Car dependence in England and Wales: spatial inequalities and implications for a just transition

Abstract

This paper explores the nuances of car dependence in England and Wales by identifying four distinct archetypes that span structural and conscious forms. Employing the 2011 England and Wales Census, archetype prevalence is mapped across the study area at the LSOA level, and a demographic analysis is performed. We find that while dependence exists across the study area, structural dependence is found more in rural areas, particularly the east coast of England and Wales, while conscious dependence is more prevalent in and around urban centers. The demographic makeup of each archetype differs significantly, with disability, socio-economic class, and ethnicity arising as notable significant indicators. This work highlights that an equitable transition to a sustainable transport system requires geographically and demographically specific policies tailored to the unique needs of each archetype. This transition away from car dependence, especially internal combustion engine vehicles, is imperative for a just and climate-resilient transport system.

1 Introduction

Both private vehicles and public transport are integral to providing access to key goods and services, facilitating social inclusion, and wellbeing (Gates et al., 2019; Cooper et al., 2019; Chatterjee et al., 2019). Due to its ability to connect people to important destinations and therefore facilitate direct and indirect enhancement of life domains, there is a well-established positive impact between transport and quality of life (Delbosc, 2012; Banister and Bowling, 2004).

This positive effect of transport on quality of life has been demonstrated across several geographies. For example, in Denmark, Hybel and Mulalic (2022) found that access to the various components of transport infrastructure (access to public transport and proximity to highways) are important factors for quality of life and Mattson et al. (2021) found similar results in the US. These benefits are found to be more prominent in vulnerable or disadvantaged groups, indicating the importance of considering socio-economic status in transport planning. Stanley et al. (2011) found that while trip making increases wellbeing, the value of trips increases with lower income. Similarly, gains in mobility are particularly important for groups such as the elderly and those in isolated rural communities through increased access to health care, psychological benefits, exercise, and community inclusion (Spinney et al., 2009; Cooper et al., 2019; Grajdura et al., 2026). Nonetheless, car ownership itself has been demonstrated to increase self-reported happiness and quality of life even when accounting for socio-economic status and income in rural China (Li et al., 2022) and in both car-oriented and transit-oriented cities in Europe, though the effect is stronger in car-oriented cities (Mouratidis, 2025).

Currently, the UK's transport system largely relies on the burning of fossil fuels in internal combustion engines, with the domestic transport sector accounting for 30% of national greenhouse gas emissions in 2024 (Department for Energy Security & Net Zero, 2025). The transport sector overtook power generation for the first time in 2021, to become the greatest single source of CO₂ emissions in the UK (Office for National Statistics, 2024). More than half (54%) of these emissions are estimated to stem from the use of private vehicles (Department for Transport, 2025d). This is reflected in the way that mobility and transport have become synonymous with private car use in the UK. Between 1971 and 2023, the proportion of households that owned at least one car increased from 52% to 78%, whilst those owning two or more vehicles increased from 8% to 34% (Department for Transport, 2025e). Indeed, a study by the RAC found that 80-90% of people say that they would find it hard to adjust to lifestyles without a car (Simpson, 2025).

However, the balance between those who produce these emissions and those who are most affected by them is deeply inequitable (Brand and Boardman, 2008; Büchs and Schnepf, 2013; Barnes et al., 2019). In the UK, half of all greenhouse gas emissions are generated by just 15% of the population (Frost and Hobbs, 2024). This discrepancy is closely related to income, with the wealthiest 1% emitting 13 times that of the lowest income quintile or 7 times the average (Frost and Hobbs, 2024). This trend is also true for NOx emissions (one third of which originate from road vehicles) and particulate matter (PM) emissions due to increased private vehicle ownership and greater travel distances of wealthier groups (Barnes et al., 2019). Moreover, younger and lower income households experience disproportionately high levels of traffic-related pollution: Barnes et al. (2019) explain that “in respect of traffic emissions, the poor pollute the least and are polluted the most”.

Given the overlapping environmentally and socially negative impacts of our current internal combustion engine-dominant transport system, it is clear that moving toward alternatives is essential if we are to develop both climate resilience and a just transport system. In light of the environmental element of this transport crisis, the UK government has proposed several policies to encourage mode and behavior shift. Some of these policies have focused on encouraging active travel and public transport. For example, in 2025, a £3 bus fare cap was introduced across England to ensure fares remained below inflation, alongside a £1 billion investment into improving bus services (Department for Transport, 2025a). Similarly, the UK government announced almost £300 million investment to improve and promote active travel in England (Active Travel England, 2025). However, despite these investments into active travel and public transport, the primary driver toward a more sustainable transport system has been the push toward electric vehicles (EVs).

The focus on EV adoption is clearly illustrated by the extensive policy and funding landscape in the UK to support this transition. Notably, in 2023, the UK Government announced a legal end date for ICEV sales, stating that all new vehicles must be 100% zero emission by 2035 (Department for Transport, 2023a). To support this, several additional policies and measures have been enacted. These include mandates for car manufacturers to have an increasing proportion of their annual new car sales to be electric up to 100% by 2035 (Department for Transport and Office for Zero Emission Vehicles, 2024), as well as investments into vehicle and battery research and development (Advanced Propulsion Centre, 2025; Faraday Battery Challenge, 2025). Furthermore, in 2025, the UK Government announced an Electric Car Grant, where up to £3,750 of the EV cost would be subsidised by the government to increase consumer uptake (Department for Transport and Office for Zero Emission Vehicles, 2025). To further facilitate uptake, public charge point regulations were published in 2023 to increase the accessibility and availability of charge points have been rolled out, including the provision of contactless payments, reliability, and helpline access, and the UK Government launched £10 million in funding to support EV charging on the strategic road network (The Public Charge Point Regulations, 2023; Department for Transport, 2025f). These policies are demonstrably helping to bring about the rapid support and uptake of EVs, with 6.24% of vehicles in the UK plug-in electric at the end of 2024, increasing from 2.26% in 2021, and the number of public charge points exploding from 28,460 to 73,000 in the same period (Zapmap, 2025a,b).

However, a direct shift from ICEVs to EVs, while reducing lifetime emissions by up to ≈50% (Kelly et al., 2023), will perpetuate current infrastructures, urban forms, and travel norms, and thus will not address the social inequalities associated with private vehicles. In much of the UK, car dominance has transmuted into car dependence, at both a societal and individual level. In this paper, car dependence is defined as where an individual or society is, or feels that they are, reliant on their car to access essential goods and services. This dependence renders mode shift difficult or impossible. As such, EV-specific challenges such as the upfront costs of EVs, requirements for home charging, and cost and availability of public charging sites make the switch difficult or undesirable for many, leaving parts of the driver population at risk of being “left behind" by the EV transition (Dsouza, 2024; Ciampoli, 2024; Transport & Environment and SKIM, 2024). It is therefore essential to understand who in the UK is dependent on their cars, and the geography of these patterns of dependence, particularly in relation to underlying socio-economic distributions, in order to understand how to equitably transform the transport system into one that is sustainable and accessible to all.

There has been a significant body of work to date exploring car dependence from a breadth of perspectives, which can be summarized into four major categories: assessing car use and ownership, land use characteristics and accessibility, practices and behaviors, and applications to policy (Van Eenoo, 2025). Alternatively, car dependence can be considered to exist within a three-level framework, comprising the micro level (regarding individuals), macro level (regarding societies), and the meso level (regarding trips and travel practices) (Mattioli et al., 2016a). A review by Cremer-Schulte et al. (2025) builds on the literature to propose a unified framework for defining, conceptualizing, and operationalising car dependence, attesting to the breadth and complexity of understanding car dependence. The framework describes internal and external components of car dependence, each with objective and subjective features, and the extensive array of metrics and methods for its measurement. The range of approaches and dimensions for quantitatively assessing car dependence in the literature is as broad as that for its conceptualization, including assessment of car use and access, accessibility outcomes, subjective perceptions, modeling of choice sets, and car dependence as an explanatory variable for wider issues - and employ an array of metrics such as transport demand and supply, land use and form, accessibility, opinions and experiences, and socio-demographics (Sierra Muñoz et al., 2024).

This paper presents a framework for understanding the geographic and demographic distribution of forms of car dependence in England and Wales to inform an equitable transition away from ICEVs. This work builds on the growing body of work looking at the drivers and prevalence of car dependence across the world, especially work by Carroll et al. (2021) in Ireland. First, four distinct archetypes of car dependence are defined using area-level data. Then, the geographical extent and spatial distribution of each identified archetype will be determined to understand their respective prevalence. Finally, a demographic analysis will be performed to elucidate socio-economic trends in vulnerability. This spatial and socio-demographic analysis of car dependence patterns will inform which areas or groups of people may be most at risk during the EV transition in the current UK policy landscape, and as a result which groups may benefit from, and require support for, moving away from dependence on private vehicles.

2 Background

2.1 Transport access inequality

Vast inequalities can be observed in how different groups of people travel. Two key factors that are commonly found to determine travel habits are income and spatial distribution. As such, geography and socio-demographics will be explored in this paper.

Affordability is found to be a primary barrier to good transport access in the UK, with lower-income groups spending proportionally more on transport and, as such, traveling less overall (Government Office for Science, 2019b). Overall, 26% of UK households have no access to a car, rising to 48% in the lowest income quantile (Chatterton et al., 2018; Stacey and Shaddock, 2015). Indeed, the costs of not only purchasing, but also insuring and maintaining a private vehicle have been found to be primary concerns for drivers (Close Brothers Motor Finance, 2025). Meanwhile, those who depend on the bus for their transport in the UK are more likely to be lower paid, turn down further education, and experience more deprivation due to transport issues (Government Office for Science, 2019b). At the other end of the scale, private vehicle ownership rates increase with income, while proportional income spent on transport decreases (Government Office for Science, 2019b; RAC Foundation, 2020).

Travel habits and the spatial distribution of jobs, services, and homes are also closely linked. This can be seen clearly when looking at housing; housing positioned further from city or commercial centers is typically more affordable, generating a trade-off between higher transport costs and lower housing costs (Lipman, 2006; Suel et al., 2024). This is demonstrated by the urban-rural divide in the UK, where rural communities especially suffer from transport poverty due to a lack of public transport links combined with a low density of opportunities (Sustrans, 2012). Sustrans (2012) defines somebody as transport poor if the following criteria apply:

• They have a low income (10% or more of their income is spent on running a car)

• They live more than one mile from a bus or train station

• It takes more than one hour to access essential services.

