Content uploaded by Alexa Delbosc
Author content
All content in this area was uploaded by Alexa Delbosc on Jun 08, 2016
Content may be subject to copyright.
Modelling the social and psychological impacts
of transport disadvantage
Graham Currie •Alexa Delbosc
Published online: 13 May 2010
Springer Science+Business Media, LLC. 2010
Abstract This paper presents the results of a research project aiming to develop a robust
empirical model to measure links between transport disadvantage (TD), social exclusion
(SE) and well-being (WB). Its principal aim is to improve on current research methods in
this field. Existing approaches derive associations between TD and its impacts through
simple comparative methods, through qualitative methods and using limited and prescrip-
tive definitions of SE. The new method draws from an interview questionnaire measuring
TD through self-reported difficulties with transport. A principal components analysis of
responses identifies four statistically significant sub-scales (transit disadvantage, transport
disadvantage, vulnerable/impaired and rely on others). SE is represented in five dimensions
including income, unemployment, political engagement, participation in activities and
social support networks. Well-being adopts standard psychological measures—‘Satisfac-
tion With Life Scale’ (SWLS), ‘Positive Affect’ (PA) and ‘Negative Affect’ (NA). Struc-
tural equation modelling (SEM) was used to model links between TD, SE and WB. A
hypothesised model proposed negative associations between SE and WB and between TD
and WB and a positive association between TD and SE. Modelling results showed that
scales used to measure TD, SE and WB were all statistically related to their underlying
concepts. Modelling of the hypothesised links between constructs was generally favourable
with a good statistical fit. However the relationship between TD and WB was not significant.
An exploratory analysis supported the hypothesis that this was caused by high reported
travel difficulties for both highly mobile and less mobile people. A revised theoretical model
explored the theory that feelings of isolation due to time poverty might be mediating the
TD-WB link. SEM analysis of the revised model confirmed a good model fit with statis-
tically significant measures between TD, time poverty and WB. Time poverty was not found
to be associated with social exclusion. The final model suggested that TD is positively
associated with SE with a measured strength of .27. SE is strongly negatively associated
with WB (-.87). TD is positively associated with time poverty (.19) while time poverty is
negatively associated with well-being (-.14). Areas for future research are identified.
G. Currie (&)A. Delbosc
Department of Civil Engineering, Institute of Transport Studies, Monash University, Building 60,
Clayton, VIC 3800, Australia
e-mail: graham.currie@eng.monash.edu.au
123
Transportation (2010) 37:953–966
DOI 10.1007/s11116-010-9280-2
Keywords Transport disadvantage Social exclusion Well-being
Structural equation model
Introduction
A range of research suggests that transport disadvantage can contribute to social exclusion
or a poor quality of life. For example the UK Social Exclusion unit found that 38% of
jobseekers say transport is a key barrier to getting a job and that transport costs are a
significant burden on students aged 16 to 18 (reported fully in 2003). Hine (2004) found
lower car ownership, lower rates of car licensing and greater reliance on walking and
public transport in disadvantaged areas of urban Scotland. Cervero (2004) identified an
increasing gap in access between suburban job opportunities and the inner-city poor. A
number of recent studies have found that transport plays an important role in quality of life
amongst the elderly (Banister and Bowling 2004; Mollenkopf et al. 2005; Spinney et al.
2009).
A necessary limitation of research in this field is the methodological approach to
identifying the association between transport disadvantage and its impacts. In general
methodologies typically involve the use of comparative analyses whereby the character-
istics of groups are compared based on contrasting spatial, mobility, access or socio-
demographic qualities (Church et al. 2000; Schonfelder and Axhausen 2003; Hine 2004;
Department for Transport 2006; Dodson et al. 2006). This approach has many practical
benefits in terms of data collection but can be simplistic in assuming that contrasts in
behaviour identified are caused by differences in access and mobility quality shown. When
causal links are explored in more depth in this research they are often based on qualitative
and anecdotal evidence based on limited samples of interview or focus group data (Young
2001; Raje
´2003; Fritze 2007; Hurni 2007; Penfold et al. 2008).
In addition many studies generally limit their definitions of social exclusion to specific
types of groups which are said to be associated with social exclusion (such as the elderly,
unemployed or ethnic minority groups). Social exclusion is rarely clearly defined and
measured and often it is assumed based on group membership.
Many of these methodological limitations are entirely understandable given the prac-
ticalities of research data collection and analysis. Despite these limitations research in this
field has provided a powerful basis for shaping policy and practice. However a more
quantitative methodology which illustrates the causal links between transport disadvantage
and its social impacts would be highly desirable. It would represent a more robust and
defendable basis for policy development and could demonstrate the relative strengths of
associations between transport disadvantage, social exclusion and well-being.
This paper presents the results of a research project aiming to develop a robust empirical
model to measure links between transport disadvantage, social exclusion and well-being. It
is part of an international study undertaken in Melbourne, Australia.
