Content uploaded by Karel Martens
Author content
All content in this area was uploaded by Karel Martens on Feb 16, 2016
Content may be subject to copyright.
1
WHY ACCESSIBILITY MEASUREMENT IS NOT MERELY AN OPTION,
BUT AN ABSOLUTE NECESSITY
Karel Martens
Faculty of Architecture and Town Planning, Technion - Israel Institute of Technology, IL
Institute for Management Research, Radboud University Nijmegen, NL
E-mail: cjcmmartens@gmail.com
Book chapter to be published in:
N. Punto and A. Hull (eds.) (2016) Accessibility tools and their applications, Routledge.
Please refer to this chapter as follows:
Martens, K. (in press) Why accessibility measurement is not merely an option, but an absolute
necessity. In: N. Punto and A. Hull (eds.) Accessibility tools and their applications, Routledge.
2
Abstract: Many authors and advocates have argued over the past decades that the traditional
mobility-centered approach to transportation planning needs to be replaced with an accessibility-
centered approach. But while the interest in accessibility is on the rise, in practice transportation
(and land use) systems still tend to be evaluated using traditional mobility-centered performance
indicators, like the level-of-service criterion or travel time savings. The aim of this chapter is to
provide an argument that the assessment of accessibility is not merely an option, but an absolute
necessity. The paper starts by distinguishing three, partly overlapping, goals of transportation
planning: activity participation, economic development, and environmental quality.
Subsequently, the interrelationship between accessibility measurement, accessibility indicators,
and each of these goals is explored. That exploration leads to the conclusion that the
measurement of accessibility is not of intrinsic value if policy makers seek to promote economic
development or environmental quality. While accessibility is related to both goals, other
performance indicators may be more suitable if decision-makers are interested in advancing
either goal. The analysis results in a radically different, and perhaps counter-intuitive, conclusion
regarding the goal of activity participation: if decision-makers are interested in the goal of
activity participation, it would be fundamentally wrong to measure a transportation (and land
use) system’s contribution to actual activity participation. The goal of activity participation
requires measurement of accessibility, as only by doing so it is possible to account for persons’
constitutive interests in the range of states they can achieve. Hence, the conclusion that
accessibility measurement is not merely an option but an absolute necessity, if decision-makers
want to take the constitutive interests of persons seriously in the design of transportation
systems.
3
Keywords: accessibility, evaluation, activity participation, economic development,
environmental quality
1. INTRODUCTION
Over the past decade, there has been an increasing interest in the evaluation of transport policies
and investments from the perspective of accessibility. The call to move away from mobility as
the prime performance indicator of transportation (and land use) planning towards accessibility
has been based on the argument that access to desired destinations is what really counts for the
users of that transport system (Levine, this book). Traditional mobility-centered performance
indicators, like the level-of-service criterion or travel time savings, do describe important
qualities of the transport network, but only capture one of the components that determine
accessibility.
Following this line of argumentation, the accessibility paradigm has gained in importance
and accessibility is now by many considered as the key performance criterion to assess the state
of the transport (and land use) system (e.g., Bertolini, le Clercq et al. 2005; Martens 2012).
Levine (this book, p. ##, emphasis in original) takes this argument a step further when he claims
that “more accessibility is inherently preferred over less” and that promoting accessibility does
not require further justification in terms of accessibility’s contribution to other desirable goals.
That claim warrants supportive argumentation, as it is by no means clear why accessibility is
inherently a desirable goal. After all, it could be argued that the true value of accessibility only
emerges if accessibility is “consumed”, that is, if people translate accessibility into actual trip-
making. Moreover, as Sager rightfully argues, it is unlikely that high levels of accessibility will
4
be sustained if little use is made of the potential it offers (Sager 2005). So the question to be
answered is why it would be sufficient for transportation planning to focus on enhancing
accessibility. The aim of this chapter is to answer this question, by exploring the ‘desirable
goals’ that have been typically ascribed to transportation planning. By answering this question, a
powerful argument will be developed in favor of measuring accessibility. By doing so, the
chapter develops a strong rationale for the accessibility analyses presented in this book, as
virtually all of these analyses focus, implicitly or explicitly, on one goal of transportation
planning where accessibility measurement is of key importance: activity participation.
The chapter is organized as follows. In Section 2, I will explore which ‘desirable goals’ have
typically been ascribed to transportation planning. I then briefly discuss the role of accessibility
indicators in policy formation (Section 3). In Section 4, I explore the relationship between
accessibility measurement, accessibility indicators, and each of the goals of transportation
planning. I then turn to exploring what this might mean for the measurement of accessibility. I
end with some concluding remarks (Section 5).
2. GOALS OF TRANSPORTATION POLICY AND PLANNING
Shiftan et al. (2007) define transportation planning as the field of governance intervention that
aims to ensure the effective and efficient movement of people and goods. This definition
accurately captures the aim of transportation planning as a technical exercise, but it fails to
specify the underlying rationale for transportation planning. As widely agreed upon, movement,
whether of people or goods, is rarely an end in itself, but typically a means to some other end
(e.g., Cervero, Neil et al. 2001; Levine, this book). In order to understand if accessibility is
5
inherently desirable, as claimed by Levine, we thus have to start by exploring these ends of
movement.
Depending on the perspective, two distinct ends can be distinguished. First, from the
perspective of a person, effective and efficient movement is important because it enables the
person to participate in activities located throughout space. All else being equal, the
enhancement of the possibilities for movement enables persons to choose from a wider range of
locations dispersed in space to carry out their activities. Persons may make use of these increased
possibilities for a variety of reasons, such as obtaining a more suitable, or better paying, job,
choosing from a wider range of consumer goods available at a larger range of locations, or more
regularly visiting friends or family located at a distance. From this perspective, the rationale of
transportation planning is thus to enable, maintain and, if possible, enhance people’s possibilities
to engage in out-of-home activities.