This definition would place 1.5 million people in England into transport poverty. The spatial distribution of transport poverty is evident in the UK: in 2019, 87% of trips in rural UK areas were by private vehicle, compared with 78% in urban areas (Government Office for Science, 2019a). This reliance on the expensive private vehicle in poorly connected communities leads to rural areas being especially prone to the “double vulnerability” of simultaneous energy poverty and transport poverty (CREDS, 2023). This also means that rural areas contribute greater transport-related emissions than more urban areas (Büchs and Schnepf, 2013).

Beyond income and spatial distribution, other external socio-cultural factors may strongly influence people's transport behaviors. A study into trans and gender non-conforming people's travel habits revealed that they are more likely to depend on public transport due to generally higher rates of disadvantage (Lubitow et al., 2020). However, due to high levels of negative experiences resulting from discrimination on public transport, these groups of people find themselves increasingly turning to private vehicles or simply traveling less, compounding systemic inequalities. A similar trend was found with women traveling at night (Plyushteva and Boussauw, 2020). Additionally, a study into homeless people's travel behaviors revealed highly spatially constrained behaviors and reiterates the need to look holistically at the system to serve marginalized communities when considering mobility developments (Jocoy and Del Casino, 2010).

2.2 Car dominance

In the UK in 2023, 60% of UK journeys and 78% of distance traveled was completed by privately owned cars in 2023, compared to just 9% by public transport and 2% by active travel modes (National Travel Survey, 2024). In this way, cars have become the primary and default mode of transport for private travel, creating car dominance. As discussed by Handy (1993) and Sheller and Urry (2000), today's car dominance can be attributed to a compounding cycle of “auto dependence” where increased car use leads to non-localized urban design that prioritizes speed over proximity, promoting accessibility for cars and hence leading to more car use. This ratchet effect is also described by Lucas and Jones (2009), where, as a result of people using their cars for trips where alternatives do exist, these alternatives are cut back, which leaves people locked in to car use. Mattioli et al. (2020) note how cars have become so embedded into society that pro-car decisions are even viewed as apolitical; the car is synonymous with growth, prosperity, and modernization.

Car dominance generates stark social inequalities and problems: as more people enter into car ownership, those who do not use cars experience a greater disadvantage due to increasingly car-centric infrastructure. This creates a “radical monopoly” as described by Ivan Illitch (2001) where even non-drivers are impacted by the dominance of cars over transportation, such that “nearly everyone—whether or not they drive—is harmed by [automobility]” (Miner et al., 2024). As summarized by Mattioli (2014): “the self-reinforcing cycle of car dependence results in an increasing intensity of car deprivation for a decreasing proportion of the population”. One major contributing factor to this inequality is the high costs of owning and running a car, estimated to cost on average over £3,300 per year in the UK by NimbleFins due to high costs of insurance, fuel, tax, parking, and other maintenance (Yurday, 2025). For lower-income households, this can pose significant financial problems if there are no alternatives available; RAC Foundation (2020) found that the lowest income households spend up to a quarter of their income on cars in 2018/19, while Chatterton et al. (2018) found that the lowest income households spent up to twice as much of their income on motoring than the wealthiest households, demonstrating the regressive and unequal nature of car dominance. Mattioli et al. (2016b) researched the incidence of households in “car-related economic stress" (CRES), finding that up to 9% of households in Great Britain were affected by CRES in 2012, with higher risk associated with lower urban density and poorer public transport provision. The social inequality surrounding CRES is clear when analyzed alongside income: while 9% of UK households experience CRES, this rises to 67% of households in the lowest income quantile (Mattioli, 2017). The financial burden of owning and running a car is often compounded by housing costs (Cao and Hickman, 2018; Robinson and Mattioli, 2020) and fluctuating fuel costs (Lovelace and Philips, 2014; Martiskainen et al., 2021). These elements of transport poverty are summarized by Lucas et al. (2016) into four main categories: affordability, mobility poverty, accessibility poverty, and exposure to externalities. These categories intersect and exacerbate one another; one such feedback loop can be seen in accessing employment in certain areas, where cars are needed to access employment, but employment is needed to afford a car (Mahieux and Mejia-Dorantes, 2017).

2.3 Car dependence

Since the overwhelming majority of people in the UK use cars to travel, car dependence manifests in many different ways across socio-demographic groups. There are several ways of characterizing car dependence. One of these ways is to distinguish between structural and conscious dependence, as proposed by Stradling (2007). In this formulation, Stradling concisely explains the difference between these forms of dependence: “the former are unable to switch modes, the latter unwilling.” Thus, structural dependence arises from a real lack of alternatives available to the individual for travel. This manifests in two key ways: infrastructural or personal. Examples of infrastructural forms of structural car dependence include insufficient or non-existent public transport provision, poor cycle infrastructure, or a lack of services within walking distance. Examples of personal forms of structural dependence include physical or mental disability prohibiting the use of public transport or active travel modes.

On the other hand, conscious dependence is where the dependence is perceived, even if not stemming from a real, physical lack of alternatives. This dependence may arise from habit, preference, or convenience. As summarized by Brindle (2003), “we are not dependents on the car as such, but rather on what it provides”. As with other possessions and objects contributing to self-identity, car pride is a strong emotive reaction that many drivers experience (Gatersleben, 2020). Indeed, Moody and Zhao (2019) found that car pride has a significant and positive relationship with household car ownership. Many symbolic aspects of driving a car, such as wealth and status symbolism, self-expression, and freedom, contribute to car ownership and use (Steg, 2005; Steg et al., 2001; Gatersleben, 2020). Moreover, Steg (2005) shows how symbolic and affective factors can be more significant than instrumental factors (e.g., cost, comfort, convenience, and safety) in determining an individual's car use for commuting. The lines between practical utility and preference are often blurred, as Kent (2014) highlights how people are attached to their cars for more than the time savings car journeys bring. The strength of these symbolic factors varies across demographics, with young, male, lower-income, frequent drivers experiencing the greatest effects. The way in which driving a car is both due to emotions and elicits emotions itself means that many people see cars as extensions of themselves or their families, and means that this form of conscious car dependence is a complex psychological trait: “[cars] are deeply embedded in ways of life, networks of friendship and sociality, and moral commitments to family and care for others” (Sheller, 2004). In more recent times, the car has also developed into a form of “coccooning”, where drivers and passengers may find momentary calm and seclusion in an increasingly busy urban environment (Wells and Xenias, 2015).

Car dependence can also be viewed from the perspective of individual versus trip dependence. Here, the distinction is made between a person who is dependent on their vehicle at all times for all trips, and a specific trip purpose that requires a vehicle (Goodwin, 1995). An individual might be dependent if they have a mobility issue that prohibits them from using public transport or active travel, or if their home is in an area poorly served by public transport. In contrast, trip-based dependence is specific to journey type; an individual may be able to use public transport to reach leisure destinations but requires a car for commuting to work, dropping a child off at school, or going to the supermarket (Stradling, 2007). The idea of certain locations (trip destinations) being car dependent is also explored by Brindle 2003), who explains how access, rather than mobility per se, is critical and people will often travel further (by car) to access a more favorable option for a destination than the most local option.

This is a problem that is not going away by itself; indeed, the majority of new build homes in the UK are built in car-dependent areas, thus perpetuating and locking in dependence for many years to come, if unmitigated (Kiberd and Straňák, 2024). In response to this growing issue, the UK Department for Transport has issued a Connectivity Tool to provide information around how well-connected places are, and what transport support may be required Department for Transport (2025b). The analysis in this paper aims to contribute to this knowledge base, to provide insights into which areas in the UK are vulnerable to car dependence and how this relates to socio-demographics, to understand the equity implications of both current connectivity and future developments.

3 Proposed dependence archetypes

Market segmentation was initially described by Smith in 1956 regarding how a heterogeneous consumer base can be disaggregated into smaller, homogeneous groups for targeted marketing strategies (Smith, 1956). This segmentation can be performed a priori, where the number and type of segments are established beforehand, or post-hoc, where the number and type of segments are determined by the data itself (Wedel and Kamakura, 2000). A priori segmentation is generated by systematically selecting relevant characteristics, while post hoc segmentation usually relies on data-driven clustering to identify groups that share similarities. Neither segmentation methodology is universally favored: while Anable (2005) argues that post-hoc methods are significantly more robust, Haustein and Hunecke (2013) conclude that no single segmentation approach is inherently superior. Population segmentation has been applied to several other fields since, including that of transport planning. For example, segmentation has been utilized in understanding daily mobility patterns and attitudes (Charleux, 2018; Jensen, 1999), mode shift potential (Anable, 2005; Arian et al., 2021), and car dependence (Mattioli, 2017; Carroll et al., 2021). The use of segmentation in transport planning is particularly useful for developing effective targeted policies. This is evidenced by its use in several policy documents at both the local and national level, such as the Transport Classification of Londoners by Transport for London (2017) or the transport user personas developed by Department for Transport (2023b).

In a paper by Carroll et al. (2021), an a priori segmentation was performed on electoral zones in Ireland. The segmentation was binary, such that zones were either in the target group or outside of the group; those in the target group were recommended as priority areas for government support and intervention. The target group, named “Forced Car Ownership”, was defined by three characteristics: high single car ownership, high deprivation, and poor public transport access. This combination of characteristics highlights a particularly vulnerable segment of the population: those who struggle to afford to drive a car, but have no choice since there are insufficient viable alternatives available. Their analysis found that rural areas are particularly at risk of forced car ownership, and found a potentially self-reinforcing causal link between deprivation and transport disadvantage. In this article, the work by Carroll et al. (2021) has been built upon and extended. The prevalence of Forced Car Ownership, defined similarly, is explored across England and Wales alongside three additional novel proposed segments, or archetypes, of car dependence. These additional archetypes are formulated in a similar way to the Forced Car Ownership archetype, but with varying socio-economic and travel characteristics. This ensures that multiple forms of car dependence are accounted for and that other potentially vulnerable groups are not overlooked.

3.1 Proposed archetypes

We propose a set of four car dependence archetypes, comprising three novel archetypes in addition to forced car ownership as defined by Carroll et al. (2021): rural car commuter, city driver, and urban car commuter. Each archetype describes different ways in which people may experience car dependence, encompassing both structural and conscious dependence. The archetypes are defined by both travel to work habits and socio-demographic characteristics, generating a description of who might be expected to experience each form. A summary of all four archetypes is shown in Table 1.

