Understanding the links between transport implementation, underlying assessment practices and social exclusion

(1)

1

University of Amsterdam

Graduate School of Social Sciences

Research Master in Urban Studies Programme

Understanding the Links Between Transport

Implementation, Underlying Assessment Practices

and Social Exclusion

Author: Pablo Moran Figueroa.

Student number: 10635246

Supervisor: Prof. Dr. Luca Bertolini

Second reader: Prof. Dr. Marco te Brömmelstroet,

(2)

2

In memory of Mauricio Aguirre Munizaga

(3)

1

I. Acknowledgements

Words are not enough to express my gratitude to all the people who made this ambitious thesis project possible. My experience at the Research Master in Urban Studies was indeed challenging. However, the enormous support I had from colleagues and lecturers at the programme always kept me going forward. This thesis document is dedicated to all of them, with joy and happiness.

I would especially like to thank my supervisor, Prof. Dr. Luca Bertolini, for his excellent guidance in this project and constant motivation. His role in this project was more than what I could’ve ever imagined. His profound human and professional skills have helped me to find a new start in my professional life. Thanks for this. My gratitude also goes to all the people who participated in this research thesis, making this project a unique experience.

I also want to dedicate this work to my family and friends spread around the world, who would always look for a moment to cheer me up, to bring me a smile. And of course I could not forget about my friends here in Amsterdam. Thanks for all the bike rides and happy moments, and also for helping me those times when things looked more complicated than usual.

Finally, I want to thank Janina, my girlfriend, for making this life so cheerful.

Words are not enough, and probably will never be. Yet in my memory, within myself, all good moments and thanks will be eternal.

Pablo Moran Figueroa Research Master in Urban Studies University of Amsterdam September 2015

(6)

4

II. Summary

The present thesis focuses on understanding the dynamics of mobility, transport and social exclusion. Its core conceptual definitions and operations are encompassed by Amartya Sen’s capability approach (1985/1987/2000/2007) – a distinct theoretical framework which focuses on the relation of various human actions and their environment components, in order to describe a state of deprivation or exclusion. Social exclusion in these respects is conceived as a process rather than a sole definition or measurement system. Instead, a framework of observation of these human actions is provided, in order to understand the different sources and factors that contribute to worsening or alleviating the manifestation of this phenomenon. In this research project, Sen’s framework has been adapted to analyse the dynamics of social exclusion from a mobility and accessibility perspective, through the use of participatory practices, GIS and qualitative assisting methodologies. The main developed methodology in this research is presented as an instrument of analysis, named TEBUPAT (Transport Environment Bottom-up Participatory Assessment Tool).

Moreover, this thesis adds a relevant component of research by also exploring how transport-planning assessment practices engage a perspective on social exclusion and its application or operationalization. This brings into further discussion how assessment practices play a role within transport implementation and/or modification projects, and more specifically, how could these drive results that have an impact in terms of social exclusion. The use of TEBUPAT in two cases of study – Santiago and Amsterdam – provides a first approach to empirical data that could help reveal these relations.

The empirical evidence gathered from both case studies is put side by side with what assessment practices are able to observe in each city. The comparison of assessment practices in both cases reveals trends in which these procedures do not only lack of the capacity to define and operationalize social exclusion, but could be neglecting a refinement of social-impact measures. The found and analysed evidence adds to a larger discussion in academia and transport-planning, with further research focused on two avenues: (1) improving the use of TEBUPAT – by improving the representativeness and theoretical saturation of data – and (2) incurring into how such a tool could be embedded in real transport-planning situations. This latter research avenue would also explore the policy implications behind these practices. It is expected that this triggers a wider theoretical and practical discussion, about the purposes of mobility planning and its social impact, present in the current technical framework of transport-planning practice.

(7)

5

1. Introduction: problem statement and research aim

Issues of mobility and its impact on society have been one of the central topics of sociological literature during the last decades (Hannam et al., 2006). As inferred from Kaufman et al. (2004), mobility has structured the paradigm of modern life, becoming an essential component of everyday activities in urban regions. “Changes in mobility patterns may be at the base of fundamental societal changes, whose consequences for people and territories are only now becoming intelligible” (Kaufman et al., 2004, p.746). However, understanding these emergent and constantly shifting dynamics stands a major challenge. One of such sets of issues where more understanding is called for is the relationship between transport and social exclusion.

The increasing attention to the links between transport systems and social exclusion has resulted in a growing body of literature that suggests a multidisciplinary approach to these matters, as presented by Beyazit (2011), Church et al. (2000), Hodgson & Turner (2003), Kenyon (2003), Lucas (2012), Mc Cray & Brais (2007), Preston & Rajé (2007), Stanley and Lucas (2008) and Stanley & Vella-Brodrick (2009), to mention some relevant contributions. Hitherto, one of the main focuses remains on determining how social exclusion may be properly defined and evaluated, as the concept is characterised by its multidimensional nature (Lucas, 2012). Thus, the manifestation of social exclusion cannot be exclusively linked to transport issues. Yet, “it is […] essential to recognise, the dynamic and relational nature of the exclusionary process where transport is concerned (Lucas, 2012, p.108)”. That is, understanding which factors and processes in transport are key to assess social exclusion. Accordingly, the claims of a more sensitive approach arise from a general critique of how social exclusion is measured or identified – differentiating this concept from poverty (Hodgson & Turner, 2003) and mainly criticising views based (solely) on income-based deprivation indicators, as these seem to provide a limited scope on the phenomenon’s extent (Preston & Rajé, 2007).

As a socio-spatial phenomenon, social exclusion might have different levels of presence throughout a city – different sets of constraining environment conditions in diverse territorial or spatial dimensions and scales (Madanipour, 2002). Transport-implementation processes might have different consequences in these respects. Likewise, people living under less benefited conditions are also expected to be more sensitive to these processes (Lucas, 2012). Changes in their mobility patterns could result in accessibility deprivation of whole communities, especially those ones that, for instance, depend on limited transport options, among other related problems (the presence of other social, economic, land-use and geography factors related to social exclusion). In short, little is known about how the implementation or modification of transport could reshape socio-spatial practices – the way in which such could have an impact in the environment components that structure mobility conditions – and how this could affect people’s condition of social exclusion (McCray & Brais, 2007).