Second, from the perspective of a business or an organization, the effective and efficient
movement of people and goods is crucial to obtain inputs for, and trade outputs generated by, the
production process and, thus, for their (economic) performance (e.g., Banister and Berechman
2000). The terms “business”, “organization” or “firm” are here interpreted in their broadest
sense, to include for instance also services and non-profit organizations. Likewise, the term
“production process” should be understood to encompass any kind of process using inputs like
labor, capital or raw materials, producing any kind of outputs, including non-tangible services.
From the perspective of the firm so understood, the augmentation of movement of people implies
a larger labor market and thus more options to attract more suitable personnel, as well as the
possibility to provide services to a larger set of potential clients dispersed over space. An
increase in the effectiveness and efficiency in the movement of goods also implies that
6
businesses can attract supplies from a larger pool of potential suppliers and can serve a larger
market area with their goods. The enhancement of transportation systems thus enables businesses
and organizations to take advantage of economies of scale and scope. Taken together, effective
and efficient movement is a crucial prerequisite for the economic functioning of cities, regions
and countries, and the augmentation of transportation systems may lead to economic growth.
These two perspectives thus lead to the identification of two ends of movement: activity
participation and economic development. These ends are often perceived as the rationales, or
meta-goals, of transportation planning and policy. In Levine’s terms, activity participation and
economic development are seen as the desirable goals to which movement, or accessibility, are
to contribute. Clearly, this is not a new insight, but the explicit formulation of these ends is
important for the argument developed in the remainder of the paper.
In the literature, often a third goal of transportation planning and policy is mentioned: the
environment or environmental quality. Clearly, however, the improvement of environmental
quality is not an end of movement. As underscored by many authors, the movement of goods and
people has tremendous impacts on the environment, in terms of air, water, soil and noise
pollution, including transportation’s contribution to greenhouse gas emissions. In light of these
tremendous impacts, it may come as no surprise that interventions in the transportation system
are often presented as a means to improve environmental quality. But if improvement of
environmental quality were interpreted literally as a goal of transportation planning, then the
most effective transportation policies would probably consist of a systematic reduction in the
possibilities for (motorized) movement. Clearly, this cannot be the goal of transportation
planning. On a closer look, it is therefore more appropriate to perceive the environment or
environmental quality as a constraint for transportation planning, as Owens (1995) has already
7
pointed out. That is, transportation planning would seek to ensure the effective and efficient
movement of people and goods within a set of environmental constraints. As Owens formulates
it: “sustainability may require that we meet the imperatives of environmental protection while
maximising accessibility within these constraints” (Owens 1995, p. 48). Depending on the
severity of the constraints, and available (technological) solutions, the effective and efficient
movement may have to be restricted – or can still be enhanced. But the meta-goals of
transportation planning – activity participation and economic development – would be guiding
what kind of measures may be considered to achieve environmental goals.
More recently transportation planning and policy has been linked to another goal, in part
due to the attention the EU has given to the concept in recent years (European Commission
2004): the promotion of social cohesion. However, on a closer look, the notion of social cohesion
can be seen as a particular way to analyze the relationship between transportation, activity
participation and economic development. The link between social cohesion and economic
development is clear in most of the economic literature, which typically defines social cohesion
in terms of the level of convergence in regional welfare levels (López, Gutiérrez et al. 2008).
Lopez et al. (2008, p. 279) also link social cohesion to activity participation by expanding the
notion of social exclusion, noting that “disparities in the access to opportunities can be seen as an
obstacle for cohesion, and transport as a tool to reduce these disparities”. Social cohesion
conceptualized in these ways thus seems to relate to both activity participation and economic
development. More precisely, the concept of social cohesion loosely links both economic
development and activity participation to notions of equity, fairness and justice, as Lopez et al.
(2008) rightfully point out. That is, social cohesion does not add a goal to transportation
planning, but qualifies the goals of activity participation and economic development. It
8
underlines that transportation policy and planning cannot only be concerned with the overall
contribution to activity participation or economic development, but has to address the
distribution of that contribution over regions and people. The notion of social cohesion implies,
so to say, a normative qualification of both goals of transportation planning (see also Martens
2012).
Based on the above, I argue that transportation planning and policy has two, partly
overlapping, goals: activity participation and economic development. Clearly, both goals directly
relate to the fundamental relationship between transportation and space: the inherently spatial
nature of society implies that movement is a conditio sine qua non for human existence. This
fundamental relationship is captured by both goals. The protection or improvement of the
environment should be perceived as a condition, or set of constraints, within which movement
can be ‘allowed’, rather than as a goal. Social cohesion, in turn, implies a qualification of the two
goals of transportation planning and policy.
In the next section, I will explore whether it is indeed sufficient to measure accessibility
in light of these goals of transportation planning, as Levine claims, or whether it is necessary to
explicitly measure to what extent the underlying goals have been achieved. I will explore this
claim for the two goals of transportation planning (activity participation and economic
development), as well as for the environmental ‘goal’ of transportation planning. The latter is
included, because of its prevalence in actual practice and its use in daily parlance.
3. ACCESSIBILITY MEASUREMENT AND ACCESSIBILITY INDICATORS
The ‘accessibility turn’ that can be discerned in research and, somewhat reluctantly, in practice,
calls for the measurement of accessibility as a crucial input for transportation planning as well as
9
land use planning (e.g., Levine and Garb 2002; Bertolini, le Clercq et al. 2005). As discussed
above, Levine adds to this claim, suggesting that the measurement of accessibility is of relevance
because accessibility is inherently desirable. Clearly, whether the former ‘weaker’ claim or the
latter ‘stronger’ claim is accepted, accessibility measurement and the development of
accessibility indicators is of crucial importance. In order to answer the question posed at the
outset, a closer look at the characteristics of indicators is of importance.
Multiple accessibility measures have been developed in the literature, including
infrastructure-based, location-based, and utility-based measures (e.g., Black and Conroy 1977;
Handy and Niemeier 1997; Geurs and van Wee 2004). Every type of accessibility measure will
result in one of more accessibility indicators. As Feitelson and Chenoweth (2002, p. 264) stress,
indicators “are instruments that are used for communicating key information about key systems
in a simplified form to policy makers and the general public”. Building on a number of authors,
Geurs and Van Wee (2004) state that the suitability of accessibility indicators depends on four
criteria: (1) theoretical basis, (2) operationalization, (3) interpretability and communicability, and
(4) usability for evaluation. In order to determine whether transportation planning can suffice by
analyzing accessibility, or should also analyze the underlying desirable goals, it is of importance
to briefly address two of these criteria: theoretical basis and usability.