The four archetypes were developed to be relevant to the context of England and Wales by drawing on both existing literature surrounding transport access and car dependence, and available area-level datasets. This ensures they are derived from real-world data and experience, thus making them useful for meaningful policy development. These four archetypes were developed in this way so as to represent a cross-section of the forms of dependence and how these intersect with traditional forms of vulnerability; an individual may be vulnerable to car dependence but not be considered vulnerable otherwise. The archetype definitions populate each quadrant of a set of deprivation/public transport access axes, as visualized in Figure 1, where the x-axis represents increasing public transport accessibility and the y-axis represents increasing levels of deprivation. The x-axis (public transport provision) serves as a proxy for the structural-conscious spectrum, with those on the left representing more structural dependence and those on the right representing more conscious dependence. Meanwhile, the y-axis (deprivation) serves as an indicator for how much of a social priority alleviating the dependence should be: those with greater levels of deprivation should receive more support for an equitable transition.

Figure 1

It is important to note that these four archetypes are not a complete segmentation: there are people within the population who will fit into none of the archetypes. These are assumed to be not car dependent, and thus not of interest for this study. It should also be acknowledged that these archetypes are not a unique way to describe all archetypes of dependence. For example, these archetypes do not explicitly account for those who are structurally dependent due to a personal reason, such as a mobility issue limiting the use of otherwise well-provisioned public transport.

3.1.1 Forced car ownership

Forced car ownership (ForcedCO) is a form of structural car dependence that has particularly important ramifications for equity. As such, the concept of forced car ownership has been explored at length in the literature. Previous studies have defined ForcedCO as a household that owns a car, despite having low income or few economic resources; by this definition, 7–8.5% of car-owning households in the UK may be in this category, either maintaining or adopting a car despite financial difficulty (Mattioli, 2017; Curl et al., 2018). These households were found to most likely include families with children and working adults, those on low-middle incomes, and those with mortgages (Mattioli, 2017). Despite experiencing lower social exclusion and material deprivation than households with no access to a car, ForcedCO households were found to be more likely to experience in-work poverty or fuel poverty.

ForcedCO is often characterized by its self-reinforcing nature, evident in the precarious balance that many working families strike between housing and transport costs; Lipman (2006) found that, in the US, as people move further from urban and employment centers in search of more affordable housing, they then must spend as much if not more on transportation. Rising costs of both transport and housing render increasing numbers of people in ForcedCO as a result. This problem is also marked in the UK, with large swathes of Greater London experiencing both car dependence and high housing costs, with increasing prevalence over time (Cao and Hickman, 2018). The archetype of ForcedCO may be further broken down into sub-groups of people, depending on their financial situation and whether they are maintaining or adopting a new car (Curl et al., 2018).

In this study, ForcedCO is defined in the same way as Carroll et al. (2021): as high levels of single car ownership, high levels of deprivation, and low levels of public transport provision. The use of high single car ownership as a defining characteristic explicitly accounts for those households that cannot afford multiple vehicles, despite potentially requiring more than one.

3.1.2 Rural car commuter

The rural car commuter (RuralCom) archetype is defined by high levels of multiple car ownership, low levels of deprivation, and low levels of public transport provision. Households that fall into this category experience a form of structural dependence on their car due to low accessibility to employment centers and other essential amenities, due to poor public transport provision in suburban or rural areas. However, this group are assumed to have moved to these locations out of choice, to escape the busy city (Morris, 2019). Due to the low levels of deprivation, it is also expected that the rural car commuter may be a person involved in counter-urbanization or ‘rural gentrification', moving away from urban centers despite affluence (Karsten, 2020; Sheppard and Pemberton, 2023). The choice involved in living away from urban centers, as opposed to a forced compromise between housing and transport costs seen in ForcedCO, is characterized through the inclusion of high multiple car ownership. As such, this archetype will likely interact with their cars and respond to incentives in a different way than those in ForcedCO. The RuralCom archetype combines elements of conscious dependency, since people have agency and make choices around their transport options, despite their car use being reinforced by structural dependency. As explored by Handy et al. (2005), choice and necessity become less clearly defined here as long-term choices around home or employment location, for example, feed into car dependency, which implies a lack of choice. Indeed, Steg et al. (2001) found that individuals themselves combine conscious and structural reasons for car use, relying on each when socially acceptable.

3.1.3 City driver

The city driver (CityDriver) archetype is defined by high levels of single car ownership, high levels of deprivation, and high levels of public transport provision. CityDriver differs significantly from both ForcedCO and RuralCom due to the fact that households falling into CD are in areas assumed to have a good provision of alternative modes of transport, meaning they are more likely to be experiencing a conscious form of dependence. For example, these could comprise people for whom driving is a hobby, status symbol, and/or socio-culturally important activity, or people whose employment is directly car-dependent (e.g., taxi driver) or indirectly car-dependent (e.g., tradesperson with tools). However, this could also include people who are unable to make use of notionally adequate public transport, such as somebody with a disability preventing them from using it. This could also include parents with young children, as this demographic is more likely to buy and own a car (McCarthy et al., 2017; Lloyd-Smith, 2020). The combination of good alternative modes and high levels of deprivation means that this archetype is expected to be prevalent in deprived inner city areas.

3.1.4 Urban car commuter

The urban car commuter (UrbanCom) archetype is defined by high levels of multiple car ownership, low levels of deprivation, and high levels of public transport provision. UrbanCom is similar to the CityDriver archetype due to the high levels of assumed alternative modes available, and thus the wide-ranging conscious form of dependence. However, critically, this archetype is characterized by low levels of deprivation and therefore the reasons driving the dependence may be slightly different. For example, driving a car in well-connected areas of a city may be seen as a status symbol or display of wealth, or a way to separate oneself from the general public (Steg, 2005; Sheller, 2004; Wells and Xenias, 2015). This archetype may also comprise more affluent parents with young children, similarly to the CityDriver archetype.

4 Methodology

4.1 Data

The primary dataset used in this work is the 2011 England and Wales Census data (hereafter referred to as the 2011 Census) at the spatial granularity of Lower Super Output Area (LSOA) as a compromise between high spatial granularity and computational efficiency (Office for National Statistics, 2021). While more recent data is available in the form of the 2021 Census, that Census was carried out during a national lockdown, with many respondents furloughed. This resulted in respondents interpreting guidelines on how to answer questions in different ways, generating a mixture of in-pandemic and pre-pandemic behaviors. As such, the ONS explains that 2021 data for characteristics including age, location, and occupation can be assumed to be unreliable and that 2011 Travel to Work areas should be employed for analytical and statistical work since these will be more representative of where people live and work (Office for National Statistics, 2022). As such, the 2011 Census was deemed more suitable for this analysis, despite the age of the dataset, for stability and internal consistency.

Both tabular demographic data and the Travel to Work dataset at the LSOA level were used from the 2011 Census. For the initial mapping of travel behavior trends, the method of travel to work data from the Travel to Work dataset was used. The distinct methods of travel to work given are: work from home, underground/light rail/metro/tram, train, bus/minibus/coach, taxi, motorcycle/scooter/moped, driving a car/van, passenger in a car/van, bicycle, on foot, and other.

To define the dependence archetypes, deprivation data from the Census was used alongside car/van availability from the Travel to Work dataset. The deprivation variable describes the overall trend of whether a household is in one, two, three, or four dimensions of deprivation as measured by the Census; these dimensions are: employment, education, health and disability, and household overcrowding. We additionally used a secondary dataset of public transport accessibility developed by Anejionu et al. (2019). This data (hereafter referred to as the public transport accessibility dataset), based on 2011 LSOA boundaries, describes how many instances of a given type of destination can be reached by public transport within a given amount of time from a given LSOA, based on service frequency and stop location. In this work, we use the number of jobs accessible by public transport within 30 min for each LSOA. The data is well aligned with that of the Census in terms of travel purpose and geographical granularity. A travel time of 30 minutes was chosen to reflect the well-documented effect known as Marchetti's constant, which finds that people, on average, spend one hour of their day traveling to work regardless of speed of travel (Marchetti, 1994).

For the socio-demographic analysis of the archetypes, the following demographic variables from the Census were included: age (by year), gender (male or female), disability (major, minor, or no impact on daily activities), ethnicity, and socio-economic classification. Ethnicity is given as five broad categories: White, Black, Asian, Mixed, and Other. The socio-economic classification is grouped into nine broad categories, from “Higher managerial, administrative & professional occupations” to “Routine occupations” and additionally includes “Never worked & long-term unemployed” and “Unclassified (including students)”. Each variable is given in terms of each LSOA's proportional population per category of each variable.

Alongside the 2011 Census, the Office for National Statistics 2011 rural/urban classification of LSOAs was used (Department for Environment, Food and Rural Affairs, 2021b). This classifies each LSOA in England and Wales on an eight-point scale, describing how urban or rural the area is, based on population size, density, settlement form, and wider context. These classifications range from “Urban major conurbation” to “Rural village & dispersed (sparse)”.

4.2 Initial mapping of trends

We used bivariate choropleth maps to examine how the use of cars for traveling to work varies both spatially and with respect to other demographic variables simultaneously: the two variables examined are deprivation and public transport accessibility to reflect the axes in Figure 1. The level of car use for traveling to work was represented by the proportion of people who use a car/van to get to work as either a driver or passenger. Deprivation was represented according to the proportion of people per LSOA reported as experiencing at least one dimension of deprivation, as defined by the Census. Each variable was categorized into three equally-sized bins labeled as “low”, “medium”, and “high” to represent relative values. This meant that each LSOA was one of nine categories (two variables, each with three categories), each represented by a corresponding color.

4.3 Archetype mapping

In order to dive deeper into these car use trends, the archetypes proposed in Table 1 were analyzed. This analysis employed the methodology developed by Noble et al. (2006) in their seminal work in measuring multiple deprivation at the small area level. The outlined methodology has since been utilized across many major UK datasets including the England Indices of Multiple Deprivation (Ministry of Housing, Communities and Local Government, 2019). Since the present archetype analysis involves the combination of multiple independent domain indicators (defining variables), this methodology was deemed appropriate.

First, each LSOA was ranked according to each variable in the archetype defining criteria, and these ranks were then normalized between 0 and 1. Single and multiple car ownership was defined according to the proportions of households with only one and more than one car, respectively; deprivation was defined according to the proportion of households in at least one dimension of deprivation; public transport accessibility was measured according to the indicators discussed above.

To combine these individual rankings into an overall indicator for vulnerability to a given archetype of dependence, the three individual rankings were each transformed according to Equation 1, where X is the resulting transformed value and R is the untransformed normalized rank. This equation transforms the ranks to an exponential distribution with a range of between 0 and 100.