In Europe, transport policy stakeholders have raised their attention to this phenomenon. Yet, in practice, the discourse of social exclusion has been actively integrated only in some few places (Geurs et al., 2009). One of these cases was the experience of the Social

(8)

6 matter of accessibility constraints for informing policy recommendations (Lucas, 2012). But even for such experience the definition of social exclusion remained blurry. In Hodgson’s and Turner’s (2003) words:

“[The] UK transport policy would appear to be developing only partially in a direction set out by the theoretical understanding of the term social exclusion. It is not addressing the issue of participation in the decision-making process of transport operation and management. In fact, in the transport sector, substantial failings in allowing people to participate in the policy process have been acknowledged and those efforts that do exist are often seen as tokenistic or placatory (Bickerstaff et al., 2002) […]” (p. 268).

As we may infer from the latter authors, the debate might be misled by a lack of attention and understanding of the processes that lead to a condition of exclusion, reflected in a lack of participatory mechanisms and the limited involvement of communities. Their focus is in fact an attempt to redefine the epistemological scope on social exclusion within urban transport-planning practices. In other words, thinking differently in terms of how social exclusion may be grasped from a conceptual dimension to its operationalization. This is a reflection that takes an important place within this research project. Withal, for other contexts with no clear assessment of social exclusion – urban transport-planning regimes probably lacking in methodological and instrumental capacities for unfolding the process of social exclusion, from a mobility perspective – the same issues would seem to apply. That being said, the main concern of this research focuses on the present uncertainty to evaluate social exclusion in relation to transport and mobility, also bringing into attention what could be the consequences of transport planning practices that might lack of a systematized approach to this phenomenon. The purpose of this research is therefore to scrutinize how social exclusion is affected by the implementation or modification of

transport projects and the capacity of transport planning regimes to address, observe and assess the process of social exclusion. Yet for examining this chain of events, I claim that it is

first required to understand what particular socio-spatial conditions of the transport

environment foster social exclusion, as highlighted by the aforementioned academic

(9)

7

1.1. Causal model and research questions

The main causal relation in study can be described in a simple X-Y relation model, illustrated in figure 1. Here, the main focus is set on understanding how the implementation or modification of transport may affect communities in terms of social exclusion. This will be done by observing “less benefited areas”, as a means to address a potentially existing condition of social exclusion – structured by a multiplicity of socio-spatial factors – and afterwards scrutinize how such phenomenon could be structured in mobility and accessibility terms1_.

Figure 1. X-Y relationship model. The arrow indicates a causal relation, where social exclusion is shown as a present or detected outcome.

However, such relation suggests a more complex interaction of causal events, as seen in the expanded model in figure 2. This research deepens transport implementation and/or modification processes by addressing the role of transport-planning regimes. More specifically, the assessment practices embedded in these latter.

Figure 2. An expanded causal relation model of figure 1. Arrows indicate causal relations. Transport planning practices are conceived as a hierarchical array of institutional, organizational and operational components. These may be structured from a greater political and policy domain – involving complex strategies for social, environmental and/or financial dimensions, their priorities and agency factors – to those tasks related to

1_{The concepts of mobility and accessibility are largely covered in the urban and transport literature, but their} relation may be easily confused as these could be used interchangeably (Ross, 2000). Mobility and accessibility actually convey fundamentally different meanings (Ross, 2000). Mobility stands for the physical act of displacement (for instance, through various means of transport), whereas accessibility is a more complex dimension of these actions (Handy, 2002). Though a specific definition of the latter term is not conveyed (Handy, 2000), this research will mainly refer to this concept with a focus on the human actions and environment components (Iwarsson & Stahl, 2003) and how these are connected to an individual’s access and capacity to improve her/his quality of life through mobility actions. These issues will be again highlighted in the first pages of the theoretical framework.

(Spatial) Social exclusion in less benefited areas Implementation or modification of

(10)

8 more technical-oriented procedures, as modelling or designing and the assessment procedures behind these.

Hitherto, the focus on assessment practices is indeed related to a larger set of components that define transport-planning practices. Now, it should be made clear that it is not possible to refer to the whole compound of transport-planning practices within this limited research project. However, a specific focus on assessment practices at the second stage (X2-Y2) is a crucial one. Exploring how these technical and/or methodological procedures are sensitive to social exclusion, is a means to address how the phenomenon is comprehended and operationalized within transport-planning regimes, and a first step into a wider task of transport-planning components to be scrutinized. Likewise, it is expected that this contributes to a better understanding of how these regimes and their practices could have an impact on social exclusion.

Moreover, a specific order of these interrelated causal relations is needed. Attending the issues of the secondary level (X2-Y2) cannot be done without scrutinizing the primary causal relation (X1-Y1). The aim of first assessing social exclusion is to develop a means to understand how implementation or modification procedures could be connected to this phenomenon, and thereafter scrutinize how assessment practices in transport planning are addressing such links. Accordingly, with regard to research design terms, this project follows a causes-of-effects model as described by Goertz and Mahoney (2012), where attention is paid on tracking the causes (X’s) of an already detected condition or outcome (Y – in this case, social exclusion). These aspects will be thoroughly covered in the methodology section.

Furthermore, the main research question takes into account the whole causal model, linking both stages.

 Main research question (MRQ) – How does the implementation or modification of

transport systems impact the conditions of social exclusion of (potentially) less benefited areas of a city and how do transport-planning assessment practices play a role in this process?

Accordingly, sub questions in the first causal relation (X1 -> Y1) should address what particular environment conditions of transport might be enhancing or alleviating social exclusion within sectors that could be catalogued as potentially experiencing such phenomenon.

 Research sub-question (RSQ) 1 – What is the role that transport systems play in the

socio-spatial configurations of society and how do these impact the less benefited areas of a city and its inhabitants, in terms of social exclusion?

With the secondary causal relation (X2 -> Y2), attention is set on planning assessment practices within transport implementation procedures.

