5.2 Sensitivity of the results
In this paper assumptions had to be made regarding several issues. In this section we will
discuss the sensitivity of the results to changes in the most important ones. In our opinion
these are the assumptions on direct and indirect land use, on (regarding indirect land use) the
percentage of the area that would be built-on should restrictions on land use be removed and
on land prices. For each of these assumptions we will have to answer two questions. First,
how large is the uncertainty on the values that were assumed, and second, how sensitive are
the results to changes in these values (expressed, when possible, in percentage change of the
results as a result of a one-percent change in the assumed values).
Unfortunately, concerning the direct and indirect land use of some types of infrastructure we
cannot be very precise about the possible variation in the assumptions. The results on trains
and aeroplanes will not be very sensitive to the assumptions however, mainly because the
assumptions on direct and indirect land use of railways and airports are based on very
accurate and reliable sources. As already mentioned in Section 3, the direct and indirect land
use of roads and waterways are based on minimum design demands. Therefore, the costs
associated with them can be considered minimum values. How much the land use based on
minimum design demands differs from actual land use by roads and waterways is unclear.
The area that would be built on should the restrictions causing indirect land use be removed
was assumed to be 50 percent of the total restricted area inside the built-up area. Outside the
built-up area this figure was assumed to be 20 percent. Regarding the second, it is clear that
the results on costs of indirect land use change proportional to a change in the assumed
values, i.e. a decline in the percentage from 50 to 40 percent would result in a downfall in
results of (50-40)/50 = 20 percent. Therefore, the more actual values differ from the assumed
values, the more sensitive the results on costs of indirect land use are. Unfortunately, we
again do not have information on the possible variation in these figures, although in our
opinion they are likely to be small. However, even if this variation is high, since costs of
indirect land use do not contribute much to the total costs of land use, let alone to the total
fixed costs of infrastructure, the sensitivity of absolute total costs of land use to these
assumptions is bound to be small.
Regarding the assumptions on land prices we can be a bit more precise. Let us focus on costs
of direct land use, being the largest cost categories. Regarding the first question, note that we
have taken averages of land prices that were found in the literature and/or provided to us
through personal contact. For direct land use inside the built-up area the land prices varied
from 18 to 27 Euro (with an average of 23 Euro), which implies a possible variation of