Transforming the ranks in this way, into a decaying exponential distribution, means each indicator has a range of between 0 and 100, and controls for any cancellation effects when summing or averaging indicators. The constant of 23 is chosen such that an LSOA will end up in the 90^th percentile of overall vulnerability if it ranks highest (100^th) in one indicator and lowest (0^th) in another, rather than canceling out to the 50^th percentile (Noble et al., 2006). The mean of these transformed ranks was then found and divided by 100, giving an overall score between 0 and 1 to describe an LSOA's vulnerability to a given dependency. These relative vulnerabilities could then be mapped to show the geographical distribution of archetype prevalence.

4.4 Demographic analysis of archetypes

In order to understand some of the underlying differences between these archetypes, an analysis and comparison of the socio-demographic makeup of the highest-scoring LSOAs in each archetype was undertaken. The LSOAs with scores greater than 2 standard deviations away from the mean score for a given archetype were considered to be the highest-scoring, and these four sub-populations were compared against both each other and the total population. The demographic variables that were further explored were: rural-urban classification, socio-economic classification, ethnicity, disability, age, and gender. This analysis comprised comparing the proportion of each subpopulation with each other, and with the total 2011 Census population. This was performed by comparing the percentage values and visualizing the results using radar charts and bar charts. The percentage values were also represented as proportional ratios to the corresponding values of the total population to more easily convey how each archetype subpopulation differs from the overall population. It is important to take both the absolute percentage and ratio together, since the scale of absolute difference will determine the ratio value as well as the relative difference.

4.5 Modelling approach

A multivariate logistic regression model was used to analyse the relationship between demographic factors and the likelihood of experiencing a given archetype of car dependence. Importantly, this methodology allows for the analysis of multiple variables at once, to help understand the independent effects of each variable, removing confounding effects. Furthermore, this method provides interpretable effect sizes in the form of odds ratios and average marginal effects with associated statistical significance. In this way, the relationship between socio-demographic variables and car dependence archetype can be more thoroughly understood.

The dependent outcome variable for each model is a binary of whether an LSOA is in the highest-scoring subpopulation for each archetype, as defined in Section 4.4. The following variables were considered for inclusion as independent variables: age, disability, ethnicity, gender, SEC, and urbanism. The “Higher managerial, administrative & professional occupations” SEC category was chosen to represent “higher” SEC classes, and the sum of the “routine occupations” and “semi-routine occupations” categories was used to represent “lower” SEC classes to reduce the complexity from nine categories. Disability that impacted daily activities a little and disability that impacted daily activities a lot were combined into a single disability indicator. Similarly, the rural-urban classification was collapsed to a binary of rural or urban, with categories A1-C2 defined as “urban” and D1-E2 defined as “rural”.

In the raw Census data, each demographic indicator is reported according to its prevalence in the LSOA. These were converted into categorical variables, with values greater than the median designated as “high” and values less than the median as “low”. For all demographic variables, the “low” value was used as the reference category. For urbanism, “rural” was used as the reference category.

To ensure that these variables did not present multicollinearity, the variance inflation factor was calculated, and variables with scores greater than 2.5 were removed. The final list of variables used in the logistic regression models was: age (under 18, over 65), disability, ethnicity (Black, Asian), gender, occupation (managerial and routine), and urbanism. The odds ratios, average marginal effects, and corresponding statistical significance were calculated for each independent variable. For each archetype model, the McFadden's R² value was calculated.

5 Results

5.1 Initial mapping of trends

Figure 2a shows the bivariate map of relative levels of car commuting with respect to relative public transport accessibility. This figure shows that in London and other large urban centers, car commuting is low while public transport accessibility (PTA) is high (in pink). This is expected due to the density of employment and housing in cities. Conversely, large parts of east, north-east, and south-west England and north and south Wales (in teal) exhibit high levels of car commuting but low levels of PTA, suggesting that car commuting is high due to insufficient public transport. These areas are therefore potential target areas for investment in public transport or active travel to encourage mode shift. The combination of high levels of car commuting and low levels of PTA suggests structural dependence may be prevalent in these areas. Interestingly, although few, there are several LSOAs with high car commuting and high PTA in the outskirts of urban centers (in navy). This suggests that a form of conscious dependence may be prevalent in these areas since there are high levels of car commuting despite notionally sufficient public transport accessibility.

Figure 2

Figure 2b shows the bivariate map of relative levels of car commuting with respect to relative deprivation. In this map, London and other urban centers show areas of high deprivation with low levels of car commute (in pink). This can be understood by the often high levels of deprivation found in inner city areas. Areas in the east and north-east of England and south Wales (in navy) exhibit high deprivation with high levels of car commute. These areas, therefore, may be vulnerable to the ForcedCO archetype due to high levels of car use despite high deprivation. In addition, many rural areas across the country (in teal) exhibit high amounts of car commuting but low levels of deprivation. These areas, therefore, may be vulnerable to the RuralCom archetype. Car use for commuting may additionally be influenced by the provision of parking at the home and employment location, which can be expected to be more prevalent in less space-constrained rural areas.

5.2 Archetype mapping

Figure 3a shows the distribution of the forced car ownership (ForcedCO) archetype across England and Wales. Though much of the country exhibits a low ForcedCO score, pockets of east England and central and south Wales stand out as relative hotspots (in yellow and orange). These are similar areas as those identified in Figure 2, thus confirming the expected result that these areas are likely to experience high levels of ForcedCO.

Figure 3

Figure 3b shows the distribution of the rural car commuter (RuralCom) archetype. London and other large urban centers show low vulnerability (in blue) while areas with the greatest vulnerability (in orange and red) are found surrounding urban centers, particularly in areas immediately outside Greater London. This demonstrates the peri-urban nature of the individuals' home location expected to comprise this archetype and highlights the commuter belt areas east of London and surrounding other major urban centers.

Figure 3c shows the urban car commuter (UrbanCom) distribution. Though a less clearly delineated distribution, in general, higher levels of this archetype can be seen in and around large cities and particularly around London. For example, many of the highest-scoring LSOAs can be found in affluent parts of the south-east of England that are well connected to London by rail, including Wokingham, Barnet, St Albans, Wycombe and Surrey Heath.

Finally, Figure 3d illustrates the clearly defined distribution of the city driver (CityDriver) archetype. While most of the area is represented in blue, showing a very low prevalence of the archetype, major towns, cities, and urban centers are clearly defined in green and yellow showing relatively higher levels. In particular, some LSOAs in central London, Birmingham, Bradford, and Leicester show high levels of this archetype (in orange), confirming the expected result that the CityDriver archetype will be found in urban centers. Areas in the east coast of England, similar to those with high vulnerability to ForcedCO, also demonstrate high levels of this archetype, despite low levels of public transport—this may be due to the extremely high levels of single car ownership and deprivation in these areas.

It is worth noting here that between these four archetypes, almost every LSOA experiences a relatively high vulnerability to at least one form of dependence - this is a testament to the high levels of car ownership throughout the study area. It can also be observed that many LSOAs appear to have a relatively high vulnerability to more than one archetype, highlighting the heterogeneity of LSOAs, and thus the potential for local differences in population giving overlapping results.

5.3 Demographic analysis of archetypes

Summary statistics describing the mean, standard deviation, and first and third quartiles for each analyzed variable and each archetype are provided in the Supplementary Material. A detailed discussion of each variable is presented in this section.

5.3.1 Rural-Urban classification

Table 2 reports the percentages of each rural-urban classification for each archetype subpopulation, alongside these values as proportional ratios of the corresponding value for the total population in brackets. Figure 4a is a radar chart visualizing these results.

Figure 4

These results support the spatial distribution shown by the maps in Figures 3a–3d. These results show that the CityDriver archetype is found overwhelmingly in major urban centers, with 78.17% of city driver LSOAs in urban major conurbations compared with 33.15% of overall LSOAs (ratio of 2.36). The RuralCom archetype is found mainly in rural areas, with 46.24% of LSOAs found in rural villages compared with 7.17% of the total LSOAs (ratio of 6.45). The ForcedCO archetype is found more in rural towns with 28.42% and 6.11% of forced car ownership LSOAs classed as rural towns and dispersed rural towns, respectively, compared with the total population's 9.18% and 0.57% (respective ratios of 3.10 and 10.78). The UrbanCom archetype has a less clearly delineated distribution, generally following the average total population, but with slightly greater representation in urban cities and towns (49% compared with 45.52%) and rural villages (12.93% compared with 7.17%), giving a ratio of 1.80 for both. These results therefore generally support the expected geographies of where these archetypes are likely to be found, and support the spatial distribution.

5.3.2 Socio-economic classification

Table 3 reports the percentages of each socio-economic classification for each archetype subpopulation, alongside these values as proportional ratios of the corresponding value for the total population in brackets. Figure 4b is a radar chart visualizing these results.

This analysis reveals clear trends: the ForcedCO archetype is over-represented in the following socio-economic classifications: small employers & account workers (ratio of 1.13), lower supervisory & technical occupations (ratio of 1.23), semi-routine occupations (ratio of 1.27), routine occupations (ratio of 1.37), and never worked & long-term unemployed (ratio of 1.15). In particular, ForcedCO has the highest absolute and relative proportions of those with lower supervisory & technical occupations, semi-routine occupations, and routine occupations compared with all other archetype subpopulations and the total population. This trend is clearly visible in the orange line in Figure 4b. As such, it can be inferred that the the ForcedCO archetype is seen more by those in “blue collar” occupations. It can be said, therefore, that those with these highlighted occupations are more likely to experience a structural form of car dependence, driven by insufficient provision of alternatives coupled with deprivation. These groups are therefore priority areas for interventions and support. However, this could also be explained by these people requiring cars for work directly in a way that may not be alleviated by increased public transport access for commuting, such as delivering items, visiting clients, or carrying equipment. In this way, the limitations of the breadth of the archetype definitions are evident; without more detailed data, it is not possible to unpick the specific causes of travel behavior.

A similar trend can be found for the CityDriver archetype, which is over-represented in the following socio-economic classifications: routine occupations (ratio of 1.04), never worked & long-term unemployed (ratio of 2.70), and unclassified (ratio of 1.63). Moreover, the CityDriver archetype has the highest absolute relative proportions of people in never worked & long-term unemployed, and unclassified groups compared with all other archetype subgroups and the total population. This trend is clearly visible in the purple line in Figure 4b. The high prevalence of those in the never worked & long-term unemployed class may also be explained via a mediator variable; for example, those with disabilities or chronic health issues may struggle to find work and may also struggle to use available public transport. Moreover, long-term unemployed people, such as full-time parents or carers, may require the use of a car for journeys to school or hospital that cannot be completed using the available public transport.

