|
"Where in the world do you want to go today?" ran the old Microsoft
ad, in a simple statement indicating the boundless possibilities of
virtual mobility. Online, you can work, shop, chat and do all kinds of
things "in" (so to speak) places that would otherwise be well out of
reach. It seems self-evident that these kinds of online activities
will have an impact on what we travel for, and how often we travel.
There are limits, of course. Faced with a choice of going to the
Maldives, or taking a virtual tour of the islands, the physical visit
is far more attractive to most people. Physical presence is
always going to be the preferred, or necessary, choice for many
activities. But increasingly online activity is preferred where
- it takes the hassle out of travelling or distribution, and/or
makes it more efficient
- it opens up new opportunities for activity that otherwise would
have been out of reach due to distance or cost.
"Virtual mobility", as the phrase implies, is not really mobility
at all. It refers to undertaking activities that previously would have
required transport to do. And in many cases, the activities undertaken
are simulations of or substitutes for physical activities.
What all this adds up to is that online activities necessarily have
both a spatial and temporal effect on our activities - we do what we
do in different places than previously, and it takes a different
amount of time to do it.
So how much does this make a difference on the roads?
Virtual mobility and flexible work
From the published research, the area where online activity is
making a measurable difference is in the area of telework. There are
many different kinds of telework, but two of the varieties have been
the subject of more intensive investigation. These are
home-based teleworking, and working at local telecentres.
Here is a sample of findings from the research:
- Teleworkers in the BT Workabout scheme reduced their travel 93
miles reduction per week for car users, 143 miles per week for rail
users
- Teleworkers at the Dutch Ministry of Transport made 17 per cent
fewer trips and cut peak-hour car travel by 26 per cent. Household
members also appeared to travel less than before
- Teleworkers in the Greater Munich Area studied in the MOBINET
project found that they reduced work trips by 43% and all trips by
19%
- Users of the California Neighbourhood Telecenters Project
demonstrated a 17% reduction in vehicle miles travelled per week
- Users of Surrey County Council’s pilot telecentre reduced the
length of their commute journey on average by 19%, and duration by
36%. Teleworking has now been introduced across the Council in the
Surrey Workstyle programme.
These headline figures mask a variety of effects in terms of
redistribution. The trips that do happen are often to different
places, at different times and, within the household, often undertaken
by different people.
It is worth noting that in most cases telework is not full-time.
Across the teleworking population at large, it seems that around 1.5
days per week (6 days per month) is emerging as a typical figure.
Although averages are higher at high tech or telecoms companies like
BT, for most individuals a day and a half is a comfortable level. It
remains to be seen whether this will change over time with improved
technologies, greater familiarity and changes to management attitudes
and workplace culture.
Looking at a wide range of studies, it is possible to generalise as
follows:
- average range 20-50 miles (person mile travelled) are saved per
teleworking occasion
- this leads to a typical range of savings of 1300-3500 miles per
teleworker per year (given the incidence of teleworking in the
studies at 1, 1.5 or 2 days per week).
It is pretty much established, then, at the individual and
household levels, that there is a significant net traffic reduction
from teleworking even when compensatory trips are taken into account.
Wider impact on the roads
A number of studies have attempted to aggregate these kind of
findings to give a picture of the effects at city, regional or
national level, and potential growth trends.
A kind of orthodoxy has been emerging, where rates of teleworking
(numbers of days per week), average distances travelled and telework
penetration rates (i.e. number of people teleworking/number of people
who could telework) are combined to give an aggregate figure. However,
this kind of approach contains many assumptions, particularly in the
attempts to find how many people do or could telework. This is
typically based on some kind of estimate about the numbers of
"knowledge workers", or guesstimating the numbers of teleworkers as a
subset of white-collar workers, and elaborating growth trends from the
growth of the services sector.
One of the key assumptions in looking at potential ceilings in the
number of teleworkers is the assumption that patterns of early
adoption amongst mainly higher earning professional and managerial
male workers are determinative of future trends. A paradoxical effect
of such assumptions is that it may over-estimate the traffic reduction
potential per teleworker (as higher earning men tend to commute longer
distances) while under-estimating the total traffic reduction
potential by not seeing the telework potential within lower paid and
lower-skilled jobs.
In this kind of estimate, there is also often very dubious use of
data that is collected for other purposes. However, in some
countries such as the UK and many other European countries, data is
now collected in the Labour Force Survey (LFS) that provides a
consistent if perhaps conservative way of measuring the numbers of
people teleworking from home for at least one day per week.
Based on figures from the British LFS, on any day there are about
1.1 million people teleworking. However, while this adds up to
quite a high number of eliminated commute round-trip journeys, in
percentage terms it currently makes only a small dent in total traffic
figures. Growth in the number of teleworkers is also outstripped by
both employment growth (there are far more new non-teleworkers than
teleworkers in the workforce) and by total traffic growth for all
purposes. It is also important to note that one telework occasion
doesn't simply translate to one eliminated commute round-trip, as
teleworkers may also "time-shift", i.e. shift their commute journey to
a different time of day.