1
1
Australian Research Council Industry Linkage Program Project LP0669046 ‘Investigating Transport
Disadvantage, Social Exclusion and Well-being in Metropolitan, Regional and Rural Victoria’, Monash
University, in association with the University of Oxford (UK), University of Ulster (UK), Department of
Transport, Victoria, the Bus Association of Victoria and the Brotherhood of St. Laurence. The principal
chief investigator is Prof. G. Currie and the project Research Fellow is Ms Alexa Delbosc. The chief
investigators are Prof. T. Richardson, Prof. P. Smyth and Dr. D. Vella-Brodrick. The partner investigators
are Prof. J. Hine, Dr. K. Lucas, Mr. J. Stanley, Dr. J. Morris, Mr. R. Kinnear and Dr. J. Stanley.
954 Transportation (2010) 37:953–966
123
The paper starts with a short review of the literature examining the relationship between
transport, social exclusion and well-being. This is followed by a description of the research
methodology. The results of the research are then presented. The paper concludes with a
summary of the key findings including a discussion of the implications of the findings for
future research in this field.
Transport and social exclusion
There has been a recent surge in literature on how transport disadvantage can exacerbate
social exclusion or reduce quality of life. Interest in reducing social exclusion stems from
French social policy (Lenoir 1974/1989) but more recently the UK has focussed a great
deal of policy attention on reducing social exclusion (UK Social Exclusion Unit 2003). The
UK is also one of the few governments to make the transport-exclusion relationship a focus
of policy (Hodgson and Turner 2003; Department for Transport 2006). The policy focus on
transport and exclusion varies greatly between nations; for example Lucas (2004) found
that Germany, Italy and Japan have developed specific policies to address transport dis-
advantage amongst the disabled and elderly whereas American policy emphasises transport
equity. Most recently the European Commission funded a comprehensive best practice
review of transport programs to reduce exclusion across Europe (Holmes et al. 2007).
Research literature on the topic uses a variety of methods to identify the relationship
between transport disadvantage and social exclusion. The most common method is com-
parative or category analysis to identify groups facing exclusion or transport disadvantage.
Spatial category analyses are often used to identify transport disadvantage based on dis-
tance from or travel time to jobs, shops, hospitals or education, either by car or public
transport (Church et al. 2000; Schonfelder and Axhausen 2003; Hine 2004; Department for
Transport 2006; Dodson et al. 2006). Other common forms of spatial analyses examine
levels of car ownership, licensing rates or public transport service levels using comparative
approaches (Currie 2004; Hine 2004; Hurni 2005; Currie 2009).
Social exclusion is also most often measured using comparative analyses but the cat-
egories are socio-economic or demographic. A common focus of the literature is the study
of transport needs for a particular group thought to be socially excluded such as ethnic
minorities (Raje
´2003), the disabled (Capability Scotland 2004; Casas 2007; Penfold et al.
2008), young people (Currie 2007; Hurni 2007) and young or single parents (Fritze 2007;
Hurni 2007; Titheridge and Solomon 2008). Research of this kind typically associates
these groups with social exclusion however the use of a comprehensive definition of social
exclusion and its components is rare in the transport literature.
Focus groups or qualitative interviews are often used in conjunction with comparative
and categorical analyses to explore the causal links between transport problems and social
impacts (Young 2001; Raje
´2003; Fritze 2007; Hurni 2007; Penfold et al. 2008).
What all of these studies lack is a quantified measure of both social exclusion and
transport disadvantage with a demonstrated empirical measure of the causal links between
these constructs. One or both of the constructs is invariably measured as a proxy or
indicator, such as car ownership or travel time for transport disadvantage or being a
member of a group thought to face exclusion.
Transport and well-being
The literature on the impact of transport on well-being is less developed than the literature on
transport and social exclusion. At the societal level, well-being and quality of life are
Transportation (2010) 37:953–966 955
123
increasingly being recognised as desired outcomes of public policy. Bhutan was the first
country to measure the quality of life of its citizens using a standard of ‘‘Gross National
Happiness’’ instead of Gross National Product (The Centre for Bhutan Studies 2009). The
OECD has since established guidelines for measuring societalwell-being (Boarini et al. 2006).
Only a few papers have examined the role transport plays in well-being and thus far the
literature has been restricted to elderly cohorts. Banister and Bowling (2004) found six
‘‘building blocks’’ to quality of life amongst the elderly and found that transport plays an
important role in maintaining quality of life. Mollenkopf et al. (2005) surveyed elderly
Europeans to create a structural equation model illustrating the strong relationship between
mobility and quality of life, taking into account factors such as socio-economic status and
residential area. Most recently Spinney et al. (2009) found a small relationship between
realised mobility (minutes spent outside the home) and feelings of life satisfaction amongst
Canadian retirees.
What these studies suggest is that when mobility is restricted amongst the elderly, their
quality of life suffers. Transport likely plays an important role in maintaining social
relationships and activities like hobbies which can reduce social isolation. However the
impact of transport on well-being has not been demonstrated in a wider population group.
It is likely that this impact is very complex; improving understanding in this area is a major
focus of this paper.