 RSQ 2 – What are the assessment practices embedded in transport implementation and

(11)

9 This last sub-question is basically addressing the internal dynamics of transport assessment practices with regard to social exclusion, observed in the general process of implementation or modification. The objective here is to understand how planning assessment practices and their understanding on this phenomenon – its conceptualization, instrumental or methodological operationalization, if existent or not – could influence and/or impact a condition of social exclusion. Figure 3 shows the relation of the presented research questions.

Figure 3. According to this scheme, the main research question (MRQ) can only be engaged once the sub questions (RSQs) are attended.

1.2. Main research aim and goals

The main purpose of the research is to contribute to the present scientific and professional debate linking transport issues and social exclusion, by (1) expanding the theoretical discussion on social exclusion, (2) defining what spatial environment conditions in transport can be classified as generators of exclusion, to finally (3) understand how planning practices and their assessment procedures could drive these results.

First, a theoretical reflection on social exclusion – its definition and engagement – is scrutinized, following the latest scientific discussions in transport literature. Thereafter, the aim is to develop a distinct epistemological approach to the phenomenon, paying attention to the related transport issues that could affect this outcome. Moreover, a context-sensitive approach is crucial. It is not expected to compare the levels of social exclusion of the chosen cases of study, but actually looking into their own processes. That is, understanding how social exclusion is constructed according to the local reality of the cases through transport implementation or modification processes. This will be done by introducing Amartya Sen’s capability approach (Sen, 1985/1987/2000/2007). Involving an individual perspective, this theoretical framework will be used to assess transport affecting the mobility and accessibility of citizens. It is an attempt to focus more on the process of social exclusion by linking socio-spatial practices with the transport environment conditions. The idea of using the capability approach is derived from the claims of a more sensitive approach to these issues, which have also stressed the need to involve participatory practices in transport planning (Hodgson & Turner, 2003; Preston & Rajé, 2007). Secondly, by identifying the conditions of transport that could be affecting social exclusion in the selected cases of study, it is expected to analyse how transport

(12)

10 planning assessment practices are capable of addressing these processes. Hence, this thesis will scrutinize both theoretical and practical issues for assessing social exclusion in transport. The main research question drives the overall purpose of this research but leaves open further development on these matters in the future. In short, this thesis is the first step into a wider task of theory and practice development.

(13)

11

2. Theoretical framework

In this section, most of the efforts will be used to present Sen’s capability approach as an alternative to evaluate social exclusion from an individual and context-sensitive perspective. In addition, I will present the theoretical dimensions of TEBUPAT (Transport Environment Participatory Bottom-up Assessment Tool), an instrument for detecting social exclusion based on the principles of the capability approach, along with other necessary conceptual definitions.

2.1. Social exclusion and transport: eyes on the process

As explained by Lucas (2012), “there is no overarching consensual view about what precisely constitutes social exclusion, but there is wide agreement that it reaches beyond a description of poverty to provide a more multidimensional, multi-layered and dynamic concept of deprivation” (p. 106). As suggested here, the study of social exclusion should involve a wide perspective on the various sources and processes that constitute the phenomenon. But still, the main concerning question remains: how to do this?

Within transport literature, social exclusion has been mostly related to accessibility issues and mobility deprivation. As explained by Kenyon et al. (2003):

[Social exclusion is the] process by which people are prevented from participating in the economic, political and social life of the community because of reduced accessibility to opportunities, services and social networks, due in whole or in part to insufficient mobility in a society and environment built around the assumption of high mobility. (Kenyon et al., 2003, p. 210, cited by Lucas, 2012, p.108)

However, strategies for detecting and deepening these interrelated conditions have been less explored (Hodgson & Turner, 2003). There has been an extensive reflection on the concept of social exclusion and its dynamic nature (Hodgson & Turner, 2003), yet, in practice, the term seems to still be “heavily reliant on income measures” (Stanley & Lucas, 2008, p. 36).

Moreover, we should consider “the relation between social exclusion, mobility and [accessibility] to be a dynamic one, and one that plays out at the level of society as a whole” (Cass et al., 2005, p.553). The nature of this multidimensional phenomenon requires an epistemologically open perspective. That is, understanding that its components are based on interactive individual and environment features, shaped by the context of study and its particularities.

2.2. Introducing Sen’s capability approach

The ‘capability approach’ (CA), developed by Amartya Sen (1985/1987/2000/2007), provides a strategic perspective for engaging the latter reflections. In the first place, the advantage of taking the scope of the CA is that we are able to assume the multidimensional

(14)

12 aspects that build social exclusion. Under this theoretical framework2_{, social exclusion is} understood as a process that is triggered by multiple individual capability failures (Sen, 2000), where the observation of diverse forms of deprivations could help us understand the structure of the final outcome. The CA focuses not on the level of income but deprivation of capabilities, e.g. access to healthcare, education, participating in the economic life and the autonomy in decision making (Zheng & Walsham, 2008, p.224). In short, this perspective allows us to engage the multiple sources or origins that generate social exclusion, focusing on the relational features of our interest, without losing sight of the overall picture of this phenomenon. The CA is thus as an open-ended theoretical approach for scrutinizing multiple constraints that affect the quality of life of individuals as part of a community. Under this logic, the constraints of an individual could speak for the problems of a whole community – “individual lives are inescapably social life” (Sen, 2000). Its main structure can be explained throughout the concepts of functionings,

capabilities and value-objects (Sen, 2007).

Functionings refer to all those acts a person is able to do as a human being. These are the

essential and basic actions that a person needs to execute in order to reach valuable life goals (Sen, 2007). More complex interactions of these functioning are grouped as

capabilities, that is, the multiple combinations of human actions. These combined actions

(capabilities) would allow someone to achieve significant or valuable goals in her/his life, defined as value-objects – the various goals, elements or events of meaningful importance to an individual. These three concepts constitute the core of the CA framework.