Geurs and Van Wee (2004) stress that the theoretical soundness of an accessibility
indicator depends on the extent to which the indicator captures the phenomenon of accessibility
in a proper way, accounting for the various components that shape accessibility (notably the
transport system, land use patterns, temporal constraints, and individual characteristics). They
thus focus on the measurement of accessibility itself. Feitelson and Chenotow (2002, p. 264)
stress that indicators “must be highly correlated or causally related to the complex issues they are
10
to portray”. From this perspective, the value of an indicator is not only determined by the
theoretical soundness of the indicator itself, but as much by its ability to help decision-makers
and others gain insight into a particular issue (see also Sen 1991, p.17). Feitelson and Chenotow
thus stress the relationship between the indicator and the goals of a policy. This argument goes
beyond the criterion of usability (see below), but emphasizes another part of the cause-effect or
action-reaction chain surrounding accessibility. Most literature on accessibility measures stresses
the factors that shape accessibility, but tends to ignore the relationship between accessibility and
the purposes, or underlying goals, of accessibility, as discussed in the previous section (to be fair,
Geurs and Van Wee do address this dimension, but under the heading of usability) (Figure 1). As
suggested by Feitelson and Chenotow, if an accessibility index is to be used for policy purposes,
the theoretical soundness of an accessibility indicator should first and foremost be determined by
its relationship with the goals of transportation planning: activity participation, economic
development, or environmental quality. We can now specify in more detail the ‘strong’ claim
made by Levine: if accessibility is inherently desirable then measuring accessibility is equated
with the measurement of the policy goal. Hence, there would be no need to explore the causal
connection with underlying ‘desirable’ policy goals. I will return to this claim below.
Insert Figure 1 approximately here
As Geurs and Van Wee (2004) rightfully point out, the usability of an accessibility
indicator depends on the purpose for which, and the context within which, the indicator is used.
They explicitly distinguish between accessibility evaluation from the perspective of activity
participation (denoted ‘accessibility as a social indicator’) and economic development
(‘accessibility as an economic indicator’). They thus do emphasis the causal connection between
11
accessibility and other desirable goals, discussed above. Geurs and Van Wee are less explicit,
however, about another fundamental element of the usability criterion. This concerns the
relationship between the information conveyed by an indicator and a decision-maker’s scope for
intervention. If accessibility is measured for policy purposes, it is of the utmost importance to
take this scope into account in the development of indices. That is, the accessibility indicator or
set of indicators derived from accessibility measurement should provide decision-makers with an
understanding of the actual situation that is as closely linked to their possibilities for intervention
as possible. The traditional mobility-centered indicators score especially strong in this
dimension: they provide information on the state of the transportation system, which can almost
directly feed into the decision-making process by transportation planning authorities. For
instance, indicators of travel time losses are not only easy to interpret, but also provide decision-
makers with clear directions regarding possible interventions. The now common criticism on this
near-direct link between the score on this indicator and the type of policy intervention that is
often considered in response, only underlines the importance of this ‘intervention scope’
dimension of indicators. Policy-making requires indicators that clearly “point out the problems
so as to galvanize the decision-makers into action”, as Feitelson and Chenotow (2002, p. 264)
rightfully stress. I have addressed this issue, which is crucial for the measurement of
accessibility, elsewhere and will not return to it in this chapter (Martens 2014).
4. ACCESSIBILITY MEASUREMENT, ACCESSIBILITY INDICATORS AND POLICY
GOALS
Let me now further develop the argument and explore if transportation planning could indeed
suffice with measuring accessibility or whether it is necessary to directly measure the extent to
12
which each of the goals of transportation is achieved. In what follows, I will take up the three
‘goals’ of transportation planning one-by-one, starting with the ‘goal’ of environmental quality.
4.1 Environmental quality
One of the reasons that the shift from a mobility-centered to an accessibility-centered assessment
of transportation policies is applauded by academics and practitioners alike, lies in the hopes that
such a shift would contribute to more environmental friendly transportation (and land use)
policies. There are, however, three reasons to reject the measurement of accessibility if we are
interested in the promotion of environmental quality.
The first argument relates to the fact that the improvement of accessibility may come
about in many ways. It may be caused by a simple improvement in travel speeds for motorized
trips, as well as by an increase in land use densities. Clearly, the former is highly likely to be
undesirable from an environmental perspective, while the latter is likely to reduce the
environmental impacts of movement. These examples underscore that there is no inherent
relation between improvement in accessibility and improvement in environmental quality. A
possible solution may lie in a different way of measuring accessibility, as argued by Cheng et al.
(2007). Rather than measuring changes in ‘average accessibility’, changes in accessibility
provided by the car could be compared with changes in accessibility delivered by the ‘green’
modes: public transport, cycling and walking (see e.g., Kwok and Yeh 2004; Benenson, Martens
et al. 2010). Benenson et al. (2010) refer to this as the measurement of the accessibility gap.
Such a comparison could give some estimate of the potential for desirable behavioral changes in
response to transportation (and land use) policies, but it would certainly require an understanding
of the causal relation between a decreasing accessibility gap and an increased ‘green’ travel
13
behavior. It is by no means certain that a clear relationship may be established, certainly if it is
taken into account that the measurement of accessibility gaps may give some indication of mode
choice, but provides little information on changes in average trip distances, which is obviously of
key importance in establishing the environmental impact of interventions in the transportation
system. Hence, an improvement of accessibility, even if measured in terms of a reduced gap
between modes, is not inherently desirable from an environmental perspective (contrary to my
own argument in Benenson, Martens et al. 2010).