Methodology
This paper aims to fill methodological gaps identified in the literature by empirically
measuring the relationship between transport disadvantage, social exclusion and well-
being using primary surveys to quantify each construct and structural equation modelling
methodology to explore the empirical links between them.
Measuring transport disadvantage, social exclusion and well-being
An interview questionnaire was adopted to collect data on transport disadvantage, social
exclusion and well-being using the approach identified below.
A household interview survey adopted a ‘follow on’ approach from an existing
household travel survey called VISTA (the Victorian Integrated Survey of Travel and
Activity, The Urban Transport Institute, 2008). Households who participated in VISTA
were given the opportunity to opt-into the household survey. This approach enabled better
targeting of the interview quotas and also reduced questionnaire time by appropriating the
travel diary elements of the VISTA travel survey.
The sample frame deliberately targeted both socially advantaged and disadvantaged
households (based on income) as well as groups who had good and bad access to transport
and walk accessibility. In general one person was selected per identified household using a
‘Kish grid’ random sampling method (Kish 1965). This acts to reduce self selection bias of
respondents with regards to issues such as reported transport problems.
The household survey was undertaken between September and December 2008 and 535
surveys were collected. Slightly over half of the sample (58%) was made up of households
below $AU 1,100 per week, the gross median income of Australians (Australian Bureau of
Statistics 2007). The majority of the sample lived in outer suburban Melbourne (77%)
including a good coverage of locations with walkable (46%) and non-walkable access
(54%) to local activity centres.
956 Transportation (2010) 37:953–966
123
Based on their records in the VISTA survey, the average trips per person per day in the
household survey sample was 3.4 and the average person kilometres was 36.8. Although it
is slightly higher than in the UK where people took an average of 2.7 trips per day and
travelled an average of 30.2 km (Department for Transport 2005), it is lower than the
average American who took 3.7 trips per day and travelled 59.4 km (Hu and Reuscher
2004). The travel habits of the sample do not appear to be unusual in an international
context.
Transport disadvantage has been measured in hundreds of different ways (Dodson et al.
2004). Often these measures look at potential indicators of disadvantage such as trip rates,
distance to services, lacking a car or poor public transport (Church et al. 2000; Currie
2004; Hine 2004; Hurni 2005; Department for Transport 2006). Less often do studies
measure the direct consequences of transport disadvantage, such as an inability to access
job interviews (UK Social Exclusion Unit 2003) or restricted access to education and
employment (Currie 2007).
In this study transport disadvantage is measured using subjective, self-reported mea-
surements. Survey participants were asked to judge potential difficulties with eighteen
different types of transport disadvantage (identified though a review of the research lit-
erature). These are listed in Table 1. They were asked to rate how easy or difficult they
found such issues as ‘getting to places quickly’ or ‘covering the costs of transport’. All
responses were subjective i.e. they record participant’s perception of their situation.
A principal component factor analysis (reported fully in Delbosc and Currie 2010)
determined that these questions could be expressed by four statistically significant
underlying factors.
2
Some 57% of the variance in responses was explained by the four
Table 1 Derived factor groups—type of transport disadvantage
Factor group name Component feature of transport disadvantage: difficulty with…
Transit disadvantage Buses/trains/trams being available at night
Buses/trains/trams being available at weekends
Buses/trains/trams operating frequently
Being able to make bus/train/tram connections
Transport disadvantage Being able to travel when you want to
Being able to get around reliably
Finding transport so you can travel
Getting to places quickly
Finding the time to travel when you need to
Vulnerable/impaired Being able to physically get onto/off buses/trains/trams
Needing help to get around on your own
Being able to understand where to go
Feeling safe from theft/attack when travelling on your own
Rely on others Having to rely on others for transport
Finding someone to provide assistance when transport is available
Covering the costs of your transport
2
Two questions were not included in these factors. ‘‘Being able to get information about buses/trains/
trams’’ had a low initial extraction to the model. The Principal Components Analysis was re-run without it.
‘‘Being able to get to bus/train/tram stops/stations’’ was excluded because it loaded equally across three
factors.
Transportation (2010) 37:953–966 957
123
factors derived. The output factors and their component transport disadvantage types are
reported in Table 1.
Social exclusion is a complex, multi-dimensional construct. It is broader than the notion
of poverty and refers to limits in societal participation and social support as a result of a
combination of factors which may include unemployment, low income, discrimination,
crime and poor skills (Cabinet Office Social Exclusion Task Force 2009). In the transport
literature social exclusion is usually defined as membership of an excluded group; in
contrast the social policy literature identifies characteristics that underpin exclusion. Its
measurement includes multiple dimensions such as economic, social and political factors
(Bhalla and Lapeyre 1997). Burchardt (2000) refined these dimensions to include income
level, unemployment, political engagement and social interaction. Specifically, social
exclusion was measured using five dimensions:
•Income: Participants were classified into four categories of non-equivalised gross
household income
•Unemployment: This included both those who were looking for work and those who
were unemployed due to disability or illness
•Political engagement: This was measured by recording recent participation in political
or community groups.