Functionings involve a wide array of conditions. They can be defined from elementary needs, as ‘being healthy’, to more specific and complex demands as being socially integrated to a community (Sen, 2007). These conditions will depend on the chosen

evaluative space (or framework of functionings), encompassed by the questions: “(1) what

are the objects of value?” and (2) “how valuable are the respective objects?” (Sen, 2007, p.272). Depending on the context or frame in which we may be assessing a social reality, our focus on the relevant functionings will be encompassed by different (local) value-objects structures. Hence, value-value-objects are all those elements that constitute a fulfilment of functionings and capabilities, which are necessary to substantially improve someone’s quality of life, depending on the (framed) context of the subject.

Furthermore, the notion of freedom is intrinsically related to capabilities. “It concerns the ‘range’ of choice a person has – not how [she/he] values the elements in that range or what she chooses from it” (Sen, 2007, p.272). In other words, it determines the way a person chooses according to her/his capabilities. If these are not present, or deprived by various factors, we can state that the freedom of this person is limited.

2_{Rather than a theory, the CA is better explained as a theoretical framework, as there are no pre-sets of its} core concepts. Hence, these may be adapted according to the needs and/or purposes of the interested researchers. In this paper, the CA is used to operationalize the concepts of mobility and accessibility as it will be further explained.

(15)

13

2.3. Mobility capabilities, accessibility deprivation and scale relations

‘Mobility capabilities’ is an adaptation of the aforementioned CA; an attempt to assess the conditions of mobility as the confluence of human actions, their needs, and environment components. The evaluative space of mobility capabilities – functionings and value objects – are therefore framed into relevant components of people’s mobility experience.

Functionings are related to the process of engaging the act of mobility with special attention to the socio-spatial determinants of such practices. That is, not solely focusing on displacement, but actually identifying the transport environment factors that affect such mobility practices. Furthermore, value-objects would refer then to the events and/or needs that are claimed with the act of mobility: valued destinations in the city, in various spaces (or territories) where objects or events are located, whose reach implies a contribution to quality of life on different levels (as reaching health or educational centres, commercial services, jobs, and so on). In addition, value-objects are set by a person’s social, cultural and economic environment – also considering how personal values and other priorities are set by the community they belong to. People from different strata in society may have different needs and desires, reflected on different spatial practices across the city; different destinations and commuting patterns. Accordingly, value objects may also be influenced or determined by the same environment conditions that affect the functionings and capabilities of an individual. The adapted principles of the CA are summarized in the following diagram.

Figure 4. A summarised diagram of the notions of the CA, applied to the concepts of accessibility and mobility.

Concordantly, referring to accessibility deprivation would mean an entanglement of diverse human actions (in relation to environment components), related to the act of mobility. In other words, a restricted set of (mobility) functionings, resulting in a diminished range of value-objects and options to reach such – a state of deprived or limited freedom.

Accessibility as the dimension of achieved (or non-achieved)

(16)

14 Moreover in this conceptual model, mobility actions (functionings) take place throughout different territorial dimensions or scales (Cass et al., 2005; Kaufman et al., 2004). The notion of scales is important in this definition, in order to pay attention on the social relations and practices of individuals throughout different spatial dimensions (Madanipour, 2002). In other words, to observe environment conditions in relation to people’s experiences under various scale relations or dimensions of the act of mobility. For such purposes, a three-level scalar approach will be used to refer to mobility experiences:

micro, meso and macro levels (see table 1)3_.

Table 1 – scalar relations of mobility capabilities.

 Micro: the proximate space and level of action of a person. For instance, within a

neighbourhood or community area.

 Meso: a larger range of mobility, beyond the proximate space, as the spatial reach among

larger district areas in a city.

 Macro: the larger mobility actions which also delimit the mobility space of a person. For

instance, the “city as a whole” or the a larger space among urban districts and/or conurbations.

Another important thing to remark on this conceptual model is that the relation among functionings, capabilities and value-objects is dynamic between scale transitions. Changes in any of these categories are reflected in patterns shifts, which finally determine how an individual is connected to the rest of the city and her/his role as a citizen (or society as a whole). Adopting this perspective is an attempt to analyse social exclusion where it might have been overlooked, not looking for evident results but hidden or neglected processes instead.

2.4. Concepts overview

The theoretical adaptation of the CA and its socio-spatial focus for framing mobility experiences is an attempt to develop an epistemologically open perspective, not establishing previous definitions or assumptions on transport environment components (for instance). Instead, the focus would be how an everyday life experience could give us details of the particularities of these environment components and their dynamics.

3_{A similar use of these terms can be found in the work of Kaufman et al (2004), but with a different use of the} terms specified here.

(17)

15 Figure 5. An overview of the main theoretical framework components. Social exclusion is evaluated from a mobility capabilities perspective, focusing specifically on the socio-spatial practices that take place within the mobility experience of an individual.

Accessibility stands for the confluence of socio-spatial practices, operationalized through the concept of mobility capabilities. Limited or deprived accessibility could speak for a state of exclusion of an individual (and her/his community). Figure 5 summarizes the concepts discussed so far in this chapter.

2.5 Introducing TEBUPAT

4

As for the methodological needs of this research, I hereby present TEBUPAT (Transport Environment Bottom-Up Participatory Assessment Tool), a planning instrument developed for assessing scenarios of potential social exclusion5_{. TEBUPAT is also based on} the CA principles and it is connected to the conceptual dimensions and reflections as presented so far. Although this tool is still in phase of development, its core theoretical and methodological procedures are applicable. Its design is mainly influenced by the late development of Qualitative GIS research practices and models, stimulating the creation of new approaches for both data gathering and analysis through participatory mechanisms

4_{The presentation of TEBUPAT is split between the theoretical framework and methodology chapters for} practical reasons: to first introduce how this is connected to the theoretical discussion presented so far and thereafter introduce its main methodological procedures for the collection and analysis of empirical evidence. 5_{The development of this tool took place during the early stage of this research, in parallel to the initial} research design, during a Research Apprenticeship in the COST Action TU1002 group, supervised by Prof. Dr. Luca Bertolini at the University of Amsterdam. Since its core functions are also based on the principles of mobility capabilities, most of the reflections highlighted in the theoretical framework are also embedded in TEBUPAT, making it suitable for the methodological purposes of this research.