A second argument concerns the fact that many interventions in the transport domain
may have little impact in terms of accessibility, but may substantially reduce the environmental
impacts of the transportation sector. Typical examples include policies to promote more
environmental friendly technologies, such as policies promoting electrical cars or less polluting
fuels. If such policies have no impact on the cost of travel at all, they will also have no impact on
accessibility at all. But even if they do change the cost of travel slightly, accessibility is unlikely
to be a good predictor of their environmental benefits.
There is a third argument why accessibility measurement, whatever shape it may take, is
not sufficient if we want to improve environmental quality. This argument runs as follows.
Ideally, one would like governments to compare policies across policy domains, if these policies
aim to achieve comparable goals. Thus, if the goal is to reduce environmental pollution, ideally
one would like to see a systematic comparison of policies that seek to curb pollution by the
transportation sector with policies addressing (heavy) industries or households, for instance.
Such a systematic comparison would require indicators that can be compared across policy
domains, rather than domain-specific indicators. Thus, an indicator of environmental pollution,
such as the level of emission, is strongly preferred over an accessibility indicator, as the latter is
14
only relevant for a limited number of domains. Clearly, even if accessibility was shown to have a
strong correlation with environmental friendly behavior in the transportation domain,
accessibility indicators would do little to enable a cross-domain comparison. While cross-domain
comparison is not very common in policy making, the use of accessibility indicators as a proxy
for transportation policies’ contribution to environmental quality would de facto rule out such a
comparison. This would limit the ability of organizations and citizens to scrutinize government
policies, for instance of organizations advocating the efficient use of government resources. The
use of accessibility indicators would thus hinder rather than enable good governance in the
environmental domain.
Clearly, then, accessibility is not inherently important from an environmental perspective
and accessibility measures and indicators are of limited importance as a metric to assess the
contribution of transportation policies towards improvements in the environment. If anything, it
seems more reasonable to draw on traditionally mobility-centered indicators, such as vehicle
kilometers travelled, than accessibility indicators if one is concerned about the environmental
impacts of transportation interventions. The argument that accessibility (measurement) is of
inherent importance will thus have to be found in (one of) the two proper goals of transportation
planning: activity participation and economic development. I will take up the latter first.
4.2 Economic development
Would it be appropriate to perceive accessibility as inherently desirable from the perspective of
economic development? And, hence, would it be appropriate to measure accessibility rather than
economic development in assessing transportation interventions? Let me try to answer these
questions.
15
At first sight, Levine’s argument seems to hold for economic development. Clearly, if
accessibility is improved, this will enable firms to attract inputs from, and trade outputs to, a
larger area. Indeed, if accessibility is improved through increased speed, a small rise in speed
may lead to a substantial increase in accessibility, due to the geometrical effect of space (the
surface area of a circle increases with the square root of the radius), even if possible changes in
land use are taken into account. Likewise, increases in the density of activities in a particular
region may substantially improve the accessibility experienced by firms already located in that
region, provided that such density increases will not decrease travel speeds too much due to
congestion. Both types of accessibility improvements may thus enable firms and organizations to
take advantage of economies of scale and scope, which have clear economic benefits. The
improvement of accessibility thus is directly related to economic development, suggesting that it
might be sufficient to measure accessibility rather than economic development itself.
Yet, while more accessibility is positive from an economic perspective, it is less clear
how much accessibility contributes to economic development. It is likely that the benefits of
accessibility are subject to some form of diminishing returns. That is, at low initial levels of
accessibility, an accessibility increment is likely to generate substantially more economic
benefits, than an identical increment at high initial levels of accessibility (see also Martens
2006). This differential impact of accessibility increments could perhaps be captured by using an
appropriate accessibility indicator that takes into account the initial accessibility level. For
instance, the ratio between accessibility increment and initial accessibility level could be used as
an accessibility indicator. Such an indicator is especially relevant if comparing policies across
different regions. Yet, it will be less easy to develop an indicator that can take account of the fact
that the relation between accessibility and economic development may well be mediated by the
16
economic structure of a region or by the prevailing institutional conditions (Banister and
Berechman 2001). For instance, some economic sectors may find it more easy to turn improved
accessibility into economies of scale or scope than others. The relationship between accessibility
and economic development is clearly mediated by multiple factors that can hardly be captured by
the accessibility measures and indicators discussed in the literature. Indeed, establishing a
relationship between (small) accessibility increments and economic development has proven to
be elusive (Banister and Berechman 2001, but see Ozbay 2003). However, since the same holds
true for the relationship between mobility and economic development, this observation does not
necessarily challenge the relevance of accessibility measurement and accessibility indicators.
Since the causal connection between accessibility and economic development may expected to
be stronger than the causal connection from mobility towards economic development (rather than
the other way around), quite the opposite position could be defended: lacking an (easy) way to
establish a relationship between improvements in mobility or accessibility and economic
development, accessibility indicators are to be preferred over mobility indicators. Indeed, while
firms may be able to capture benefits of scale and scope by increasing their mobility, it may be
expected that accessibility improvements will often precede a growth in mobility. This is a
powerful argument in favor of measuring accessibility rather than mobility, and for measuring
accessibility rather than the elusive impact of (incremental) accessibility improvements on
economic development.
This claim can also be challenged, however, for two reasons. First, as is the case for
environmental quality, policies to improve accessibility are only one way to stimulate economic
development. For instance, investments in research and development, in higher education, or in
deregulation may also be conducive to economic growth. While the notion of accessibility may
17
bring transportation and land use interventions into a coherent framework, it does little to bridge
the divides between such a broad range of policy domains. By using an accessibility indicator as
a proxy for economic development, it will be impossible to compare such interventions with
improvements in accessibility. As argued before, this is undesirable from a perspective of good
governance. This suggests that an explicit measurement of the expected economic benefits from
accessibility improvements is warranted, e.g. in terms of job growth or increase in GDP, rather
than the measurement of accessibility per se.