•Participation: Participants were asked if they have been excluded from a range of
activities such as hobbies, sport and visiting libraries
•Social support: This was measured by asking how easily people could get help from
others if they needed it.
The measurement of well-being at the individual level is a mature research topic in
social psychology (Kahn and Juster 2002). The two papers to date that relate transport to
well-being used empirical measures of ‘affect’, ‘satisfaction with life’ and ‘quality of life’
(Banister and Bowling 2004; Mollenkopf et al. 2005). For this study two measures of well-
being are adopted:
•Satisfaction With Life Scale: Participants indicate how much they agree with five
statements about their life conditions and how close their life is to their ideal (Diener
et al. 1985)
•Positive and Negative Affect Schedules: Participants rate how much they generally feel
a range of positive and emotions (Watson et al. 1988). This measure is made up of two
independent subscales
The ‘Satisfaction With Life Scale’ (SWLS) and ‘Positive Affect’ (PA) and ‘Negative
Affect’ (NA) Schedule are standard measures for measuring subjective well-being in the
psychology literature (Diener 1984; Lucas and Diener 1996). Taken together these scales
measure subjective well-being or quality of life.
Structural equation modelling
Structural equation modelling (SEM) is a statistical methodology that examines the
underlying structural relationship between variables and displays these relationships pic-
torially. Variables can be conceptualised in one of two ways:
•Latent variables are theoretical constructs that cannot be directly observed. Well-being,
social exclusion and transport disadvantage are all examples of these constructs.
958 Transportation (2010) 37:953–966
123
Because these constructs cannot be observed directly they cannot be measured directly
and instead they must be operationally defined by observable behaviours (Byrne 2001).
•Observed variables are the observed indicators of underlying latent constructs. Self-
report ratings of happiness, participation in political activities and having to rely on
others for transport are examples of observed indicators of wellbeing, social exclusion
and transport disadvantage, respectively.
Confirmatory factor analysis methods are used to form latent variables using observed
variables as factors. Once the variables within the model have been defined, regression
methods are used to examine hypothesized relationships between variables.
SEM methodology requires the construction of hypothesised relationships between
constructs which are then tested using the modelling results. It is hypothesised that
experiencing difficulty in accessing activities would result in lower measurements of
subjective well-being (Fig. 1). This hypothesis is based on research demonstrating the
importance of mobility on quality of life amongst the elderly (Banister and Bowling 2004;
Mollenkopf et al. 2005) though this connection has not been demonstrated in a broader
population. Furthermore it was hypothesised that social exclusion will lead to decreases in
well-being based on the negative impacts social exclusion has on individuals. It was further
hypothesised that transport disadvantage would increase social exclusion, though the
evidence linking the relational aspects of social exclusion with transport disadvantage is
still emerging (Stanley and Stanley 2007).
Results
Table 2shows the results of the measures identified for the field survey sample as a whole.
PA, NA and SWLS were skewed toward higher well-being which is typical of international
samples (Cummins 2001). Social exclusion measures showed a greater variation in results.
Political engagement was low overall, the mean participation score was near the middle of
the scale and social support scores were quite high. The average scores of transport
disadvantage were toward the middle of the scale with the highest average scores reported
for transit disadvantage.
Fig. 1 Hypothesised model
Transportation (2010) 37:953–966 959
123
Model fit
The hypothesised model was tested using the statistical analysis package AMOS 17.0.
Maximum likelihood method estimation was used. Figure 2shows a visual representation
of the resulting model.
The evaluation of the overall hypothesised model resulted in a significant chi-square
(v
2
=182.81; df =51; p\.0001), suggesting that the fit of the model to the data is not
adequate. However it has been long recognised that v
2
has been proved to be an unrealistic
measure of statistical significance due to its extreme sensitivity to large sample sizes
(MacCallum et al. 1996). For this reason, additional indicators of model fit are usually
consulted.
Three measurements of model fit were used. The Goodness of Fit Index (GFI) and
Adjusted Goodness of Fit (AGFI) indexes indicate a good ‘‘fit’’ if their values are above .90
(Byrne 2001); GFI for the model was .942 and AGFI was .912 indicating a good fit. The
root mean square of the error of approximation (RMSEA) was .070, considered a rea-
sonable fit (MacCallum et al. 1996).
All observed variables in the model showed highly significant relationships (p\.001)
to their respective latent variables (Fig. 2, represented in circles). This means that the
observed variables are all methodologically sound indicators. The measures for ‘‘transport
disadvantage’’ had the strongest relationships to their construct, ranging from .48 to .84.
Measures for ‘‘subjective well-being’’ were also strong though NA was lower at -.36.
