Conclusions
Accessibility impacts of land-use and transport
changes, for example, those due to policies, are often
evaluated using accessibility measures, which researchers
and policy makers can easily operationalise and
interpret, but which generally do not satisfy theoretical
criteria. We have described a set of theoretical criteria
related to the different components of accessibility from
the perspective of evaluating land-use and transport
changes. In other words, an accessibility measure should
be sensitive to changes in the quality of transport services
(transport component), the amount and distribution
of the supply of and demand for opportunities
(land-use component) and temporal constraints (temporal
component). It should also take individual needs,
preferences and abilities into account (individual component).
These criteria are not considered absolute;
applying the full set of criteria would imply a level of
complexity and detail that can probably never be
achieved in practice. Thus in practical applications,
different situations and study purposes demand different
approaches. However, it is important to recognise the
implications of ignoring one or more of these criteria.
In conclusion, infrastructure-based accessibility
measures, such as average speed on the road network,
are easy to interpret and communicate but are not very
useful for evaluating the accessibility impacts of landuse
and transport policy plans since the measures lack
the land-use component, and temporal and individual
elements. As a result, they may lead to inaccurate or
even misleading results if these shortcomings are not
recognised and described. More complex location- and
utility-based accessibility measures can be considered
effective measures of accessibility, which can also be
used as input for social and economic evaluations. That
is they overcome the most important shortcomings of
infrastructure-based measures, and can be computed
with state-of-the-practice land-use and transport data
and models. Moreover, utility-based measures capture
the valuation of accessibility by individuals, providing a
useful basis for user-benefit evaluations of both land-use
and transport investments. An important remaining
theoretical shortcoming is the exclusion of individuals’
spatial–temporal constraints typically included in person-
based accessibility measures. Person-based measures
are potentially very useful for social evaluations,
and may also be tied to the utility-based approach,
which opens up the possibility of using them in economic
evaluations. However, person-based measures
still have important disadvantages related to data
availability and complexity, restricting applications to
relatively small regions and subsets of the population.
Furthermore, to date, state-of-the-art, activity-based
transport models have not yet been able to link daily
activity patterns with long-term spatial behaviour of
household and firms, an important shortcoming in
evaluations of land-use and transport investments.