In contrast, the RuralCom archetype is over-represented in higher and lower managerial, administrative & professional occupations (ratios of 1.71 and 1.32 respectively), intermediate occupations (ratio of 1.04), and small employers & account workers (ratio of 1.37). Similarly, the UrbanCom archetype is over-represented in higher and lower managerial, administrative & professional occupations (ratios of 2.03 and 1.40 respectively), intermediate occupations (ratio of 1.03), and small employers & account workers (ratio of 1.03). UrbanCom has the highest proportions of the higher and lower managerial, administrative & professional occupations compared to all other archetype subgroups and the total population, while RuralCom has the highest proportions of small employers & account workers. This trend is clearly visible in the blue and green lines in Figure 4b.

RuralCom and UrbanCom show similar trends in occupation: both have significantly greater representation in higher and lower managerial, administrative & professional occupations, with a stronger effect for UrbanCom. This outcome is expected due to both archetypes being defined by low levels of deprivation. These results also show that those in the aforementioned occupations may be vulnerable to both structural dependence (RuralCom) or conscious dependence (UrbanCom). Greater detail in data is therefore required to understand if there is an association between certain attitudes or characteristics and travel behavior. The high prevalence of small employers & account workers may be partially understood by acknowledging that this class, including agricultural employers and workers, would be expected to be found in more rural areas with poor public transport and own multiple vehicles for use on agricultural land.

5.3.3 Ethnicity

Table 4 reports the percentages of each broad ethnic group for each archetype subpopulation, alongside these values as proportional ratios of the corresponding value for the total population in brackets. Figure 4c is a radar chart visualizing these results.

The results of this analysis show that the ForcedCO, RuralCom, and UrbanCom archetypes all have similar proportions of White people with ratios of 1.07, 1.10, and 1.04, respectively. The absolute values all exceed 90%, implying that these archetypes are experienced mostly by White people, and may be partially explained by the fact that, on average, White people have greater access to cars and vans than Asian and Black people (Department for Transport, 2024). The CityDriver archetype, however, bucks this trend, with a much smaller relative White population (ratio of 0.56) and larger Black (ratio of 4.32), Asian (ratio of 4.22), Mixed (ratio of 2.07), and Other ethnicity populations (ratio of 4.04). This effect can be understood by examining the definition and geographical distribution of the archetype: CityDriver is defined by high deprivation and is found mostly in urban centers. In the UK, while 9% of White people live in the most deprived neighborhoods, this increases to 15.7% of Asian people and 15.2% of Black people, with particularly pronounced effects for people of Pakistani and Bangladeshi heritage (31.1% and 19respectively) (Ministry.3% of Housing, Communities and Local Government, 2020). Meanwhile, urban areas in the UK are comprised of 18.3% ethnic minority groups on average, compared with 3.2% in rural areas (Department for Environment, Food and Rural Affairs, 2021a). Thus, it would be expected that the CityDriver archetype, with high deprivation and high prevalence in urban centers, would have a greater representation of ethnic minorities. The CityDriver archetype is assumed to be a conscious form of dependence due to the high provision of public transport. One potential hypothesis for explaining the relatively high proportion of those with Asian heritage in the CityDriver archetype could be the importance of driving in some subcultures of British Asian and British Muslim culture (Alam, 2016; Peaceophobia, 2025). On the other hand, the effect could also be driven by people in ethnic minorities requiring their cars, despite the provision of public transport; for example, taxi drivers in the UK comprise 53.8% ethnic minorities (compared to 19.9% across all occupations), and are the most ethnically diverse occupation according to Norrie (2017).

5.3.4 Disability

Table 5 reports the percentages of people with disabilities that cause major and minor impacts on daily activities for each archetype subpopulation, alongside these values as proportional ratios of the corresponding value for the total population in brackets. Figure 5 is a bar chart visualizing these results.

Figure 5

ForcedCO shows greater levels of people with a disability compared to the total population for both major and minor impacts on daily activities (ratios of 1.41 and 1.27, respectively). In general, those with disabilities have lower rates of access to a car or van than people without disabilities, and so this result is contrary to the general trends between car use and disability (Department for Transport, 2025c). This may be in part explained by the high deprivation included in the definition of the ForcedCO archetype; those with disabilities or disabled family members are more likely to be in poverty and thus will be more likely to fall into this archetype (Matejic and Clark, 2022).

Conversely, the CityDriver archetype does not show the same trend as ForcedCO, despite also being defined by high levels of deprivation: while CityDriver has slightly higher levels of disability that has major impacts on daily life than the total population (ratio of 1.06), the levels are slightly lower for disability that has minor impacts on daily life (ratio of 0.91).

Both the RuralCom and UrbanCom archetypes display similar trends: the prevalence of disability is less than the total population for disabilities that have both major and minor impacts on daily life. This may be partially explained by the low levels of deprivation in both of these archetypes, thus lower expected prevalence of disability. For RuralCom, this effect is clear for both major and minor impacts on daily activities (ratios of 0.59 and 0.83, respectively). The effect is more pronounced for UrbanCom for both major and minor impacts (ratios of 0.48 and 0.68, respectively).

This analysis shows that the archetypes defined by high levels of deprivation have a greater prevalence of people with disabilities than the general population, and vice versa. However, this effect is partially mediated by the level of public transport accessibility: those with poor public transport show greater disability prevalence and those with good public transport show lower disability prevalence. This highlights the equity-based imperative for action to support those experiencing ForcedCO, since this archetype is disproportionately experienced by those with disabilities who may be vulnerable.

5.3.5 Age

Table 6 shows the proportion of each archetype that falls into each broad age category: under 18, 18 to 65, and over 65 years of age. It can be seen that variation between the archetypes is minimal, with limited variation from the total population. Nonetheless, some trends can be seen when examining the proportional ratios compared to the total population, shown in brackets. Indeed, ForcedCO has a greater proportion of people over 65 (ratio of 1.41) while RuralCom and UrbanCom have smaller proportions over 65 (ratios of 0.59 and 0.48, respectively), and CityDriver has a greater proportion of people under 18 (ratio of 1.18). Thus, the ForcedCO archetype is more prevalent in elderly people, the RuralCom and UrbanCom archetypes are more prevalent in children and working-age adults, and the CityDriver archetype is more prevalent in children. The high prevalence of children in areas with high public transport access (UrbanCom and CityDriver) may again be due to the effect of parenthood causing individuals to buy and own a car.

5.3.6 Gender

In this analysis, gender proportions across the archetypes rendered little variation from the total population, with all values within 1.3% of the total population's proportion (with ratios ranging 0.98 - 1.01), as shown in Table 7. This is likely an artifact of using area-aggregated data, where gender variation across LSOAs is unlikely to be significant.

5.4 Logistic regression

The results of the logistic regression presented in Table 8 reinforce the trends found in subsection 5.3, showing that archetypes of car ownership correlate with demographic characteristics. All four models gave McFadden's pseudo R-squared values 0.181 ≤ R² ≤ 0.361, indicating that the logistic regression models have moderate to good fit.

For the ForcedCO (forced car ownership) archetype, disability, ethnicity, occupation, and urbanism were all significant variables. Those LSOAs with high prevalence of disability and high prevalence of routine occupations have significantly greater odds of experiencing ForcedCO. An LSOA with high levels of disability has 2.18 times the odds of experiencing ForcedCO than those with low levels, all else being equal, representing an average marginal probability increase of 1.7%. Meanwhile, those with greater proportions of people identifying as Black or Asian, with a managerial occupation, or in an urban setting have significantly lower odds of experiencing ForcedCO. An LSOA with high levels of individuals with a managerial SEC has 0.21 times the odds of experiencing ForcedCO (a 3.5% average marginal decrease in probability) and an urban LSOA has 0.19 times the odds of experiencing ForcedCO (a 4.4% average marginal decrease in probability). These results also confirm the result that ForcedCO is less prevalent in urban areas and with people identifying as Black or Asian while those with disabilities are more at risk.

The results for the RuralCom (rural car commuter) archetype model show that LSOAs with high prevalence of managerial occupations have 4.57 times the odds of experiencing RuralCom than those with low prevalence, all else being equal, representing an average marginal probability increase of 4.1%. LSOAs with high levels of under 18s are also more susceptible to RuralCom with 2.36 times the odds of experiencing RuralCom (a 3.3% average marginal increase in probability). Meanwhile, those in urban areas and with a high prevalence of routine occupations have significantly lower odds of experiencing RuralCom: urban LSOAs have 0.18 times the odds of experiencing RuralCom (a 6.5% average marginal decrease in probability) while LSOAs with high levels of individuals with routine occupations have 0.11 times the odds (a 5.3% average marginal decrease in probability). The model also reveals that LSOAs with greater proportions of women, disabled people, and those identifying as Black have significantly lower odds of experiencing RuralCom, while those with greater prevalence of children and the elderly have greater odds. These results also confirm that this archetype is experienced less in urban areas and by those identifying as Black, and those with disabilities.

The UrbanCom (urban car commuter) model tells a similar story to that of RuralCom: LSOAs with greater prevalence of women, disability, people identifying as Black, and those with routine occupations have significantly lower odds of experiencing UrbanCom. Meanwhile, LSOAs with greater prevalence of those identifying as Asian, children, and those with managerial occupations are significantly more likely to experience UrbanCom. Indeed, an LSOA with high levels of individuals with managerial occupations has 22.09 times the odds of experiencing UrbanCom (a 5.2% average marginal increase in probability), while an LSOA with high levels of disability has 0.05 times the odds of experiencing UrbanCom (a 5.5% average marginal decrease in probability), illustrating the significance of these variables. Here, LSOAs in an urban setting are significantly more likely to experience UrbanCom, confirming the mapping results.

Finally, the CityDriver (city driver) model shows that LSOAs with higher prevalence of people with disability, people identifying as Black and Asian, and children have significantly greater odds of experiencing CityDriver. Meanwhile, LSOAs with greater prevalence of women, over 65s, and those with managerial or routine occupations are significantly less likely to experience CityDriver. These results confirm that CityDriver is largely experienced by those in deprived urban areas, such that urban LSOAs have 6.45 times the odds of experiencing CityDriver (a 4.4% average marginal increase in probability).