(18)

16 (Cope & Elwood, 2009; Talen, 2000; Kahila & Kyttä, 2009; Knigge & Cope, 2006; Kwan, 2008). Individual and/or group interviews, plus participatory mapping instances – from previously chosen spatial units and their communities, as neighbourhoods6_{– serve as the} main input of data. TEBUPAT’s goal is to portray the spatial practices of communities as a means to understand the implications of transport elements in the composition of their mobility capabilities (considering experiences under all the detected transport modes in the area of study and its fabric). Part of the analysis results are shown as graphic visualisation models – GIS-based maps – bringing together the relation of the detected functionings with the spatial dimensions of the portrayed mobility experiences.

For identifying and categorizing functionings, TEBUPAT establishes a four-dimensions list based on the work of Cass et al. (2005)7_{. These are presented as (1) financial, (2) physical,} (3) organizational and (4) temporal factors. Table 2 describes these categories.

Table 2, Original source: TEBUPAT (Moran, 2014). Identifying individual functionings composing mobility capabilities.

Financial: Actions related to the use of monetary resources and the way that such are administered by each individual. It may also refer to the capacity of an individual to sustainably generate the necessary financial means for her/his mobility needs.

Physical: Actions related to the basic displacement functionings of a person. It is also relevant to consider health problems of the users that might result in physical limitations or other issues as fatigue.

Organizational: The way individuals organize with regard to the available means of transport and the agency factors linked to such tasks. E.g.: Car-pooling or adapting to the local public transport supply. It is a dimension which refers to a person’s own organizational components in her/his everyday life (as working or contacting/meeting other people), also considering community ties or other social factors that could influence these functionings. Temporal: “[The] degree to which people do or do not have control over, or flexibility built into, their

temporal regime” (Cass et al., 2005, p.551). How an individual has to adapt to the temporal regimes of transport and its schedules (as with public transport). A temporal regime can therefore be understood as the broader system of available transport, and how an individual is subjected to such.

Based on Cass et al., 2005, pp. 548-552

These functionings categories are universal under different modes of transport. Basically, this list works as a coding scheme for depicting the composition of mobility capabilities, being such the confluence of individual and environmental factors. Applying this coding scheme requires to first recognize all present modes of transport and their spatial components. Furthermore, functionings not only derive from the use of each transport mode, but also from combinations of these ones. As an example, let us consider walking as

6_{Details about the units of observation and selection of interviewees are covered throughout the next sections} (methodology chapter).

7_{These authors also reflect on the multidimensional significance of access – and therefore accessibility is here} interpreted in the same terms. These four dimensions are an attempt to broaden the scope on access in terms of human action and citizenship – multiple levels and conditions of human action and their socio-spatial determinants.

(19)

17 a transport mode that is required to reach and use of other modes as bus transport or other private mobility means as taxis. Moreover, there could be other modes that directly or indirectly affect the relations among other means (i.e.: the replacement of ‘walking’ with ‘cars’, due to infrastructure conditions; the existence of a constrained time schedule for accessing the available public transport options; etc.)

The notion of value-objects, as aforementioned, is also grasped in TEBUPAT as the identification of ‘valued destinations’. These are directly related to the confluence of specific functionings and therefore mobility capabilities. In other words, the way in which mobility capabilities behave – considering the grouped functionings in temporal, organizational, physical and financial terms – may speak for people’s actual capacity to engage mobility and may determine their freedom (accessibility), to reach other (primary or secondary) valued destinations, as visiting relatives or leisure events. Figure 6 summarizes the theoretical dimensions of TEBUPAT, which will be referred later the methodological section.

Figure 6. Original source: TEBUPAT (Moran, 2014). General composition of functionings for defining mobility capabilities and reaching value objects. These last are defined as multiple destinations. Micro, meso and macro levels are shown as different scales of observation and representation of data in a GIS embedded visualization model.

(micro/meso/ macro levels)

(20)

18

3. Methodology

The present section is focused on deepening issues related to concepts operationalization, case selection, definition of units of observation, data gathering and analysis procedures.

3.1. Causal model stages and units of analysis

Different methodological procedures apply to the different stages of the causal model in study (as presented in section 1). The use of TEBUPAT applies in the first stage, for the assessment of social exclusion, whereas the second stage mostly consists of a deductive coding process for deepening assessment practices in transport-planning regimes. The units of analysis of this research – where the comparison of case studies will be taking place – are set on this latter stage (the analysis of assessment procedures). Details on the relation of these two stages – in order to answer to the main research question – will be detailed further ahead in this same section.

3.2 Case selection

The target population in this research consists of large mobility regimes – or cities – with a special focus on the structure of transport and mobility (observed from a macro perspective). It is also considered that the presence of social exclusion is inherent in these large and rapidly growing and/or developing urban environments (Cassiers & Kesteloot, 2012), yet each case may have its particularities with regard to the presence of such phenomenon.

Since the chosen cases should be able to explain the chain of events that drives transport-planning assessment practices into specific conditions that alleviate or worsen social exclusion, the logic of pathway cases as described by Gerring (2006) is what best suits the purposes of this research. Pathway cases are useful to point and describe particular X’s that might be generating the same outcome, “where X[a] and not X[b] is likely to have caused a positive outcome Y=1” (Gerring, 2006, p.90), in order to elucidate the underlying nature of the described causal mechanism. In other words, this case selection type deepens – as its name suggests – the pathways in which a cause (or groups of causes - X) could lead to the theorized outcome, Y (Gerring, 2007). Pathway cases in fact are able to elucidate such theories by scrutinizing the inferred causal relations – the connections of particular X’s and the expected Y, and therefore compatible with the purpose of generating hypotheses by the end of this research.

Strictly speaking, pathways cases should always be chosen from cross-case inferences “or on the basis of strong deductive hunches” (Gerring, 2007, p.238). In addition, the chosen cases should not be outliers of the population in study, but instead typical cases – or cases whose scores are close to a regression line in the tradition of cross-case inferences (Gerring, 2007)8_{. This adds to the idea that a causal outcome is assumed valid but the} actual mechanism is not explicit or fully scrutinized. Accordingly, the aforementioned

8_{It should be noted that the use of pathway cases here is not set to limit the scope of the research, but actually} bringing a tangible strategy for unfolding the processes and mechanisms behind the outlined causal model.