The measures for social exclusion, though still highly statistically significant, had lower
values ranging from .26 to .44, suggesting that this construct is not as homogenous as the
other two. This is consistent with Burchardt (2000) who emphasises that social exclusion is
a multi-faceted construct and found that correlations between social exclusion factors were
Table 2 Variable frequencies
Measure Scale Mean Standard
deviation
Well-being
Positive affect (PA) 1–5 3.6 .59
Negative affect (NA) 1–5 1.7 .52
Satisfaction with life scale (SWLS) 1–7 5.2 1.1
Social exclusion
Income ($AU) (household income measured in four
categories)
n/a 58% below
$1,100 p/w
n/a
Unemployed (percent unemployed (categorical variable)) n/a 5% n/a
Political engagement (number of activities engaged in) 0–5 1.5 1.3
Participation (number of activities engaged in) 0–5 2.5 1.4
Social support score 4–12 10.7 1.3
Transport disadvantage
Transit disadvantage 1–5 2.9 .88
Transport disadvantage 1–5 2.3 .76
Vulnerable/impaired 1–5 2.3 .64
Rely on others 1–5 2.6 .74
960 Transportation (2010) 37:953–966
123
below .40. The fact that these fairly independent indicators were all related to the
underlying construct is quite significant to the social exclusion literature.
High levels of social exclusion were a very strong predictor of subjective well-being at
-.76, p\.001. There was a small but significant relationship between transport disad-
vantage and social exclusion at .18, p\.05. However one relationship in this model was
not statistically significant. The relationship between transport disadvantage and well-
being was low and did not quite meet the criteria for statistical significance (p=.07).
Links between transport disadvantage and well-being have been established in the
literature, but only amongst elderly populations and even then the effect sizes have been
small. The null result in the structural equation model suggests one of three alternatives:
that the nature of the measurement has not captured this relationship, that the relationship
is more complex or indirect than our current model allows or that transport has no influence
on well-being in a heterogeneous population.
Past research linking transport and well-being used objective mobility measurements
whereas this model used a subjective measurement of transport difficulty. To test the
relationship between subjective transport disadvantage and mobility a measure of realised
mobility was incorporated into the analysis.
3
Out of the survey sample, 517 had filled out a
travel diary for the 2008 VISTA survey. A version of the model represented in Fig. 2was
run with the number of daily trips regressed onto the Transport Disadvantage latent var-
iable. No significant relationship resulted between realised trips and transport disadvan-
tage; people who travel a lot were just as likely to state that they experience transport
disadvantage as people who travelled little or none at all.
Perhaps, then, the lack of relationship between subjective TD and WB is partially
explained by this subjective nature. Someone who is extremely disadvantaged may have a
low expectation of the number of activities they should be able to access. Conversely
someone who is extremely advantaged may expect to access many more activities and may
even view traffic congestion as a significant ‘‘difficulty’’. However, the significant rela-
tionship between TD and SE suggests that this measure, whilst subjective, is capturing
some aspect of ‘‘real’’ difficulties.
Fig. 2 Results of initial model
3
Similar tests were incorporated into the SEM modelling in Mollenkopf et al. (2005) work.
Transportation (2010) 37:953–966 961
123
Another possible explanation for the lack of the TD-WB link is that it was clear that
many of those reporting transport problems actually had good realised mobility, had good
access to transport, were employed and had income. From this perspective well-being may
be being influenced by wider life influences and a TD-WB link may be obscured by these
factors.
Building from the above findings it was hypothesised that other underlying factors
might be mediating the relationship between transport disadvantage and well-being, factors
that were not included in the hypothesised SEM model. That is, transport disadvantage
may only lower well-being if the disadvantage causes an increase in a third factor, for
example, social isolation. The findings from the literature amongst the elderly suggest this
as a possible link.
The household survey contained a set of questions measuring social isolation. Partici-
pants were asked to identify if any of a range of issues caused them to feel ‘‘isolated or cut
off from society’’. The most common reason people cited was ‘‘not having enough time,’’
an issue of time poverty.
An analysis of responses to the time poverty related question was undertaken (a full
detailing of this analysis is presented in Currie and Delbosc 2010). The most common time
poor groups lived in households with children, were employed, took more trips and were
slightly younger. They also had lower ratings of life satisfaction and well-being. Overall
these results suggest that people who are not commonly seen as disadvantaged (the
employed and those with higher incomes) can have feelings of isolation associated with
time poverty, and that time poverty can reduce their well-being.
These findings supported the use of time poverty as an additional theoretical construct
that may mediate the influence of transport disadvantage on subjective well-being. If
transport disadvantage leads to feelings of time poverty, then subjective well-being may be
compromised.
Figure 3shows the revised model with time poverty playing a mediating role between
transport disadvantage and well-being. The overall model shows a good statistical fit that is
nearly identical to the original model, as are the relationships between the observed
variables and their latent variables (see Fig. 2). The relationships between transport dis-
advantage and social exclusion (.27) as well as between social exclusion and subjective
well-being (-.87) are stronger than in the original model. Those associated with subjective
transport disadvantage were slightly more likely to say they didn’t have enough time (.19,
p\.001) and people displaying time poverty had slightly lower ratings of subjective well-
being (-.14, p\.01). Measures of model fit were slightly lower (reflecting a less parsi-
monious model) but still within the range of a good fit.