Second, and related, even if we are willing to settle for a comparison of interventions
within the transportation (and land use) domain only, the measurement of accessibility would be
of limited importance in the assessment of economic development. After all, even within the
transportation domain, it is reasonable to require that policy interventions deliver a positive
benefit to society. This is certainly a minimal requirement for a policy seeking to promote
economic development (although it may not be the only requirement for such a policy). It would
thus be insufficient to compare policies in terms of their contribution to accessibility and
conclude that the policy with the highest contribution is by definition worthwhile, even if
accessibility improvements would be measured in terms of the ratio mentioned above or a
comparable indicator. Clearly, it would require some form of assessment of the accessibility
improvements in terms that enable a comparison with the cost that have to be incurred to realize
the accessibility improvements. Given the aim of economic development, an assessment along
the lines of standard cost-benefit analysis might be warranted, suggesting that the contribution of
accessibility increments to economic development will have to be monetized (Banister and
Berechman 2000). This in turn implies that, while the relationship between investments in
infrastructure and economic development is difficult to determine, there is no way to avoid the
18
necessity to establish the strength of this relationship. The measurement of accessibility as a
proxy for economic development will simply not do.
Taken together, I conclude that it is not sufficient to merely measure accessibility if we
are interested in promoting economic development. While I agree with Levine that accessibility
is inherently of value from the perspective of economic development, it does not follow that we
can suffice with the measurement of accessibility.
4.3 Activity participation
Let me now explore the relationship between accessibility and the third goal of transportation
(and land use) planning: activity participation. An in-depth of analysis of this relationship will
make it possible to determine whether the measurement of accessibility is indeed sufficient from
the perspective of activity participation or whether measurement of activity participation itself is
necessary.
Let me start the argument with the uncontroversial observation that accessibility is a
prerequisite for activity participation. We may expect that at extremely low levels of
accessibility, a person’s activity participation will be severely restricted. Indeed, while it is hard
to imagine in actual practice, a person with ‘zero’ accessibility is completely confined to her
place of origin and is therefore unable to participate in any out-of-home activities. But at any
level beyond that absolute minimum, a person will be able to participate in at least some
activities. It is also likely that there are strong diminishing returns to increased accessibility. That
is, activity participation may be expected to increase as accessibility improves, but the
relationship will be a concave one (Martens 2006). Like in the case of economic development, a
given accessibility increment will have a much stronger impact on activity participation at a low
19
initial accessibility level than at a high initial accessibility level. Indeed, if activity participation
is defined in terms of only frequency and duration, it may be expected that the marginal benefits
of increased accessibility rapidly decrease. This conclusion is supported to some extent by
studies exploring the notion of stable travel time budgets (e.g., Mokhtarian and Chen 2004; van
Wee, Rietveld et al. 2006), but more explicitly by recent work exploring the link between
accessibility and activity participation (Weis 2012; Weis and Axhausen 2012; Hu and Giuliano
2014). The latter studies find, by and large, that increases in accessibility have only limited
impact on activity participation, if at all. This, empirically supported, observation seems to
suggest that the link between accessibility and activity participation is less straightforward as
may be expected and that the assessment of policy interventions in terms of accessibility may
lead to a substantial over-estimation of the possible impacts on activity participation. In other
words, it suggests that accessibility indicators, however defined, may not be a proper proxy to
assess the contribution of transportation (and land use) policies to activity participation.
Yet, this argument changes if activity participation is defined in qualitative rather than
quantitative terms. Clearly, citizens are not only interested in the frequency and duration of
activity participation, but also in its characteristics. For instance, the value of participation in the
job market is not merely a product of frequency and duration, but also depends on the type and
quality of the work. We can hypothesize that the relationship between accessibility and the
quality of activity participation may be quite strong. Indeed, it may be expected that, all else
being equal, a higher level of accessibility will enable people to derive more benefit from activity
participation. For example, a higher level of accessibility may enable people to obtain a job that
better matches their skills, to engage in leisure activities that more closely match their
preferences, or to visit a more diverse set of family and friends over a given period of time. The
20
benefits of higher accessibility may be even more intangible, for instance if higher accessibility
makes a person more confident about future possibilities for employment, health care, leisure, or
even social contacts (benefits often captured under the heading of ‘option value’ in economic
terms). Clearly, these potential benefits would not be captured by merely measuring the quantity
of out-of-home activity participation in terms of frequency and duration. While the relationship
between accessibility and the quality of activity participation will also be a concave one,
showing a pattern of diminishing returns, it may be expected that benefits may be reaped over a
larger range of initial accessibility levels (Figure 2).
Insert Figure 2 approximately here
The latter argument may suggest that the best way to measure the benefits of
transportation (and land use) interventions would be to directly measure the satisfaction which
people derive from activity participation, instead of using the proxy of accessibility. Or it could
be argued that a particular type of accessibility indicator should be used, i.e. an accessibility
indicator that can measure the satisfaction derived from activity participation (such as utility-
based accessibility indicators). There are, however, strong philosophical arguments to reject such
a proposal. I will not restate all the arguments here (for a brief discussion, see Martens and
Golub 2012), but will summarize the most important ones. First, Sen (2011) has argued
persuasively that welfare or utility is an unsuitable guide to policy, if only because a person may
adjust his expectations to his condition. As Cohen (1993) formulates it, the fact that a person has
learned to live under harsh conditions, and to smile courageously in the face of it, should not
nullify his claim to a better life. In light of this possibility, the measurement of the satisfaction a
21
person derives from activity participation may be expected to lead to skewed results, as the
satisfaction may be strongly shaped by a person’s expectation of what is normal, which in turn
may be heavily determined by the circumstances the person has become accustomed to. Research
actually shows that satisfaction depends much more on expectations, than on the quality of the
product or service being delivered (see Cardozo 1965 and a range of subsequent studies). The
result of measuring satisfaction may thus be that policy attention is diverted to persons with high
expectations regarding activity participation and the related performance of the transportation
system, rather than to persons who have learned to accept a relatively poor transportation system
and the limitations it imposes on the range of activities they can perform on a given day.