Our findings demonstrate that the four archetypes encompass large swathes of the population. Women are less likely to experience all forms of dependence. Those with disabilities are more likely to be vulnerable to structural forms of dependence associated with higher deprivation (ForcedCO and CityDriver); this may be partially explained by higher rates of poverty in disabled communities (Matejic and Clark, 2022). Those identifying as Asian are more likely to be vulnerable to more urban (conscious) forms of dependence (UrbanCom and CityDriver), while those identifying as Black are more likely to be vulnerable to CityDriver. This implies a divide amongst ethnic groups, aligning with the general trend that minority ethnic groups tend to live more in urban areas (Department for Environment, Food and Rural Affairs, 2021a). Children are more likely to be vulnerable to car dependence across the board, especially archetypes associated with higher incomes (RuralCom and UrbanCom). Meanwhile, people over the age of 65 are more likely to be vulnerable to RuralCom and much less likely to be vulnerable to CityDriver. This corresponds to the fact that, in the UK, older individuals tend to live in more rural areas and experience less deprivation (Department for Environment, Food and Rural Affairs, 2021a; Joseph Rowntree Foundation, 2024). As might be expected, those with managerial occupations are more likely to experience forms of dependence associated with higher income (RuralCom and UrbanCom) and less likely to experience ForcedCO. The opposite is true for those in routine occupations. Both are less likely to experience CityDriver, however, which may be understood by considering that other occupation types are mostly prevalent in CityDriver, such as those who have never worked, are long-term unemployed, or are unclassified (see Figure 4b).

6 Discussion

In this spatial and socio-demographic analysis of the Census Travel to Work data, a novel set of four archetypes was proposed to unpack the subtleties of why and how different groups of people may rely on their cars for transport. The analysis found that, while dependence in general is widespread, each archetype is more prominent in distinct geographies across England and Wales, and the form of dependence (structural or conscious) and priority level for intervention (vulnerable groups or not) varies across socio-demographic groups. This indicates that policies to encourage mode shift must be geographically and demographically specific.

Rural areas, in general, were found to experience more structural dependence: rural towns experience more of the ForcedCO (forced car ownership) archetype, arising from poor public transport infrastructure and deprivation, and rural villages experience more of the RuralCom (rural car commuter) archetype, arising from low deprivation and potential rural gentrification. The east coast of England and South Wales were particularly vulnerable to ForcedCO, thus these are priority areas for investment into public transport infrastructure. Areas in the south and south-east of England, particularly surrounding Greater London, are particularly vulnerable to RuralCom, indicating that development of public transport for commuting from these areas to urban centers should be a priority. In contrast, urban areas experience more conscious forms of dependence in general, with major urban conurbations experiencing more of the CityDriver (city driver) archetype, arising from high levels of deprivation alongside car ownership.

In addition, this analysis has shown that those with disabilities, in lower socio-economic classifications, and identifying as White are more vulnerable to ForcedCO, and hence would benefit the most from support and infrastructure investment. Meanwhile, those with disabilities, identifying as Black and Asian, children, and those with “lower” occupations are more vulnerable to CityDriver. Further investigation into the reasons behind why these people have high levels of car ownership despite the notional availability of public transport and high levels of deprivation is required; it may be inferred that the reason is partly due to the high levels of disability making the public transport not accessible to these people, though other hidden reasons beyond this may also be at play. The RuralCom and UrbanCom archetypes exhibit similar demographic patterns, though UrbanCom shows greater magnitudes in the logistic regression. These patterns show that men, children, people with managerial professions, people without disabilities, and people who do not identify as Black are more likely to be vulnerable to both the RuralCom and UrbanCom archetypes. Further investigation into why part of this shared demographic group is found in rural areas(RuralCom) and why part is found in urban areas(UrbanCom) is required.

This research used the 2011 Census data which, at the time of writing, is fourteen years old. Despite the sound reasoning behind this choice, it is acknowledged that the transport landscape differs vastly between 2011 and today. In particular, the long-term travel and equity impacts of the COVID-19 pandemic, such as certain professions retaining the ability to work from home, will not be captured by the 2011 Census.

The area-aggregated nature of the data employed generates some interpretability issues in the presented results. This aggregation means that individual circumstances, demographics, or behaviors cannot be extricated, and some effects are likely obscured through aggregation. For example, a vulnerable household living in an LSOA that is otherwise affluent will not be captured, and these micro-level differences will be lost. Detailed analysis for understanding why certain behaviors arise, and how these are linked to circumstances, is not possible; unpicking the drivers of the conscious dependence archetypes, especially is not possible at this scale. An LSOA can exhibit multiple characteristics simultaneously, but it will not be possible to understand whether these are due to the individuals with these characteristics simultaneously, or multiple separate groups of individuals with a given characteristic. This also leads to non-mutually exclusive archetypes, leading to reduced statistical analysis avenues and confounding results. To address these issues, future work could be done to combine area-level and individual-level data, to allow for individual-level insights while preserving nationwide scope.

Nevertheless, this paper reveals insights into the potential equity and justice implications of the current and future EV transition, and highlights where the UK Government's policies might be adapted or extended to support the most vulnerable people. Due to the structural nature of the ForcedCO and RuralCom archetypes, these groups will likely find mode shift difficult and therefore be required to switch to an EV. Thus, areas experiencing ForcedCO (in the east coast of England, mid- and south Wales), and those experiencing RuralCom (areas surrounding major urban centers) are priority areas for public charge point installation and electricity grid reinforcement. However, those in ForcedCO may find it difficult to make the switch to more expensive EVs or have access to at-home charging; this would leave people in these geographies, those with disabilities, and those with routine occupations at a disproportionately greater risk of vulnerability to transport poverty and being left behind by the transition. Therefore, for a just transition, these groups will require greater support to transition away from ICEVs, in the form of both better provision of public transport where possible, as well as greater financial support for purchasing and fuelling an EV for those unable to use public transport. Meanwhile, the UrbanCom and CityDriver archetypes, defined as conscious forms of dependence, could be encouraged to move away from private vehicles toward existing public transport where possible. Further work is required to understand the specific drivers of these apparently conscious forms of dependence, and to understand which, if any, unreported structural variables may be driving this dependence, to be able to provide truly equitable interventions; for example, individuals may experience indirect structural barriers to transport, such as location of employment relative to home location. The CityDriver archetype is more likely to experience externalities such as these, meaning that people with disabilities, children, and those identifying as Black and Asian, will likely require greater support due to higher deprivation levels.

In summary, car dependence and the push toward EVs impacts different groups differently and, without tailored support for specific geographies and socio-demographic groups, there is a risk that large parts of the population will be left vulnerable to transport poverty or entrenched inequality. These heterogeneities must be considered by policymakers at the national level to ensure the transition to a resilient, decarbonised transport system is just for all.

References

• 1

  Active Travel England (2025). Almost £300 Million to Gear Up New Walking, Wheeling and Cycling Schemes. London: GOV.UK.

  • Google Scholar

• 2

  Advanced Propulsion Centre (2025). Automotive Transformation Fund - DRIVE35. Available online at: https://www.apcuk.co.uk/funding/automotive-transformation-fund/ (Accessed September 9, 2025).

  • Google Scholar

• 3

  AlamY. (2016). Automatic transmission: ethnicity, racialization and the car. Identities: Global Stud. Cult. Power25, 284–301. doi: 10.1080/1070289X.2016.1232197

  • CrossRef
  • Google Scholar

• 4

  AnableJ. (2005). ‘Complacent Car Addicts' or ‘Aspiring Environmentalists'? Identifying travel behaviour segments using attitude theory. Transp. Policy12, 65–78. doi: 10.1016/j.tranpol.2004.11.004

  • CrossRef
  • Google Scholar

• 5

  AnejionuO. C.SunY.ThakuriahP. V.McHughA.MasonP. (2019). Great Britain transport, housing, and employment access datasets for small-area urban area analytics. Data Brief27:104616. doi: 10.1016/j.dib.2019.104616

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6

  ArianA.Meiyu PanM.ChiuY.-C. (2021). Personas: a market segmentation approach for transportation behavior change. Transport. Res. Record2675, 172–185. doi: 10.1177/03611981211028623

  • CrossRef
  • Google Scholar

• 7

  BanisterD.BowlingA. (2004). Quality of life for the elderly: the transport dimension. Transp. Policy11, 105–115. doi: 10.1016/S0967-070X(03)00052-0

  • CrossRef
  • Google Scholar

• 8

  BarnesJ. H.ChattertonT. J.LonghurstJ. W. S. (2019). Emissions vs exposure: Increasing injustice from road traffic-related air pollution in the United Kingdom. Transp. Res. Part D: Transp. Environm. 73, 56–66. doi: 10.1016/j.trd.2019.05.012

  • CrossRef
  • Google Scholar

• 9

  BrandC.BoardmanB. (2008). Taming of the few–The unequal distribution of greenhouse gas emissions from personal travel in the UK. Energy Policy36, 224–238. doi: 10.1016/j.enpol.2007.08.016

  • CrossRef
  • Google Scholar

• 10

  BrindleR. (2003). Kicking the habit (Part 1): Some musings on the meaning of 'car dependence'. Road & Transp. Res. 12, 61–73. Available online at: https://www.proquest.com/docview/215247024/citation/9D50C92AB54148B5PQ/1 (Accessed August 20, 2024).

  • Google Scholar

• 11

  BüchsM.SchnepfS. V. (2013). Who emits most? Associations between socio-economic factors and UK households' home energy, transport, indirect and total CO2 emissions. Ecol. Econ. 90, 114–123. doi: 10.1016/j.ecolecon.2013.03.007

  • CrossRef
  • Google Scholar

• 12

  CaoM.HickmanR. (2018). Car dependence and housing affordability: an emerging social deprivation issue in London?Urban Stud. 55, 2088–2105. doi: 10.1177/0042098017712682

  • CrossRef
  • Google Scholar

• 13

  CarrollP.BenevenutoR.CaulfieldB. (2021). Identifying hotspots of transport disadvantage and car dependency in rural Ireland. Transp. Policy101, 46–56. doi: 10.1016/j.tranpol.2020.11.004

  • CrossRef
  • Google Scholar

• 14

  CharleuxL. (2018). Deriving mobility archetypes from household travel survey data. Prof. Geographer70, 186–197. doi: 10.1080/00330124.2017.1338588

  • CrossRef
  • Google Scholar

• 15

  ChatterjeeK.ClarkB.NguyenA.WishartR.GallopK.SmithN.et al. (2019). Access to Transport and Life Opportunities. London: NatCen Social Research. Available online at: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/831766/access_to_transport_report.pdf (Accessed July 17, 2024).