The results of the revised model suggest that time poverty is a factor which acts to link
transport disadvantage and well being. However the size of the links modelled while
significant are not very large. The link between time poverty and well being seems a
reasonable conceptual one since its theoretical foundation in this research is that those
feeling isolated and cut off from society due to lack of time might reasonably be thought to
have lower quality of life and well-being. The transport disadvantage-time poverty link is
perhaps more intriguing. As noted a high share of those reporting subjective transport
disadvantage were very mobile and were employed. It is not difficult to see links between
long distance commuting, lack of time and mobile working people in congested cities. It
seems likely that it is this kind of linkage which is influencing this outcome. However it
would be interesting to explore links between TD and WB with a larger sample of more
socially disadvantaged and less mobile groups.
962 Transportation (2010) 37:953–966
123
Summary and conclusions
This paper presents the results of a research project aiming to develop a robust empirical
model to measure links between transport disadvantage, social exclusion and well-being. A
major motivation for the work is the desire to improve methodologies applied in this area.
Existing approaches tend to derive associations between transport disadvantage and its
impacts through simple comparative assessments, through the use of limited qualitative
methods and using prescriptive definitions of the concept of social exclusion.
The method involved the application of an interview survey questionnaire measuring
transport disadvantage through a range of questions exploring self reported difficulties
associated with transport. A principal components analysis of responses identified four
statistically significant sub-scales of disadvantage (transit disadvantage, transport disad-
vantage, vulnerable/impaired and rely on others). Social exclusion was represented by five
dimensions including income, unemployment, political engagement, participation in
activities and by the strength of social support networks. Well-being was measured on
standard scales from psychology including ‘Satisfaction With Life Scale’ (SWLS),
‘Positive Affect’ (PA) and ‘Negative Affect’ (NA).
Quantitative assessment of links between transport disadvantage (TD), social exclusion
(SE) and well-being (WB) was undertaken using structural equation modelling (SEM).
Consistent with SEM method, a hypothesised model proposed negative associations between
SE and WB and between TD and WB with a positive association between TD and SE.
Modelling results showed that the structure of theorised constructs for TD, SE and WB
were all good indicators for their underlying constructs. The transport disadvantage con-
structs (transit disadvantage, transport disadvantage, vulnerable/impaired and rely on
others) were all related to the underlying concept of ‘‘transport disadvantage’’ (p\.001).
Low income, unemployment, low political engagement, low participation and low social
support were all related to the underlying concept of ‘‘social exclusion’’ (p\.001) and the
PA, NA and SWLS scales were all related to ‘‘subjective well-being’’ (p\.001).
Modelling of the hypothesised links between constructs was generally favourable and
the statistical fit of the model was considered to be good. The link between transport
disadvantage and social exclusion was modest (-.18) and highly statistically significant
Fig. 3 Revised model including time poverty
Transportation (2010) 37:953–966 963
123
(p\.001). The relationship between social exclusion and well-being was stronger (-.76)
and also significant. However the relationship between transport disadvantage and well-
being just missed the cut-off point for statistical significance (p=.07).
An exploratory analysis suggested that part of this ‘missing link’ could be due to the
subjective nature of the transport disadvantage measure. Past research that linked transport
to well-being in elderly populations used measures of realised mobility, yet we found no
relationship between subjective reports of transport disadvantage and realised trip rates.
People who travelled a lot were just as likely to think they had transport problems as people
who only travelled a little.
Furthermore it was hypothesized that subjective transport disadvantage may only
influence well-being if it increased feelings of social isolation. The household survey found
that the most common reason people reported for feeling isolated or cut off from society
was time poverty (‘not having enough time’).
A revised theoretical model explored the theory that time poverty might be acting to
influence the TD-WB link. The revised model suggested that transport disadvantage is
positively associated with social exclusion (.27) and that social exclusion is strongly
negatively associated with well-being (-.87). These values are quantifiable demonstrations
of the degree of influence one factor has over another. A -.87 relationship suggests that
being socially excluded greatly contributes to poor well-being. The link between transport
disadvantage and social exclusion is smaller, suggesting that transport disadvantage has a
small but significant contribution to social exclusion.
Transport disadvantage was positively associated with time poverty (.19) suggesting that it
has a small but significant influence on making people feel so out of time that they became
socially isolated. Furthermore, time poverty has a small negative influence on well-being
(-.14). Therefore, this model hasdemonstrated that one way subjective transportdisadvantage
can lower well-being isif it causes people to becomeso time poor they are cut offfrom society.
No link between time poverty and socialexclusion was found; indeed those who were time poor
in this sample were working, had higher income and were relatively more mobile than others.
This suggests a negative association between time poverty and social exclusion.
The relative difference in effect sizes between the four concepts is an important con-
tribution to transport research as it has real implications for prioritising social policy. The
strong relationship between social exclusion and well-being suggests that policies that
successfully reduce social exclusion may greatly improve quality of life. The smaller link
between transport disadvantage and social exclusion in this heterogeneous population
suggests that much more is involved in social exclusion than just transport. If transport
policies are going to reduce social exclusion, they should be carefully targeted for maxi-
mum societal benefit. And finally, in a heterogeneous population, subjective feelings of
transport disadvantage may only result in lower well-being if they have a flow-on effect to
feelings of time poverty or social isolation.
