduce the space-time fixity of activities they have a direct effect on the timing of activity episodes (Schwanen and Kwan, 2008). ICTs however may also increase the efficiency with which activities are performed, thereby reducing their duration. And since short activity episodes are more readily slotted into individuals’ activity schedules than longer ones, ICTs may have an indirect effect on the timing of activity episodes. Information on the number and/or duration of activities is insufficient for determining whether they are rescheduled to alternative moments. The distribution of the sizes of the activity episodes only tells us something about the duration of these activity episodes. By taking into account the temporal distances between the different activity episodes, which are calculated in the configuration dimension, the exact timing of these activity episodes can also be determined.
In short, we believe that the three dimensions should be analysed simultaneously even though they capture different aspects of activity fragmentation. The first two dimensions describe how much a certain activity is fragmented, whereas the third explains in what way the activity is fragmented and provides valuable insights into the spatial and temporal patterns that are formed by the different activity episodes and locations.
2.3 Factors Associated with Fragmentation
Based on previous research on determinants of activity and travel behaviour we expect that besides ICTs, other factors including sociodemographic factors, characteristics of the built environment (Hanson, 1982; Lu and Pas, 1999), factors concerning the day of the week (Yamamoto and Kitamura, 1999) and attitudinal variables (Farag et al., 2007) might also be related to the fragmentation of activities. Like Mokhtarian et al. (2006) we expect ICTs primary impact on activities is “to expand an individual’s choice set” (page 263), whereas characteristics of both the activity and the individual determine whether or not these new options will actually be chosen. Because the adoption of ICT and their use depends among others on socio-demographic factors, the effects of these variables should be controlled in an analysis of the associations between ICT ownership and use and activity fragmentation. This section provides a brief summary of the potential relations between other determinants of activity and travel behaviour and fragmentation.