(21)

19 description of the population in study and the selection of cases are akin to the use of such case type, although we might not be able to fully grasp cross-cases assumption at this stage.

Furthermore, both similarities and differences of the contextual factors in each case study may reveal important elements, able to explain the outcome target Y (the impact on social exclusion). Moreover, it is expected that the study of two cases in this opportunity may also draw conclusions on how other cases belonging to the same population could be studied in the future.

3.2.1 Defining units of observation

The first research sub-question explicitly brings the “role of transport systems in the

socio-spatial configurations in society” into relation with the existence and impact on “less benefited areas and its inhabitants, in terms of social exclusion”. The focus on transport (or

the components of a mobility experience) is subordinated to, first, recognize less benefited socio-spatial conditions (in relation to the total extension of a case study).

For achieving the described purposes, I initially focus on a macro analysis of mobility conditions within cities (mobility regimes) used to describe how individuals from particular areas of a city may be potentially experiencing transport related social exclusion – by observing time and distance factors, the transport fabric quality, diverse private and public options, financial issues and so on. This is a process I may refer as “zooming criteria”. The units of observation in this research are conceived within these parameters – a systematic approach to spatial units identified as areas of the city, relevant to the purposes of this research9_{. Table 3 summarizes the criteria used for zooming into} the units of observation in each case study10_{. These should be understood as abstract} categories and their main objective is to help us have a broad view of how inequalities may be shaped as mobility issues in the case studies. The main data set should consist of spatial or GIS data, as satellite images and data on the mobility network components (to visualize the present modes of transport and how they play a role in the composition of the whole mobility network). Other sources as Openstreet, Wikimapia and/or Google Maps should be used as well, for identifying land-use components and general infrastructure conditions. In addition, data indicating income averages, population density and territorial administration boundaries (districts) are required. All these data together should allow us to have an initial overview of the described factors.

The units of observation shouldn’t be confused as isolated sub-cases, but as territorial units within each mobility regime (or case studies). In addition, the in-depth study of such

9_{It should be noticed that these guiding criteria are focused on a spatial basis. In other words, the units of} observation are being chosen, fundamentally, by spatial criteria, instead of focusing first on individual or community profiles. The spatial insight prevails in the election of such units, thenceforth focused on the relation of the detected communities with their transport environment conditions.

10_{These criteria were initially inspired by several accessibility measures that have also brought into attention}

the mentioned factors in distinct forms (Schönfelder & Axhausen, 2003; Geurs & van Wee, 2004; van Wee et al., 2001). Yet the list of contextual factors here presented does not follow the same measurement conventions used by the cited authors. It seeks, instead, to identify such macro conditions as potential limitations at the individual level, as similarly reflected by Kwan and Weber (2003).

(22)

20 units of observation is focused on unveiling the process of social exclusion by analysing mobility capabilities – the “how” question – and not necessarily on quantifying such findings. Likewise, these units should not be expected to be explanatory of the whole mobility conflicts in a case study, but instead explore different variables that come together and might have an impact in the composition of limited mobility capabilities. Again, I am looking into particularities that may be of relevance to explain how a transport planning system could generate, alleviate or increase social exclusion. Therefore, it should be warned that the selection of units of observation follows a concordant observation of the listed factors, but not necessarily meaning these territorial units are able to grasp the whole components of mobility conditions across a whole case study. Besides, the contingency of social exclusion is not expected to be the same across every zone of a city, but on specific areas whose socio-spatial conditions may be akin to the manifestation of such phenomenon (observed through the listed factors). All in all, what I am looking for is the relation of different mobility issues, which fetch together different functionings relations (brought together as mobility capabilities).

Table 3 – Contextual factors for approaching to areas of interest (initial guiding criteria for defining units of observations)

(F1) Time and distance factors

 Regular commuting trips, connectivity dependence to city hubs, or general metropolitan & inter-district mobility patterns*

 Existence of considerable distances and access disparities among different territorial units.

(F2) Connectivity factors

 Differences in public and private mobility networks – quality of service and its access.

 Unattended transport demands and/or limited transport options (isolation of an area).

 General physical accessibility issues, related either to the urban structure of an area or geographical conditions.

(F3) Land-use and urban fabric factors

 Access to public amenities, services and facilities agglomerations (existence or inexistence of facilities for education, health, leisure, jobs and other valuable activities or events)

 Quality of public spaces and transport fabric within the surroundings of a neighbourhood/district area.

(F4) Potential financial constraints

 Allocation of low and middle-low income groups – possible financial problems in relation to mobility access**

* This background data should be able to elucidate the relation of different territorial units with regard to the existence of hubs (services, jobs, education, etc.) and how mobility is charted across these areas, under different modes of transport.

** The presented categories should also be analysed in terms of functionings to be explored, their relations and particular limitations among these, defining the condition of mobility capabilities. Moreover, contrasting and overlapping these factors may be useful to obtain different functionings settings – comparing layers of information that may reveal important relations to be deepened.

(23)

21

3.3. First stage: using TEBUPAT for depicting mobility-based capabilities

This section is focused on the application of TEBUPAT for the first stage of analysis, which applies after selecting units of observation.

3.3.1. Data gathering: conducting interview and mapping sessions

What mainly characterizes the application of TEBUPAT is the interpretation of individual and/or group experiences, gathered through interviews and participatory mapping activities from previously chosen units of observation. The ideal subjects for these activities are expected to be “exhaustive observers” of the current conditions of the area, also demonstrating a high capacity of description of their environment. However, a minimum or maximum of participants is not determined in this occasion. Due to the experimental conditions of this approach, theoretical saturation is improved by verifying the veracity and concordance (similarities and differences) of the given testimonies, through various interview sessions.