It should be noted that these findings apply to a fairly broad sample that includes both
advantaged and disadvantaged populations. Subjective measures of transport disadvantage
may mean very different things to different social groups. A highly mobile, middle-class
family may consider themselves ‘‘disadvantaged’’ by traffic congestion or having to wait
for a late-night train home from the city. A pensioner without a car may consider them-
selves highly ‘‘advantaged’’ just to have a bus that comes by once an hour on weekdays.
This is a considerable limitation of modelling across a broad population.
With this in mind there may be merit in returning to the method of examining these
relationships within disadvantaged populations. Transport expectations may be more
consistent and comparable in this context. A special survey of extremely disadvantaged
964 Transportation (2010) 37:953–966
123
individuals (including the disabled, carers, the unemployed and single parents) is currently
underway as part of the authors’ research. Exploring the interactions between transport,
social exclusion and well-being within this group and comparing the findings to the
broader sample will bring considerable insights.
References
Australian Bureau of Statistics: Household income and income distribution, 2005–06. Report 6523.0 (2007)
Banister, D., Bowling, A.: Quality of life for the elderly: the transport dimension. Transp. Policy 11, 105–
115 (2004)
Bhalla, A., Lapeyre, F.: Social exclusion: towards and analytical and operational framework. Dev Change
28, 413–433 (1997)
Boarini, R., Johansson A., d’Ercole, M.M.: Alternative measures of well-being. OECD Social, Employment
and Migration Working Papers No. 33 (2006)
Burchardt, T.: Social exclusion: concepts and evidence. In: Gordon, D., Townsend, P. (eds.) Breadline
Europe: The Measurement of Poverty, pp. 385–403. Policy Press, Bristol (2000)
Byrne, B.M.: Structural equation modelling with AMOS: basic concepts, applications and programming.
Mahwah, Lawrence Erlbaum Associates (2001)
Cabinet Office Social Exclusion Task Force. Context for social exclusion work. http://www.
cabinetoffice.gov.uk/social_exclusion_task_force/context.aspx (2009). Accessed 6 Feb 2009
Capability Scotland: The influence of public transport provision on the social exclusion of disabled people in
Scotland (2004)
Casas, I.: Social exclusion and the disabled: an accessibility approach. The Prof Geogr 59(4), 463–477
(2007)
Cervero, R.: Job isolation in the US: narrowing the gap through job access and reverse-commute programs.
In: Lucas, K. (ed.) Running on Empty: Transport Social Exclusion and Environmental Justice, pp. 181–
196. The Policy Press, Bristol (2004)
Church, A., Frost, M., Sullivan, K.: Transport and social exclusion in London. Transp. Policy 7(3), 195–205
(2000)
Cummins, R. A.: Life satisfaction: measurement issues and a homeostatic model. In: Zumbo, B. (ed.) Social
Indicators and Quality of Life Research Methods: Methodological Developments and Issues, Year-
book, 1999. Kluwer Academic, Amsterdam (2001)
Currie, G.: Gap analysis of public transport needs: measuring spatial distribution of public transport needs
and identifying gaps in the quality of public transport provision. Transp. Res. Rec. 1895, 137–146
(2004)
Currie, G.: Young Australians: no way to go. In: Currie, G., Stanley, J., Stanley, J. (eds.) No Way to Go:
Transport and Social Disadvantage in Australian Communities. Monash University ePress, Clayton
(2007)
Currie, G.: Quantifying spatial gaps in public transport supply based on social needs. J. Transp. Geogr.
(2009). doi:10.1016/j.jtrangeo.2008.12.002
Currie, G., Delbosc, A.: Quantifying Links Between Time Poverty, Well Being and Social Exclusion.
Transportation Research Board Annual Meeting, Washington DC, USA (2010)
Delbosc, A., Currie G.: Transport problems that matter—social and psychological links to transport dis-
advantage. J. Transp. Geogr. (2010). Published on line doi:10.1016/j.jtrangeo.2010.01.003
Department for Transport: Focus on personal travel: including the report of the National Travel Survey
2002/2003 (2005)
Department for Transport: Guidance on accessibility planning in local transport plans. http://www.dft.
gov.uk/pgr/regional/ltp/accessibility/guidance/gap/ (2006). Accessed 2 June 2009
Diener, E.: Subjective well-being. Psychol. Bull. 95(3), 542–575 (1984)
Diener, E., Emmons, R.A., Larsen, R.J., Griffin, S.: The satisfaction with life scale. J. Personal. Assess.