Second, the measurement of satisfaction from activity participation negates the
constitutive interest people have in not only the actual state they achieve (the satisfaction they
derive from of a particular set of activities), but also, and perhaps even more so, in the range of
states they could achieve. The argument here is that “doing x and choosing to do x are, in
general, not equivalent” (Sen 1988). Sen (1988) gives the example of fasting. The decision to
stop eating can only be described as “fasting” if the person really does have the alternative of
eating more. Indeed, if the opportunity of normal eating disappears, the person may no longer be
able to fast. The loss of the opportunity to eat freely is thus a substantive loss even for the person
who chooses not to eat but instead to fast. The general point here is that persons have a
constitutive interest in the range of states they could achieve. The value of the chosen option
depends not merely on the characteristics of that option, but as much on the range of options
from which it was chosen. For example, it makes a fundamental difference whether a person
visits his grandmother everyday because she is the only family member he is able to visit given
his time and money budget, or whether the same person consistently chooses to visit his
22
grandmother from a range of other family members. The latter person is clearly better off,
because he has the freedom to choose from multiple options and can execute his agency to give
direction to his life. It would be fundamentally wrong to evaluate the situation of both persons
merely based on their actual level of activity participation, as it would fail to take into account
the fundamental difference in their freedoms and in their ability to execute their agency. This line
of argument obviously also holds for other types of out-of-home activity participation, such as
work or leisure.
The importance of accessibility thus does not merely lie in the fact that a person makes
use of it by visiting a particular place or carrying out a particular out-of-home activity, but even
more in the fact that the person has the option to reach a range of places. The value of the former
cannot be appreciated without taking the latter into account. Thus, from the perspective of policy
evaluation it is crucial to measure persons’ possibility to engage in out-of-home activities. The
measurement of the satisfaction derived from activity participation, or for that matter the
measurement of the quantity of activity participation (frequency and duration), would misdirect
the analysis to something that only partially captures the interests of the person. In contrast, by
analyzing accessibility, the potential value of actual activity participation is implicitly also
captured, as the opportunity set will include a person’s choice for a particular activity.
It may perhaps be illuminating to rephrase this argument in terms of the capability
approach. Sen (1990) makes a distinction between so-called “capabilities” and “functionings”.
Functionings represent what an individual has succeeded in being or doing, while the notion of
capabilities captures the range of beings and doings a person could achieve. Translated to the
domain of transportation, functionings can be equated with actual activity participation (‘what a
person has succeeded in doing’), while accessibility captures a person’s capabilities (‘the range
23
of doings a person could achieve’). Sen and Nussbaum, the major proponents of the capability
approach, make the case that evaluation of policy interventions should principally focus on
capabilities rather than functionings, because the capability set reflects a person’s freedom to
choose between different ways of living. Defending the importance of capabilities, Sen refers to
Marx’s call to replace “the domination of circumstances and chance over individuals by the
domination of individuals over chance and circumstances” (Marx 1846 in Sen 1990). By
focusing on capabilities in the evaluative exercise, the person and his agency are given center
stage. Likewise, by focusing on accessibility rather than activity participation, the person is given
center stage in the assessment of transportation (and land use) interventions.
The argument presented here thus leads to a counter-intuitive conclusion: if we are
interested in the goal of activity participation, it would be fundamentally wrong to measure
actual activity participation, whether defined in terms of quantity or quality. The goal of activity
participation actually requires us to measure accessibility, as only this measurement would
account for the range of states a person can achieve. Thus, from the perspective of the goal of
activity participation, and hence from the perspective of the person, accessibility is indeed
inherently important and more accessibility is indeed inherently preferred over less, as Levine
has thus correctly claimed.
The careful reader may have noted that this conclusion was already implicitly present in
the way I have defined the goal of activity participation. In contrast to the goals of economic
development and environmental quality, I have refrained from using terms like ‘improvement’ or
‘increase’. The two reasons why I have done so are now clear. First, as argued at the beginning
of this section, under circumstances typical for developed societies, few improvements in
accessibility, whether brought about by land use or transportation interventions, will actually
24
lead to an increase in the frequency or duration of activity participation across the board. Second,
the goal of transportation planning should not be to increase actual activity participation, but
merely to expand the range of possibilities for activity participation. Both the empirical and the
philosophical argument have led to the ‘neutral’ formulation of the goal of activity participation.
It may be clear that mobility-centered indicators, however defined, cannot replace
accessibility indicators as a way to capture person’s possibilities for activity participation.
Mobility-centered indicators, such as distance travelled or number of trips, clearly provide little
information about the range of opportunities people could choose from. A high level of mobility
if measured in distance travelled may well be the result of an abundance of possibilities
translated into a high level of activity participation, as of a lack of possibilities forcing a person
to travel over long distances to engage in any activity. Even the measurement of the number of
trips provides little information about the choices available to people, as argued at the outset of
this section. The arguments that have been put forward in the literature against the use of
mobility-centered indicators thus particularly hold in relation to the goal of activity participation.
Three further notes are in place, before drawing conclusions.
First, the use of accessibility indicators does enable comparison across policy domains of
interventions to enlarge the possibilities for activity participation. Clearly, the number of policy
domains that explicitly aims to extend these possibilities is limited. In addition to transportation
and land use planning, it includes urban service delivery policies. Accessibility indicators enable
a relatively easy comparison of transportation policies vis-à-vis interventions in the land use
system, as has been argued extensively in the literature. Urban service delivery policies may
25
shape a person’s possibilities to engage in out-of-home activities through the location of services
(a land use policy), the delineation of opening and closing hours, or price setting. Each of these
could be captured with a suitably defined accessibility indicator, making it directly possible to
compare land use, transportation and urban service delivery interventions with one another. Such
accessibility indicators could even be used to assess the impacts of regulations regarding
maximum working hours on possibilities for activity participation, underlining the power of
accessibility indicators in relation to activity participation across policy domains.