A model questionnaire was designed to allow these candidates to bring detailed testimonies (see table 4), allowing them to scrutinize the aspects they may consider of more relevance in their mobility experiences. This model questionnaire however doesn’t deepen aspects as gender differences, age constraints or other components related to more specific social conditions that could impact mobility capabilities. By the same token, the criteria for choosing participants are also subjected to bias. Nevertheless, the main objective is to bring, as aforementioned, a first impression of the composition of mobility capabilities and its analysis. What should be stressed here are the experimental conditions in which such approach is conducted. These issues should, in any case, be addressed in future research linked to the further presented findings.

As for the specific structure of the main activities, figure 7 provides an extensive view of the steps and goals that should be addressed in each session.

(24)

22 Table 4 – Model questionnaire for semi-structured focus groups.

First section:

 [Brief introduction of the participants and their backgrounds]

 How would you describe your proximate surroundings [neighbourhood/district location]?

Second section:

 How much do you use your area? / Are there any particular activities or events you regularly attend?

 Can you describe your daily activities?

 What means of transport do you use to reach these destinations?  How much time does it take you to reach these destinations?

 In your daily travels, which moments are more and/or less comfortable?  How important is to keep a regular schedule of these trips?

 Are there any particular issues that result in an uncomfortable experience while travelling?  Do you have any alternative choices to reach these destinations?

 In your surrounding area, which are the spots you frequently use to access transport?

Third section:

 What other places of the city do you have to visit and for what particular reason?  Do you visit relatives (or friends) often? / Where are they located in the city?  Which areas of the city do you visit for leisure / relaxation?

 Are you interested in a particular place you’d like to visit often?

Figure 7.Original source: TEBUPAT (Moran, 2014). A structural overview of individual and/or group interviews, divided in three sections. The first line with grey circles corresponds to the actual use of the questionnaire and its objectives, whereas the second line with yellow circles correspond to the objectives of the mapping activities. Although these lines are conducted in parallel, their objectives are intertwined.

mobility

Relation to the local neighbourhood and district area

Requirements, need and problems in engaging these activates

(25)

23

3.3.2. Data analysis: coding cycles, networks and interpretation

The first steps consist of the classification of quotes from the transcribed interview files, in each of the explored units of observation. At this point, it is important to identify the different modes of transport highlighted by the participants’ given descriptions, detecting both spatial and social components of their mobility actions (or functionings).

Basically, all relevant messages are filtered and grouped under the mentioned modes of transport, listed as categories (i.e: public transport [metro, bus]; private mobility [car, taxi]; non-motorized [walking, biking], etc). If combinations of different transport modes are described, these quotes are dynamically linked to each other. In addition, relevant infrastructure elements (as BRT corridors or urban highways) may also be sorted as specific categories. Furthermore, the classification of destinations – value-objects – is organized through a simple descriptive coding structure, with codes summarizing the most important destinations and how these are linked to a multiplicity of needs and/or significant actions, as the location of jobs, services, social networking (as visiting relatives or friends) and other related events. These data are arranged in a network view in Atlas.ti, the main CAQDAS (computer-assisted qualitative data analysis software program) used in this research (see figure 8). These quoting and coding structures would also allow a first interpretation of the relations among functionings, as represented in figure 9.

Further on, the interpretation of participatory maps into a common GIS-embedded visualisation scheme is conducted in parallel to the aforementioned steps. The main purpose of this is to visually depict the relation among functionings in space, under different scales (micro, meso and macro). Bringing into relation such scales is also a means to understand how the units of observation are embedded in the identified mobility regime and its structure. Figure 10 is a summary of the main graphic indicators used in this stage.

Hence, the interpretation of mobility capabilities is a result from these two data visualization modes, characterised by the use of both traditional qualitative data (interviews transcriptions) and a spatial insight retrieved from participant’s maps, conforming a mixed methods approach. The reader should expect this analysis to be found in a detailed description of the identified components of the mobility experience in each case, the relation among functionings, plus the interpretation of participatory maps (and these latter shown as evidence of the analysis process).

The final descriptions of mobility capabilities are not only subjected to what people may identify as an existing component of their mobility experience, but can also address missing elements. Therefore, this analysis can also relate functionings to the lack of certain components, as the inexistent of transport options or an inexistent connection to the metropolitan transport network within the neighbourhood or district area – as examples of an extreme limitation of diverse functionings. It should be also warned that these descriptions of mobility capabilities are subjected to a series of limitations, starting with the fact that these analyses are based on the data provided by a reduced set of participants.

(26)

24 Figure 8. Main data (network) arrangement – quoting and descriptive coding.

Figure 9. Hypothetical view of functionings relations, which later will compose the main written analysis.

(27)

25 Figure 10. Original source: TEBUPAT (Moran, 2014). TEBUPAT considers the use of symbols for spatial data representations. Here, symbols are meant to identify different functionings, linked to transport environment conditions.

3.4. Second stage analysis: transport planning assessment procedures

The leading research sub-question 2 (What are the assessment practices embedded in

transport implementation and how are these related and sensitive to the outcome of social exclusion?) seeks to identify assessment practices and scrutinize their sensitiveness to the

main phenomenon in study. In order to detect assessment practices or procedures, it is necessary to first identify the institutional and agency factors to which these belong. The detection of these assessment practices should be followed by an analysis focused on the internal dynamics of these procedures – as decision making processes and the relation between planners’ discourse on social exclusion, contrasted with what is actually conducted (or not) in the detected (or related) assessment procedures, within an implementation or modification process.

(28)

26

3.4.1 Data gathering

The main body of data for answering the leading sub-question comes from interviews with transport-planning practitioners. The interviewees are members from the different planning bureaus involved in implementation procedures, identified from the general institutional framework of the transport planning regime in each case. I also consider other figures who might be politically and/or technically involved in these matters, whose testimonies could be of relevance for understanding how transport planning practices work, but maintaining a focus on assessment procedures. Furthermore, policy documents and operational management reports are included, as these sources highlight key issues with regard to methodological and/or instrumental components of assessment practices in transport planning.

