in order to save 1 euro/day on housing costs. This result seems to point into the direction of
people being more sensitive to travel costs than to housing costs2.
Beside the monthly cost component, the benefit of having one extra bedroom, of not living in
a big city (on average negative valuation) and/or living in a small town are compared to the
influence of travel time and travel costs. Travel costs seem to be quite important. This may
implicate that people do not want to pay much on commuting. They do not want to spend
(extremely) high costs in order to be able to live at a certain location or to extend the number
of bedrooms. Furthermore, results in general indicate that (at least for the respondents) travel
times are relatively unimportant compared to the location related variables but also in
comparison to travel cost. The relative importance of travel time versus travel cost can be
observed into more detail by computing a value of time (VOT). Values of travel time saved
(VOT) indicate the amount of money people want to pay in order to save a certain amount of
travel time. Therefore, the VOT gives an indication of the importance of travel time in
relation to travel costs. Low values of time for example, indicate that people are relatively
more cost than travel time sensitive. In case of location decisions such a low value of time
could mean that people would prefer a relatively longer commuting time (and maybe
distance) with lower travel costs above a shorter commuting time with higher travel costs. The
VOT is computed as follows:
coeff.traveltime
VOT =---ʌ'-----------* 60 [euro/hour]
coeff .travelcost
In fact two different concepts of the value of time exist: the marginal and the non-marginal
value of time. Most studies focus on the marginal value of time, indicating as formulated
before the amount of money people want to pay in order to save a certain amount of travel
time. This marginal VOT is often indicated by the term ‘value of travel time saved’ (see also
2 These results must be handled with some care. First of all, in order to convert monthly housing cost to cost per
day, one has to know the (average) number of commute trips that are made on a monthly basis. In this case a
multiplication with 20 (5 day working day, 4 weeks/month) has been used, but this choice remains somewhat
arbitrary. In the second place, the set-up of the experiment aimed at making differentiations in monthly cost on
average comparable to travel cost variations. This ‘comparability’ could not be guaranteed before the experiment
started. After the data had been derived for example, the mean trip length was found to be substantial higher than
expected. But, in the same way also monthly housing costs were somewhat underestimated in advance.
However, it is expected that this uncertainty (i.e. attached to these two mentioned aspects) alone cannot lead to
the observed large difference in valuation between housing cost and travel cost.
17