Second, it could be argued that an explicit weighing of the monetary costs of
transportation (or land use) interventions and the benefits of improved accessibility is also
necessary in the case of activity participation, as was suggested in relation to economic
development. This is clearly a valid argument, implying that it is indeed necessary to develop an
approach to systematically evaluate the costs against the benefits of interventions seeking to
expand persons’ range of possibilities for activity participation. However, such an approach does
not necessarily require that the benefits of accessibility are translated into monetary terms. For
instance, cost-effectiveness analysis could be used successfully to directly compare the
contribution of various transportation or land use policies to possibilities for activity participation
(see Martens, Di Ciommo et al. 2014). Even if it would be argued that cost-benefit analysis is the
(only) appropriate approach to systematically compare the societal contribution of various policy
options, the measurement of accessibility still remains of crucial importance. After all, above it
was argued that the goal of activity participation requires the measurement of accessibility. The
challenge is thus not to replace the accessibility measurement with a more ‘appropriate’ index as
an input for cost-benefit analysis (as was the case for economic development), but to monetize
the improvements in accessibility. This will require a substantial research effort, as most of the
26
intellectual energy has so far been invested in assessing the monetary value of travel time
savings rather than of accessibility improvements (but see for an application rooted in notions of
welfare and satisfaction, Geurs, Zondag et al. 2010; Nahmias-Biran, Sharaby et al. 2014). The
notion of diminishing returns, discussed above, would obviously have to play a crucial role in
this monetizing exercise, as I have argued elsewhere (Martens 2006).
Finally, note that the focus on accessibility does not imply that the measurement of
activity participation itself is completely meaningless. Since the measurement of accessibility in
all its dimensions is extremely complex, as the various chapters in this book underline, it is by no
means certain when an accessibility indicator adequately captures the range of possibilities for
out-of-home activity participation. In the development of appropriate accessibility indicators, it
may thus be very valuable to also measure actual activity participation or even the satisfaction
people derive from activity participation. Both can serve as an indication – no more, but also no
less – of the situation of different (groups of) persons in terms of the range of possibilities for
out-of-home activity participation. However, since different people may substantially differ in
the extent to which they translate accessibility into activity participation, and processes of self-
selection may be at work, the measurement of activity participation should never be interpreted
in isolation from the appropriate assessment of accessibility.
5. CONCLUSION
Policy-making requires indicators that clearly “point out the problems so as to galvanize the
decision-makers into action”, as Feitelson and Chenotow (2002, p. 264) have rightfully stressed.
In this chapter, I have explored whether accessibility indicators can indeed point to the problems
that transportation (and land use) planning seek to address. For this purpose, I have distinguished
27
three ‘goals’ of transportation planning: activity participation, economic development, and
environmental quality. Subsequently, I have explored the interrelationship between accessibility
measurement, accessibility indicators, and each of these goals. That exploration has led to the
conclusion that the measurement of accessibility, operationalized whatever way in accessibility
indicators, is not of intrinsic value if policy makers seek to promote economic development or
environmental quality. Indeed, indicators drawing on the mobility-centered approach to
transportation may do a better job in “galvanizing decision-makers into action” if the quality of
the environment is at stake. Accessibility measurement is more intrinsically related to economic
development, but an adequate assessment of interventions in the transportation system cannot
suffice with the assessment of accessibility alone – it would require a systematic analysis of
accessibility’s impacts on economic development itself.
The exploration has resulted in a completely opposite conclusion regarding activity
participation. Based on a number of arguments, I have concluded that the measurement of
accessibility is important in and of itself if policy makers are interested in the goal of activity
participation. That claim has been formulated in even stronger terms: if policy makers are
interested in the goal of activity participation, it would be fundamentally wrong to measure
actual activity participation. The counter-intuitive conclusion is that the goal of activity
participation requires decision-makers to measure accessibility, as only by doing so it is possible
to account for persons’ constitutive interests in the range of states they can achieve. Indeed, from
the perspective of activity participation, and hence from the perspective of the person, the
increase in accessibility is inherently desirable, and not merely because of its predictive power in
terms of the quantity or quality of out-of-home activity participation, as correctly suggested by
Levine (this book). ‘Objective social science’ need not be uncomfortable with that claim.
28
It has taken a lengthy argument to make this point. Proponents of accessibility measures,
among which the authors of the various chapters of this book, may perhaps find the argument of
little help, as they are already convinced about the need to measure accessibility rather than
mobility. However, there is no widespread agreement yet in the academic or professional
community about the role of accessibility measurement in transportation (and land use) planning.
It has been the aim of the chapter to deliver a powerful argument why accessibility measurement
is not merely an option, but an absolutely necessity, if one is interested in the contribution of
transportation (and land use) interventions to out-of-home activity participation.
ACKNOWLEDGEMENTS
Some of the ideas presented in this chapter have been developed within the framework of the EU
funded COST Action ‘Transport Equity Analysis’ (TUD 1209) (http://teacost.eu/). Some of the
ideas presented in this chapter are discussed more elaborately in a forthcoming book by the
author, titled ‘Transport justice: designing fair transportation systems’, to be published by
Routledge in 2016.
REFERENCES
Banister, D. and J. Berechman (2000). Transport investment and economic development,
Psychology Press.
Banister, D. and Y. Berechman (2001). "Transport investment and the promotion of economic
growth." Journal of Transport Geography 9(3): 209-218.
Benenson, I., K. Martens, et al. (2010). "Measuring the Gap Between Car and Transit
Accessibility: Estimating Access Using a High-Resolution Transit Network Geographic
Information System." Transportation Research Record: Journal of the Transportation
Research Board(2144): 28-35.
Bertolini, L., F. le Clercq, et al. (2005). "Sustainable accessibility: a conceptual framework to
integrate transport and land use plan-making. Two test-applications in the Netherlands
and a reflection on the way forward." Transport Policy 12(3): 207-220.
29
Black, J. and M. Conroy (1977). "Accessibility measures and the social evaluation of urban
structure." Environment and Planning A 9: 1013-1031.
Cardozo, R. N. (1965). "An experimental study of customer effort, expectation, and satisfaction."
Journal of marketing research: 244-249.
Cervero, R., J. S. Neil, et al. (2001). Transportation Planning. International Encyclopedia of the
Social & Behavioral Sciences. Oxford, Pergamon: 15873-15878.