Interviews are guided by a semi-structured questionnaire which is adapted to every session, considering the type of institution being represented by the interviewee(s), plus the kind of tasks that are in charge of the reached planning unit(s). Table 5 – the model questionnaire – displays the main points to be covered in each interview. It should be noted that the questions in this table start by addressing the larger tasks of the selected transport planning units (questions 1 to 3), in order to further explore how the concept of social exclusion may be present (question 4) in the general structure of planning procedures, assessment practices and also planner’s discourse on the understanding of this phenomenon. Further on, the rest of the questions have a more specific insight in the actual methodologies and instruments that may be involved in the general implementation or modification procedures of transport systems.

Table 5. Model questionnaire for planning practitioners and other related figures.

1. What is the main purpose of the planning unit? (In terms of operational tasks, infrastructure management, etc.)

2. How has it changed during the last decades / Has it remained essentially the same during the last decades / what things have changed in terms of the unit’s focus and jurisdiction? 3. What are the main priorities in the process of transport design and its further

implementation?

4. What is the vision of the institution (or planning unit) on social exclusion? / Is assessing “social impact” a part of the tasks of this unit and how is it carried out? / Is it part of an integrated methodology? / Is it a policy aim and how is so present in the tasks of this institution (planning unit)?

5. In regards to the previous question, can you describe what (assessment) methodologies and/or instruments are involved in the tasks of transport planning and implementation? / Are there additional tasks that your unit has to take charge of?

6. How does this unit relate to other planning bureaus, also involved in matters of transport planning?

7. How do transport planning tasks are intertwined with land-use planning issues in this city?*

* These questions are referential, and differences may arise depending on the identified planning units and the very same interview sessions.

(29)

27

3.4.2. Data analysis

After the sources have been gathered and transcribed, the analysis procedures will be guided by a semi-structured deductive coding scheme. The main purpose of these coding operations can be summarized in three points, which will be also referred as the main coding categories at this stage: (1) to detect the general operational procedures of the reached planning units; (2) identifying the main priorities of the transport design and implementation processes, plus detecting the existence (or lack) of assessment procedures in the general modification and/or implementation stages. The last objective consists of (3) detecting the level of awareness to the phenomenon of social exclusion and assessing the sensitiveness of the related mechanisms (if detected). Figure 11 provides an overview of these coding categories, the main objectives, and their order of procedure.

Figure 11. A diagram of the coding objectives and categories. Whereas the first coding objectives pay attention to the general structure of transport-planning practices and the specific components to be observed (highlighted in yellow), the third objective is actually a prelude to more in-depth analytical deductions. The third objective can be only reached once the first categories are scrutinized and clearly distinguished.

The reader should be aware that the relation between these categories/objectives is linear. The first and second categories are characterized by the use of ‘open’, ‘descriptive’ and ‘process’ coding operations – all methods of data classification and organization that are suitable for a first approach into data (Saldaña, 2009), as these may assist the identification of the procedures of interest, as highlighted in the previous figure.

(30)

28 The third coding objective is also the opening stage for an in-depth analysis of the concept of social exclusion within these transport-planning assessment practices. More specifically, this stage works with axial and pattern coding operations – data analysis methods focused on rearranging the strands of information to focus on the patterns of interest (Saldaña, 2009). These coding sequences should focus on the definitions, links, and the operationalization of social exclusion or other related concepts, reflections or operations carried out by transport-planning practitioners. The matters discussed at this point are, basically, the presence, definition, structure and application of the concept of social exclusion in these planning units – its foundational and practical aspects, if existent. If social exclusion is not present in any form, the analysis will focus on why priorities, decision making processes and/or transport planner’s discourse may not consider this concept relevant to their practices.

3.5. Comparative analysis and answering the main research question

The purpose of the main research question (How does the implementation or modification

of transport systems impact the conditions of social exclusion of (potentially) less benefited areas of a city and how do transport-planning assessment practices play a role in this process?) is to bring empirical evidence and theoretical reflections on how the detected

conditions of social exclusion – assessed through mobility capabilities in the first stage of the causal model – could be related to transport planning assessment operations in general implementation stages (present along with priorities, decision making processes and planners’ discourse on social exclusion).

Figure 12. Sources and objectives of the comparative analysis ought to bring the first reflections, evidences, and theoretical proposal to answer the main research question.

The comparison of the units of analysis should bring into discussion how each regime is addressing the concept of social exclusion, reflected through the current state of

(31)

29 assessment practices in each transport planning regime, in light of the findings of TEBUPAT in each case. This means systematically linking the results from X1-Y1 to the units of analysis in X2-Y2, in order to check how assessment practices may be sensitive to the identified mobility capability sets. This will be done by bringing a summary of the main findings in each stage of the causal model, specified in a set of tables, with the case studies shown side by side. This visualization of these data will be useful to see trends in the assessment procedures (in each regime) and highlight what are these capable to observe if compared to the found mobility capabilities in each unit of observation. Figure 12 illustrates these steps. This comparison should assist the generation of hypothesis, addressing the detected results and the ways in which such issues could be connected to the larger array of institutional and agency factor of transport planning, with regard to social exclusion. Withal, the generation of hypothesis should also pay attention to the limitations within this research design.

3.5. Chosen cases and comparative remarks: Santiago and Amsterdam

The chosen cases of study are Santiago, the capital city of Chile, and Amsterdam in the Netherlands. Both mobility regimes fit into the criteria of the aforementioned case selection (section 3.2) – typical cases that have experienced an important urban growth. Yet the reader might consider that the urban evolution of both cities stand major differences, with comparison issues that should be carefully addressed. In answer to these matters, what is being compared is not, in any case, the level or particularities of social exclusion of both cases. As stated earlier, the focus is instead realizing how each transport planning regime is aware and sensitive to the existence of social exclusion within their contexts. Similarities and differences of these case studies are therefore bounded to the already presented research design settings.

The context of Amsterdam may be referred as a consolidated mobility regime, with the existence of a dense urban transport and mobility network. Santiago on the other hand is a rather developing context in this sense. During the last decades, the latter case has experienced vast infrastructural and operational design shifts in its mobility regime. Moreover, its transport planning regime is also in a developing stage.

I argue that comparing a consolidated and a developing transport-planning regimes may bring valuable reflections to the generation of hypotheses, also addressing the topic to a larger global context of transport-planning practices. These issues will be deepened in the following analysis sections.