49(1), 71–75 (1985)
Dodson, J., Gleeson, B., Sipe N.: Transport Disadvantage and Social Status: A Review of Literature and
Methods. Urban Policy Program, Griffith University (2004)
Dodson, J., Buchanan, N., Gleeson, B., Sipe, N.: Investigating the social dimensions of transport disad-
vantage—I. Towards new concepts and methods. Urban Policy Res. 24(4), 433–453 (2006)
Fritze, J.: You Might as Well Just Stay Home: Young Mums and Transport in Victoria, pp. 1–40. Victorian
Council of Social Services, Melbourne, Australia (2007)
Transportation (2010) 37:953–966 965
123
Hine, J.: Transport disadvantage and social exclusion in Urban Scotland. Built Environ. 30(2), 161–171
(2004)
Hodgson, F.C., Turner, J.: Participation not consumption: the need for new participatory practices to address
transport and social exclusion. Transp. Policy 10, 265–272 (2003)
Holmes, K., Clifford, S., Gregory, K., Mabelis, J.: ECLIPSE: Good Practice Review—Final Version Year
Two. European Commission (2007)
Hu, P.S., Reuscher T.: Summary of Travel Trends: 2001 National Household Travel Survey. U.S.
Department of Transportation (2004)
Hurni, A.: Transport and Social Exclusion in Western Sydney. Australasian Transport Research Forum,
Sydney (2005)
Hurni, A.: Marginalised groups in western Sydney: the experience of sole parents and unemployed young
people. In: Currie, G., Stanley, J., Stanley, J. (eds.) No Way to Go: Transport and Social Disadvantage
in Australian Communities. Monash University, Melbourne (2007)
Kahn, R.L., Juster, F.T.: Well being: concepts and measures. J. Soc. Issues 58(4), 627–644 (2002)
Kish, L.: Survey Sampling. Wiley, New York (1965)
Lenoir, R.: Les exclus, un franc¸ais sur dix, 2nd edn. Editions du Seuil, Paris (1974/1989)
Lucas, K.: Transport and Social Exclusion: A Survey of the Group of Seven Nations. FIA Foundation,
London (2004)
Lucas, R.E., Diener, E.: Discriminant validity of well-being measures. J. Personal. Soc. Psychol. 71(3), 616–
628 (1996)
MacCallum, R.C., Browne, M.W., Sugawara, H.M.: Power analysis and determination of sample size for
covariance structure modelling. Psychol. Methods 1, 130–149 (1996)
Mollenkopf, H., Baas, S., Marcellini, F., Oswald, F., Ruoppila, I., Szeman, Z., Tacken, M., Wahl, H.-W.:
Mobility and quality of life. In: Mollenkopf, H., Marcellini, F., Ruoppila, I., Szeman, Z., Tacken, M.
(eds.) Enhancing Mobility in Later Life. IOS Press, Amsterdam (2005)
Penfold, C., Cleghorn, N., Creegan, C., Neil, H., Webster, S.: Travel behaviour, experiences and aspirations
of disabled people. National Centre for Social Research (NatCen). http://www.dft.gov.uk/pgr/
scienceresearch/social/travelbehaviours (2008)
Raje
´, F.: The impact of transport on social exclusion processes with specific emphasis on road user charging.
Transp. Policy 10, 321–338 (2003)
Schonfelder, S., Axhausen, K.: Activity spaces: measures of social exclusion? Transp. Policy 10, 273–286
(2003)
Spinney, J.E.L., Scott, D.M., Newbold, K.B.: Transport mobility benefits and quality of life: a time-use
perspective of elderly Canadians. Transp. Policy 16(1), 1–11 (2009)
Stanley, J., Stanley, J.: Public transport and social policy goals. Road Transp. Res. 16(1), 20–30 (2007)
The Centre for Bhutan Studies: Gross national happiness. http://www.grossnationalhappiness.com/ (2009).
Accessed 29 May 2009
The Urban Transport Institute: Victorian integrated survey of travel & activity 2007–08: Survey procedures
and documentation, The Victorian Department of Transport (2008)
Titheridge, H., Solomon, J.: Social Exclusion, Accessibility and Lone Parents. UK-Ireland Planning
Research Conference, Belfast (2008)
UK Social Exclusion Unit: Making the Connections: Final Report on Transport and Social Exclusion. Office
of the Deputy Prime Minister (UK) (2003)
Watson, D., Clark, L.A., Tellegen, A.: Development and validation of brief measures of positive and
negative affect: the PANAS scales. J. Personal. Soc. Psychol. 54(6), 1063–1070 (1988)
Young, D.: The life and death of cars: private vehicles on the Pitjanjatjara Lands, south Australia. In: Miller,
D. (ed.) Car Cultures, pp. 35–58. Oxford, Berg (2001)
Author Biographies
Graham Currie Professor of Public Transport, Institute of Transport Studies (ITS), Monash University. He
directs the Social Research in Transport (SORT) research clearinghouse (www.sortclearinghouse.info), has
membership of several TRB transit committees, the UITP academic network and is Vice Chair of the US
TRB Light Rail Transit Committee.
Alexa Delbosc Research Fellow at ITS. She has a Psychology BA (Lewis and Clark College USA) and a
Social Psychology MA by research (Harvard University) recently published in the Proceedings of the
National Academy of Sciences. Her research concerns the social implications of public transport and
transport disadvantage.
966 Transportation (2010) 37:953–966
123