  • Google Scholar

• 16

  ChattertonT.AnableJ.CairnsS.WilsonR. E. (2018). Financial Implications of Car Ownership and Use: a distributional analysis based on observed spatial variance considering income and domestic energy costs. Transp. Policy65, 30–39. doi: 10.1016/j.tranpol.2016.12.007

  • CrossRef
  • Google Scholar

• 17

  CiampoliP. (2024). Survey Respondents Cite EV Purchase Price as Significant Barrier to Ownership. Washington, DC: American Public Power Association.

  • Google Scholar

• 18

  Close Brothers Motor Finance (2025). Insurance Hikes Overtake Fuel Prices as Leading Concern for Motorists in 2025.

  • Google Scholar

• 19

  CooperE.GatesS.GrollmanC.MayerM.DavisB.BankiewiczU.et al. (2019). “Transport, health and wellbeing: An evidence review for the Department for Transport,” in Technical Report, NatCen. London: NatCen Social Research. Available online at: https://assets.publishing.service.gov.uk/media/5dd6b167e5274a794517b633/Transport__health_and_wellbeing.pdf (Accessed July 17, 2024).

  • Google Scholar

• 20

  CREDS (2023). “Fuel and transport poverty in the UK: findings report – CREDS,” in Technical Report. CREDS, University of Oxford, Oxford. Available online at: https://www.creds.ac.uk/fair/fuel-and-transport-poverty-in-the-uk-findings-report/ (Accessed November 21, 2023).

  • Google Scholar

• 21

  Cremer-SchulteM.van WeeB.HeinenE. (2025). Stuck in the driver's seat: a conceptualisation for understanding car dependence and its determinants. Transp. Rev. 45, 173–193. doi: 10.1080/01441647.2024.2430007

  • CrossRef
  • Google Scholar

• 22

  CurlA.ClarkJ.KearnsA. (2018). Household car adoption and financial distress in deprived urban communities: a case of forced car ownership?Transp. Policy65, 61–71. doi: 10.1016/j.tranpol.2017.01.002

  • CrossRef
  • Google Scholar

• 23

  DelboscA. (2012). The role of well-being in transport policy. Transp. Policy23, 25–33. doi: 10.1016/j.tranpol.2012.06.005

  • CrossRef
  • Google Scholar

• 24

  Department for Energy Security &Net Zero (2025). Provisional UK greenhouse gas emissions Statistics: 2024.

  • Google Scholar

• 25

  Department for Environment Food and Rural Affairs (2021a). Rural Population and Migration. London: GOV.UK. Available online at: https://www.gov.uk/government/statistics/rural-population-and-migration/rural-population-and-migration (Accessed September 10, 2025).

  • Google Scholar

• 26

  Department for Environment Food and Rural Affairs (2021b). Rural Urban Classification. London: GOV.UK. Available online at: https://www.gov.uk/government/collections/rural-urban-classification (Accessed September 10, 2025).

  • Google Scholar

• 27

  Department for Transport (2023a). Government Sets Out Path to Zero Emission Vehicles by 2035. London: GOV.UK.

  • Google Scholar

• 28

  Department for Transport (2023b). Transport and Environment Statistics: 2023.

  • Google Scholar

• 29

  Department for Transport (2024). Driving Licences and Access to Vehicles.

  • Google Scholar

• 30

  Department for Transport (2025a). Bus Passengers to Save Hundreds as Government Prevents Cliff-Edge Fare Hike. London: GOV.UK.

  • Google Scholar

• 31

  Department for Transport (2025b). Connectivity Tool.

  • Google Scholar

• 32

  Department for Transport (2025c). Disability, Accessibility and Blue Badge Statistics, England, 2023 to 2024.

  • Google Scholar

• 33

  Department for Transport (2025d). Greenhouse Gas Emissions from Transport in 2023.

  • Google Scholar

• 34

  Department for Transport (2025e). NTS 2024: Household Car Availability and Trends in Car Trips.

  • Google Scholar

• 35

  Department for Transport and Office for Zero Emission Vehicles (2024). Pathway for Zero Emission Vehicle Transition by 2035 Becomes Law. London: GOV.UK.

  • Google Scholar

• 36

  Department for Transport and Office for Zero Emission Vehicles (2025). Discount of up to £3,750 on electric cars set to slash costs for thousands. London: GOV.UK.

  • Google Scholar

• 37

  Department for Transport. (2025f). £10 Million for Cutting Edge Tech to Help Futureproof England's Electric Vehicle Chargepoint Supply and Keep Drivers Moving.

  • Google Scholar

• 38

  DsouzaR. (2024). Barriers to EV Adoption: A Reality Check for the Industry. London: YouGov.

  • Google Scholar

• 39

  Faraday Battery Challenge (2025). Faraday Battery Challenge. Available online at: https://www.ukri.org/what-we-do/browse-our-areas-of-investment-and-support/faraday-battery-challenge/ (Accessed September 8, 2025).

  • Google Scholar

• 40

  FrostS.HobbsM. (2024). Moving Together: A People-Focussed Pathway to Fairer and Greener Transport. London: IPPR.

  • Google Scholar

• 41

  GaterslebenB. C. M. (2020). “Social-symbolic and affective aspects of car ownership and use,” in International Encyclopedia of Transportation 2021, 81–86.

  • Google Scholar

• 42

  Gates (2019). “Transport and inequality: An evidence review for the Department for Transport,” in Technical Report. London: NatCen.

  • Google Scholar

• 43

  GoodwinP. (1995). Car dependence. Transp. Policy2, 151–152. doi: 10.1016/0967-070X(95)96744-6

  • CrossRef
  • Google Scholar

• 44

  Government Office for Science (2019a). “Future of mobility,” in Technical Report. London: GOV.UK.

  • Google Scholar

• 45

  Government Office for Science (2019b). Inequalities in Mobility and Access in the UK. London: GOV.UK.

  • Google Scholar

• 46

  GrajduraS.LanzDuret-HernandezJ.EspelandS.BarajasJ.RowangouldD. (2026). “Having your own vehicle... it changes everything.” Rural car access, (im)mobility, and quality of life. J. Transp. Geography130:104447. doi: 10.1016/j.jtrangeo.2025.104447

  • CrossRef
  • Google Scholar

• 47

  HandyS. (1993). A cycle of dependence: automobiles, accessibility, and the evolution of the transportation and retail hierarchies. Berkeley Plann. J. 8:1. doi: 10.5070/BP38113081

  • CrossRef
  • Google Scholar

• 48

  HandyS.WestonL.MokhtarianP. L. (2005). Driving by choice or necessity?Transp. Res. Part A: Policy Pract. 39, 183–203. doi: 10.1016/j.tra.2004.09.002

  • CrossRef
  • Google Scholar

• 49

  HausteinS.HuneckeM. (2013). Identifying target groups for environmentally sustainable transport: assessment of different segmentation approaches. Curr. Opini. Environm. Sustainab. 5, 197–204. doi: 10.1016/j.cosust.2013.04.009

  • CrossRef
  • Google Scholar

• 50

  HybelJ.MulalicI. (2022). Transportation and quality of life: evidence from Denmark. Transp. Res. Part A: Policy Pract. 157, 107–125. doi: 10.1016/j.tra.2021.12.003

  • CrossRef
  • Google Scholar

• 51

  IvanI. (2001). Tools for Conviviality. London: Marion Boyars Publishing Ltd.

  • Google Scholar

• 52

  JensenM. (1999). Passion and heart in transport — a sociological analysis on transport behaviour. Transp. Policy6, 19–33. doi: 10.1016/S0967-070X(98)00029-8

  • CrossRef
  • Google Scholar

• 53

  JocoyC. L.Del CasinoV. J. (2010). Homelessness, travel behavior, and the politics of transportation mobilities in Long Beach, California. Environm. Plann. A: Econ. Space, 42, 1943–1963. doi: 10.1068/a42341

  • CrossRef
  • Google Scholar

• 54

  Joseph Rowntree Foundation (2024). UK Poverty 2024: The Essential Guide to Understanding Poverty in the UK. York: Joseph Rowntree Foundation.

  • Google Scholar

• 55

  KarstenL. (2020). Counterurbanisation: why settled families move out of the city again. J. Hous. Built Environm. 35, 429–442. doi: 10.1007/s10901-020-09739-3

  • CrossRef
  • Google Scholar

• 56

  KellyJ.ElgowainyA.IsaacR.WardJ.IslamE.RousseauA.et al. (2023). Cradle-to-Grave Lifecycle Analysis of U.S. Light-Duty Vehicle-Fuel Pathways: A Greenhouse Gas Emissions and Economic Assessment of Current (2020) and Future (2030-2035) Technologies. Argonne: Argonne National Laboratory. Available online at: https://greet.anl.gov/publication-c2g_lca_us_ldv (Accessed September 1, 2025).

  • Google Scholar

• 57

  KentJ. L. (2014). Driving to save time or saving time to drive? The enduring appeal of the private car. Transp. Res. Part A: Policy Pract. 65, 103–115. doi: 10.1016/j.tra.2014.04.009

  • CrossRef
  • Google Scholar

• 58

  KiberdE.StraňákB. (2024). “Trapped behind the wheel.” in Technical Report. London: New Economics Foundation.

  • Google Scholar

• 59

  LiS. A.GuanX.WangD. (2022). How do constrained car ownership and car use influence travel and life satisfaction?Transp. Res. Part A: Policy Pract. 155, 202–218. doi: 10.1016/j.tra.2021.11.014

  • CrossRef
  • Google Scholar

• 60

  LipmanB. J. (2006). A Heavy Load: The Combined Housing and Transportation Burdens of Working Families. Washington, DC: Centre for Housing Policy.

  • Google Scholar

• 61

  Lloyd-SmithP. (2020). Kuhn-Tucker and multiple discrete-continuous extreme value model estimation and simulation in R: The rmdcev Package. R J. 12:251. doi: 10.32614/RJ-2021-015

  • CrossRef
  • Google Scholar

• 62

  LovelaceR.PhilipsI. (2014). The ‘oil vulnerability' of commuter patterns: A case study from Yorkshire and the Humber, UK. Geoforum51, 169–182. doi: 10.1016/j.geoforum.2013.11.005

  • CrossRef
  • Google Scholar

• 63

  LubitowA.AbelsonM. J.CarpenterE. (2020). Transforming mobility justice: Gendered harassment and violence on transit. J. Transp. Geography82:102601. doi: 10.1016/j.jtrangeo.2019.102601

  • CrossRef
  • Google Scholar

• 64

  LucasK.JonesP. (2009). The Car in British Society. Calicut: RAC Foundation.