So there will need to be significant growth in teleworking, both in
total numbers and the number of days each person teleworks, to begin
to make a significant impact.
Rebound effects
There has been much debate about complementary or generative
effects of telework and other online activities on transport. These
tend to focus on:
- Trip-chaining/multi-purpose trips: commuters often
combine other activities with their commute journey, like shopping,
school escort, etc. So when the commute trip is eliminated, a
substitute trip for such purposes may still be needed.
- Extra business trips: for teleworkers who visit clients,
eliminating the commute trip and trips back to the office creates
time for further trips to clients
- Convenience of car at home – for teleworker and
household: when the car is not used for commuting, it is up for
grabs by the other household members for their use. Or if the
teleworker is usually a public transport user, he/she may be tempted
to use the car to, say, drive the kids to school
- Relocation potential: a lot of literature assumes that
teleworking will enable people to live further from their offices,
and make fewer but longer journeys
- Latent demand: other drivers occupy the roadspace vacated
by teleworkers in congested areas.
A figure of around 50-60% "clawback" of travel savings through
newly generated trips is often cited by commentators, largely on the
basis of papers from a group of American researchers.
However, close inspection of the papers concerned shows that these
figures are almost entirely speculative, however plausible the rebound
effects might be. There are no measured studies as yet that
demonstrate such effects. Some measured studies, on the contrary,
indicate opposite trends, i.e. that households as a whole contract
their "activity spaces" and their total travel when telework takes
place. There is a lot more work to be done in this field.
A key question is "how far to cast the net when looking for rebound
effects?" The following diagram outlines one way to segment the
various kinds of potential rebound effects from online activities:
At the moment research has only penetrated in to the
first order rebound effects (rebound 1) and, in a few studies, into
the second level in terms of some household effects. For level 3 and
beyond there is an urgent need to capture new data and to develop new
modelling techniques before any conclusions can be drawn.
For the column marked "Rebound X", there must be
serious questions about the coherence of the speculation, as one
becomes further removed from the original online activity in question,
and other variables are involved - not to mention questions of
causality.
For example, one study modelling the transport effects
of e-commerce in the computer industry, while showing a net transport
reduction, cancels these out by factoring in the following rebound
effect. As online businesses are around 10% more efficient, it
suggests, they can manage with around 10% fewer workers. These
workers are going to go out and do other jobs, hence create more
traffic, etc.
Is this a legitimate rebound effect to factor in? At
best, it is a complicated one, as it attributes their travel behaviour
in one company as a rebound effect of their not working for another
company! It also assumes that in their new situation that work
arrangements do nothing to minimise their need to travel. At some
point one needs to draw the line and say "This far and no further" in
attributing rebound effects.
For individuals and companies, such rebound effects
may be academic, someone else's problem. The raw travel savings
are sufficient incentive to change the way they work.
Transport effects from e-commerce and e-services
Research into the transport effects of e-commerce and
e-services are very much in their infancy, as are the technologies and
applications themselves. We will reserve more detailed discussion of
these for a future article. For the moment we can summarise the
state of the research as for e-commerce follows:
- General finding: delivery trips trips increase, in particular
lights good vehicle deliveries, while personal shopping trips
decrease. But studies tend to be based on speculative modelling,
rather than observation.
- There are major variations in estimates of reductions in
shopping trips, depending whether researchers emphasise continuing
need to buy other goods, or are optimistic about the range of goods
that will become available and user uptake.
- There are major variations in estimates about distribution
trips, depending on weight given to potential increases in
e-business efficiency and existing carries having the ability to
absorb increases into existing delivery patterns
- A few recent studies using life-cycle analysis for energy use in
traditional versus online retail indicate a similar total
energy use, but energy consumption from transport reduces
- Some expect e-commerce greatly to increase air freight, as
consumers and industries source products from further afield. Bad
for the global environment, but good news for surface transport
policy, if local sourcing decreases??
For the impact from e-services there is an absence of research that
deals with real numbers as opposed to theorising. But it is worth
noting that the UK government is allocating £6 billion to developing
e-government at all levels, and this is largely based on eliminating
physical contact between bureaucracies and citizens - which implies a
substantial transport reduction effect. One major difference between
this and e-commerce is that at the end of the online transaction, much
e-commerce still requires the distribution of a physical product,
whereas e-government and other e-services mostly involve a
"dematerialised" end-product.
Conclusions
We can summarise the state-of-the-art as follows:
- Home-based and telecentre-based teleworking have a significant
direct travel reduction effect, which is confirmed in empirical
studies
- The reduction effect endures when household activity is also
measured
- There are significant redistributive effects – destinations,
time, mode and journey length
- Other forms of teleworking are under-researched, but business
case studies indicate significant travel savings for the
organisations concerned
- Research into the transport effects of e-commerce and e-services
is at yet in its infancy
- Most figures given for wider rebound effects are speculative
- Transport studies tend to be weak on variations in technology
and management in telework and e-commerce implementations, so
studies do not necessarily compare like with like.
The upshot of this is that research does support the idea of using
telework as a tool to reduce travel - but also that there is a lot
more research that needs to be done before we see the full picture.
 |
|
