ICT FOR LIFE

2 Impacts of ICT on activity-travel behavior: literature review

In the past decades, a large body of literature on the transportation impacts of ICT
has been produced. Many studies focus on specific applications of ICT, such as
telecommuting (or teleworking), teleshopping (or e-shopping), teleconferencing, etc.
Hamer et al (1991) studied the impact of teleworking on the travel behavior of
telecommuters and their family members. They found that teleworking resulted in
significant decreases in the total number of trips made by teleworkers (17%) and
peak-hour traffic by car (26%). Pendyala et al (1991) investigated the impact of
telecommuting on spatial and temporal patterns of household travel. They found
that the trip making and the total distance traveled by telecommuters were significantly
reduced. Balepur et al (1998) found that person-miles and vehicle-miles
traveled by center-based telecommuters were significantly fewer on telecommuting
days than on non-telecommuting days. While these studies show that telecommuting
leads to net reduction of trips and distance traveled, others argue that the impact of
telecommuting is rather complex in its nature and magnitude. Salomon and Salomon
(1984) suggested that the journey to work acts as a desired buffer between home and
work and thus the complete replacement of commuting trips may not be desired.
Harvey and Taylor (2000) suggested that there were clear differences in the levels of
social interaction between those who work at home and at conventional work place,
and there was a tendency for persons with low social interaction to travel more. They
argued that working at home might not diminish travel but rather simply change itspurpose. Mokhtarian and Salomon (2002) argued that the reported net impact of
substitution of telecommuting might be less if the induced demand, latent demand,
and long-term residential relocation effects were taken into account.
Different from that of telecommuting studies, findings of teleshopping (or
e-shopping) research seem less optimistic on the substitution effects, though
empirical evidence was not available until recently. Mokhtarian (2004) suggested
that ICT leads to the fragmentation and recombination of the basic elements of the
shopping process. She argued that there are many factors influencing the interaction
between e-shopping and travel: some result in reduced travel while others lead to
increased travel. As a result, the combined outcome may be negative (i.e. more
travel may be generated). Farag et al (2004) found that e-shoppers were people who
like to shop in general, suggesting that e-shopping might not lead to the reduction of
shopping trips. Douma et al (2004) revealed that e-shopping had little substitution
effects, rather, the Internet was used to modify shopping behavior by gathering
information of products or making trip more efficient. Visser and Lanzendorf (2004)
argued that business-to-consumer (b2c) e-commerce might increase the average trip
length and car use for shopping trips, because b2c e-commerce enhanced the current
process of decentralization and suburbanization of distribution systems. Farag et al
(2007) studied the interactions between e-shopping and in-store shopping and found
that searching online increases shopping trips, which in turn increases buying online,
suggesting complementarity effects between different shopping modes. Though only
a few, the reported studies on the impacts of teleconferencing on travel also suggest
limited substitution effects at best, and some complementarity (Mokhtarian and
Salomon 2002).
The studies on specific applications of ICT have contributed a great deal to our
knowledge about the direct impacts of ICT on physical travel. Nevertheless, by
focusing on specific application, these studies fail to recognize the interrelationships
among different forms of communication and are not able to reveal the combined
and system wide effects of such relationships (Mokhtarian and Salomon 2002). Thus,
instead of focusing on a particular type of ICT application, a number of studies adopt
the comprehensive approach by looking at all forms of communication and trying to
unveil the relationship between transportation and telecommunications in a broad
context. Selvanathan and Selvanathan (1994) adopted the aggregate approach to
analyze the relationships in consumer demand for private transportation, public
transportation and communication. They found that at consumer level the demands
for private transportation, public transportation and communication were pairwise
substitutes and the exponential growth in communication was at the expense of the
two types of transportation. Also taking the aggregate approach but from the
industry’s perspective, Plaut (1997) studied the relationships in industrial demand
for transportation and communication and found that the demand for transportation
and communication at industrial level was of complementarity. A similar aggregate
approach (only using activity indicators rather than monetary ones, in contrast to the
preceding two studies) was taken by Choo and Mokhtarian (2005) who examined at
the national level the causal relationships between telecommunications and travel,
considering supply, demand and costs and controlling factors on economy activities
and socio-demographics. The study found that the net effects of telecommunications
on travel were complementarity rather than substitution. Mokhtarian and Meenakshisundaram
(1999) studied at disaggregate level the interactions between three
types of communications, namely personal meetings, surface mails and electronicmails (phone, fax, e-mail). They found lagged complementarity effects across different
types of communication but no relationships between electronic forms of
communication and personal meetings or trips. Hjorthol (2002) examined the relations
between the different uses of home computers (for private purpose as well as
for paid work) and daily travel in Norway. The study found that the access to and use
of home computers had no significant substitution effects on travel; people who
owned home computer made less work trips, but they re-scheduled their activity and
as a result the total number of daily trips did not change.
Apart from defining ICT application in a broad way, findings of the studies on
single ICT application may be further enriched by adopting the holistic approach
and broadening the scope of investigation from single trip purpose to activity and
time use patterns. ‘The analysis of total human activity may lead to different conclusions
from the analyses of single trip purpose’ (Salomon 1985). As argued by
Vilhelmson and Thulin (2001), activities motivate travel and the effects of ICT on
travel are channeled through changing or facilitating activity participation and time
use. On the one hand, ICT may replace travel through substituting physical activities
(e.g., a banking activity at the bank) by ICT-based activities (e.g., e-banking). The
replacement leads not only to the direct substitution effects, but also indirect generation
or modification effects. On the other hand, ICT may facilitate physical travel
by providing real-time traffic information to help trip makers decide on departure
time, route and transport mode, etc. It is therefore difficult, if not impossible, to have
a thorough understanding about these possible transportation impacts of ICT, if we
do not analyze how ICT changes and facilitates individuals’ activity participation and
time allocation. The activity-based framework should thus be used to capture the
ICT and travel interactions (Golob and Regan 2001). In recent years, few attempts
have been made along this line. Kim and Goulias (2004) studied the effects of
changes in the availability and access to ICT on activity, travel and modal split.
Srinivasan and Athuru (2004) studied the adoption of ICT, participation in physical
and virtual activities and their impacts on travel patterns. The study, however, did
not simultaneously model the interrelationship between ICT usage, activity participation
and travel pattern. Consequently, the complex causal relationships between
ICT usage, time use and travel behavior could not be identified.
Despite of the drawbacks, these recent attempts have demonstrated that the
holistic approach of studying the interactions between ICT and transportation is a
promising avenue towards a better understanding of the transportation impacts of
ICT. More complete models that examine the relationships between ICT, activity,
travel and socio-economics should be developed. Such models should help identify
the complex causalities in terms of direct and indirect effects between ICT and
travel, controlling the confounding factors (Mokhtarian 2003).