Cheng, J., L. Bertolini, et al. (2007). "Measuring Sustainable Accessibility." Transportation
Research Record 2017: 16-25.
Cohen, G. A. (1993). Equality of what? On welfare, goods, and capabilities. The quality of life.
A. K. Sen and M. C. Nussbaum. Oxford, Oxford University Press: 9-29.
European Commission (2004). "A new partnership for cohesion :convergence competitiveness
cooperation. Third report on economic and social cohesion. Office for Official
Publications of the European Communities, Luxembourg. ."
Geurs, K., B. Zondag, et al. (2010). "Accessibility appraisal of land-use/transport policy
strategies: More than just adding up travel-time savings." Transportation Research Part
D: Transport and Environment 15(7): 382-393.
Geurs, K. T. and B. van Wee (2004). "Accessibility evaluation of land-use and transport
strategies: review and research directions." Journal of Transport Geography 12(2): 127-
140.
Handy, S. L. and D. A. Niemeier (1997). "Measuring accessibility: An exploration of issues and
alternatives." Environment and Planning A 29(7): 1175-1194.
Hu, L. and G. Giuliano (2014). "Poverty concentration, job access, and employment outcomes."
Journal of Urban Affairs.
Kwok, R. C. W. and A. G. O. Yeh (2004). "The use of modal accessibility gap as an indicator for
sustainable transport development." Environment and Planning A 36(5): 921-936.
Levine, J. and Y. Garb (2002). "Congestion pricing's conditional promise: promotion of
accessibility or mobility?" Transport Policy 9(3): 179-188.
López, E., J. Gutiérrez, et al. (2008). "Measuring regional cohesion effects of large-scale
transport infrastructure investments: an accessibility approach." European Planning
Studies 16(2): 277-301.
Martens, K. (2006). "Basing transport planning on principles of social justice." Berkeley
Planning Journal 19: 1-17.
Martens, K. (2012). "Justice in transport as justice in access: applying Walzer's 'Spheres of
Justice' to the transport sector." Transportation 39(6): 1035-1053.
Martens, K. (2014). Accessibility and potential mobility as a guide for policy action. Paper
presented at the AESOP Conference, 9-12 July 2014, Utrecht, the Netherlands.
Martens, K., F. Di Ciommo, et al. (2014). Incorporating equity into transport planning: utility,
priority and sufficiency approaches. Paper presented at the PANAM conference, 11-13
June 2014, Santander, Spain.
Martens, K. and A. Golub (2012). A justice-theoretic exploration of accessibility measures.
Accessibility Analysis and Transport Planning: Challenges for Europe and North
America. K. T. Geurs, K. J. Krizek and A. Reggiani. Cheltenham, Edward Elgar.
Mokhtarian, P. L. and C. Chen (2004). "TTB or not TTB, that is the question: a review and
analysis of the empirical literature on travel time (and money) budgets." Transportation
Research Part A: Policy and Practice 38(9-10): 643-675.
30
Nahmias-Biran, B. H., N. Sharaby, et al. (2014). "Equity Aspects in Transportation Projects:
Case Study of Transit Fare Change in Haifa." International Journal of Sustainable
Transportation 8(1): 69-83.
Owens, S. (1995). "From ‘predict and provide’ to ‘predict and prevent’?: Pricing and planning in
transport policy." Transport Policy 2(1): 43-49.
Ozbay, K., D. Ozmen-Ertekin, et al. (2003). "Empirical analysis of relationship between
accessibility and economic development." Journal of urban Planning and Development
129(2): 97-119.
Sager, T. (2005). "Footloose and Forecast-free: Hypermobility and the Planning of Society."
European Journal of Spatial Development September 2005(17): 1-23.
Sen, A. (1988). "Freedom of choice: concept and content." European Economic Review 32(2):
269-294.
Sen, A. (1991). "Welfare, preference and freedom." Journal of Econometrics 50(1–2): 15-29.
Sen, A. (1990). Development as capability expansion, in Keith Griffin and John Knight (eds),
Human Development and the International Development Strategy for the 1990s, London:
Macmillan, pp. 41-58.
Sen, A. (2011, originally published 1980). Equality of what? Tanner lectures on human values.
S. McMurrin. Cambridge, Cambridge University Press: 195-220.
Shiftan, Y., K. J. Button, P. Nijkamp (2007). Transportation planning. Edward Elgar,
Cheltenham.
van Wee, B., P. Rietveld, et al. (2006). "Is average daily travel time expenditure constant? In
search of explanations for an increase in average travel time." Journal of Transport
Geography 14(2): 109-122.
Weis, C. (2012). Activity oriented modelling of short-and long-term dynamics of travel
behaviour, Diss., Eidgenössische Technische Hochschule ETH Zürich, Nr. 20346, 2012.
Weis, C. and K. W. Axhausen (2012). "Assessing Changes in Travel Behavior Induced by
Modified Travel Times: A Stated Adaptation Survey and Modeling Approach." disP-The
Planning Review 48(3): 40-53.
31
Transport
component
Land use
component
Temporal
component
Individual
component
Activity
participation
Economic
development
Environmental
quality
Accessibility
Theoretical relation between components and accessibility
Theoretical relation between accessibility and policy goals
FIGURE CAPTIONS
Figure 1 – TWO COMPONENTS OF THE THEORETICAL SOUNDNESS OF AN
ACCESSIBILITY MEASURE [please use full caption as described below]
Figure 1 Two components of the theoretical soundness of an accessibility measure. Most of the
studies on accessibility measurement have focused on the left-hand side of the diagram, but have
ignored the right-hand side of the diagram. The left-hand side of the diagram is based on Geurs
and Van Wee (2004).
32
Figure 2 – RELATIONSHIP BETWEEN LEVEL OF ACCESSIBILITY AND DERIVED
BENEFITS [please use full caption as described below]
Figure 2 The relationship between level of accessibility and the benefits derived from that level
of accessibility in terms of (a) the quantity of activity participation (straight line) and (b) the
quality of activity participation (dotted line). Figure adapted from Martens 2006.