  • Google Scholar

• 65

  LucasK.MattioliG.VerlinghieriE.GuzmanA. (2016). Transport poverty and its adverse social consequences. Proceedings of the Institution of Civil Engineers - Transport, 169, 353–365. doi: 10.1680/jtran.15.00073

  • CrossRef
  • Google Scholar

• 66

  MahieuxA.Mejia-DorantesL. (2017). Regeneration strategies and transport improvement in a deprived area: what can be learnt from Northern France?Regional Stud. 51, 800–813. doi: 10.1080/00343404.2016.1177174

  • CrossRef
  • Google Scholar

• 67

  MarchettiC. (1994). Anthropological invariants in travel behavior. Technol. Forecast. Soc. Change47, 75–88. doi: 10.1016/0040-1625(94)90041-8

  • CrossRef
  • Google Scholar

• 68

  MartiskainenM.SovacoolB. K.Lacey-BarnacleM.HopkinsD.JenkinsK. E. H.SimcockN.et al. (2021). New dimensions of vulnerability to energy and transport poverty. Joule5, 3–7. doi: 10.1016/j.joule.2020.11.016

  • CrossRef
  • Google Scholar

• 69

  MatejicP.ClarkT. (2022). From Disability to Destitution. York: Joseph Rowntree Foundation.

  • Google Scholar

• 70

  MattioliG. (2014). Where sustainable transport and social exclusion meet: households without cars and car dependence in Great Britain. J. Environm. Policy & Plann. 16, 379–400. doi: 10.1080/1523908X.2013.858592

  • CrossRef
  • Google Scholar

• 71

  MattioliG. (2017). ‘Forced car ownership' in the uk and germany: socio-spatial patterns and potential economic stress impacts. Social Inclusion5, 147–160. doi: 10.17645/si.v5i4.1081

  • CrossRef
  • Google Scholar

• 72

  MattioliG.AnableJ.VrotsouK. (2016a). Car dependent practices: findings from a sequence pattern mining study of UK time use data. Transp. Res. Part A: Policy Pract. 89, 56–72. doi: 10.1016/j.tra.2016.04.010

  • CrossRef
  • Google Scholar

• 73

  MattioliG.LucasK.MarsdenG. (2016b). “The affordability of household transport costs: quantifying the incidence of car-related economic stress in Great Britain,” in 48th Annual Universities' Transport Study. Leeds: Bristol Publisher.

  • Google Scholar

• 74

  MattioliG.RobertsC.SteinbergerJ. K.BrownA. (2020). The political economy of car dependence: a systems of provision approach. Energy Res. Soc. Sci. 66:101486. doi: 10.1016/j.erss.2020.101486

  • CrossRef
  • Google Scholar

• 75

  MattsonJ.BrooksJ.GodavarthyR.QuadrifoglioL.JainJ.SimekC.et al. (2021). Transportation, community quality of life, and life satisfaction in metro and non-metro areas of the United States. Wellbeing, Space Soc. 2:100056. doi: 10.1016/j.wss.2021.100056

  • CrossRef
  • Google Scholar

• 76

  McCarthyL.DelboscA.CurrieG.MolloyA. (2017). Factors influencing travel mode choice among families with young children (aged 0–4): a review of the literature. Transp. Rev. 37, 767–781. doi: 10.1080/01441647.2017.1354942

  • CrossRef
  • Google Scholar

• 77

  MinerP.SmithB. M.JaniA.McNeillG.Gathorne-HardyA. (2024). Car harm: a global review of automobility's harm to people and the environment. J. Transp. Geography115:103817. doi: 10.1016/j.jtrangeo.2024.103817

  • CrossRef
  • Google Scholar

• 78

  Ministry of Housing Communities and Local Government (2019). “English indices of deprivation 2019: technical report,” in Technical Report. London: GOV.UK.

  • Google Scholar

• 79

  Ministry of Housing Communities and Local Government (2020). People Living in Deprived Neighbourhoods. London: GOV.UK.

  • Google Scholar

• 80

  MoodyJ.ZhaoJ. (2019). Car pride and its bidirectional relations with car ownership: Case studies in New York City and Houston. Transp. Res. Part A: Policy Pract. 124, 334–353. doi: 10.1016/j.tra.2019.04.005

  • CrossRef
  • Google Scholar

• 81

  MorrisE. A. (2019). Do cities or suburbs offer higher quality of life? Intrametropolitan location, activity patterns, access, and subjective well-being. Cities89, 228–242. doi: 10.1016/j.cities.2019.02.012

  • CrossRef
  • Google Scholar

• 82

  MouratidisK. (2025). Transport and quality of life: the car and its link to subjective well-being, health, and life domains. Transp. Policy168, 101–111. doi: 10.1016/j.tranpol.2025.04.014

  • CrossRef
  • Google Scholar

• 83

  National Travel Survey (2024). National Travel Survey: 2023. London: Department for Transport.

  • Google Scholar

• 84

  NobleM.WrightG.SmithG.DibbenC. (2006). Measuring multiple deprivation at the small-area level. Environm. Plann. A: Econ. Space38, 169–185. doi: 10.1068/a37168

  • CrossRef
  • Google Scholar

• 85

  NorrieR. (2017). “The two sides of diversity,” in Policy Exchange.

  • Google Scholar

• 86

  Office for National Statistics (2021). 2021 Census - Census of Population- Data Sources- Home- Nomis- Official Census and Labour Market Statistics. Office for National Statistics, UK. Available online at: https://www.nomisweb.co.uk/sources/census_2021 (Accessed January 16, 2025).

  • Google Scholar

• 87

  Office for National Statistics (2022). Travel to Work Quality Information for Census 2021. Office for National Statistics, UK. Available online at: https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/methodologies/traveltoworkqualityinformationforcensus2021 (Accessed January 16, 2025).

  • Google Scholar

• 88

  Office for National Statistics (2024). Atmospheric Emissions: Greenhouse Gases by Industry and Gas - Office for National Statistics. Office for National Statistics, UK. Available online at: https://www.ons.gov.uk/economy/environmentalaccounts/datasets/ukenvironmentalaccountsatmosphericemissionsgreenhousegasemissionsbyeconomicsectorandgasunitedkingdom (Accessed January 13, 2025).

  • Google Scholar

• 89

  Peaceophobia (2025). Peaceophobia. Botswana: Common/Wealth. Available online at: https://commonwealththeatre.co.uk/shows/peaceophobia/ (Accessed September 5, 2025).

  • Google Scholar

• 90

  PlyushtevaA.BoussauwK. (2020). Does night-time public transport contribute to inclusive night mobility? Exploring Sofia's night bus network from a gender perspective. Transp. Policy87, 41–50. doi: 10.1016/j.tranpol.2020.01.002

  • CrossRef
  • Google Scholar

• 91

  RAC Foundation (2020). “Transport poverty 2018-19,” in Technical Report. London: RAC Foundation. Available online at: https://www.racfoundation.org/research/economy/transport-poverty-2018-19 (Accessed January 16, 2025).

  • Google Scholar

• 92

  RobinsonC.MattioliG. (2020). Double energy vulnerability: Spatial intersections of domestic and transport energy poverty in England. Energy Res. Soc. Sci. 70:101699. doi: 10.1016/j.erss.2020.101699

  • CrossRef
  • Google Scholar

• 93

  ShellerM. (2004). Automotive emotions: feeling the car. Theory, Cult. Soc. 21, 221–242. doi: 10.1177/0263276404046068

  • CrossRef
  • Google Scholar

• 94

  ShellerM.UrryJ. (2000). The city and the car. Int. J. Urban Regional Res. 24, 737–757. doi: 10.1111/1468-2427.00276

  • CrossRef
  • Google Scholar

• 95

  SheppardD.PembertonS. (2023). The actions of key agents in facilitating rural super-gentrification: evidence from the English countryside. J. Rural Stud. 97, 485–494. doi: 10.1016/j.jrurstud.2023.01.004

  • CrossRef
  • Google Scholar

• 96

  Sierra MuñozJ.DubozL.PucciP.CiuffoB. (2024). Why do we rely on cars? Car dependence assessment and dimensions from a systematic literature review. Eur. Transp. Res. Rev. 16:17. doi: 10.1186/s12544-024-00639-z

  • CrossRef
  • Google Scholar

• 97

  SimpsonA. (2025). Four-in-10 Say They Have No Alternative but to Drive to Work. RAC Media Centre. Available online at: https://media.rac.co.uk/four-in-10-say-they-have-no-alternative-but-to-drive-to-work (Accessed August 3, 2025).

  • Google Scholar

• 98

  SmithW. R. (1956). Product differentiation and market segmentation as alternative marketing strategies. J. Market. 21, 3–8. doi: 10.1177/002224295602100102

  • CrossRef
  • Google Scholar

• 99

  SpinneyJ. E. L.ScottD. M.NewboldK. B. (2009). Transport mobility benefits and quality of life: a time-use perspective of elderly Canadians. Transp. Policy16, 1–11. doi: 10.1016/j.tranpol.2009.01.002

  • CrossRef
  • Google Scholar

• 100

  StaceyT.ShaddockL. (2015). Taken for a Ride - How UK Public Transport Subsidies Entrench Inequality. London: Equality Trust.

  • Pubmed Abstract
  • Google Scholar

• 101

  StanleyJ. K.HensherD. A.StanleyJ. R.Vella-BrodrickD. (2011). Mobility, social exclusion and well-being: Exploring the links. Transp. Res. Part A: Policy Pract. 45, 789–801. doi: 10.1016/j.tra.2011.06.007

  • CrossRef
  • Google Scholar

• 102

  StegL. (2005). Car use: lust and must. Instrumental, symbolic and affective motives for car use. Transport. Res. Part A: Policy Pract. 39, 147–162. doi: 10.1016/j.tra.2004.07.001

  • CrossRef
  • Google Scholar

• 103

  StegL.VlekC.SlotegraafG. (2001). Instrumental-reasoned and symbolic-affective motives for using a motor car. Transp. Res. Part F: Traffic Psychol. Behav. 4, 151–169. doi: 10.1016/S1369-8478(01)00020-1

  • CrossRef
  • Google Scholar

• 104

  StradlingS. (2007). “Determinants of car dependence,” in Threats from Car Traffic to the Quality of Urban Life, eds. T. Gärling, and L. Steg (Emerald Group Publishing Limited), 187–204.

  • Google Scholar

• 105

  SuelE.LynchC.WoodM.MuratT.CaseyG.DennettA. (2024). Measuring transport-associated urban inequalities: where are we and where do we go from here?Transp. Rev. 44, 1235–1257. doi: 10.1080/01441647.2024.2389800

  • CrossRef
  • Google Scholar

• 106

  Sustrans (2012). “Transport poverty in England,” in Technical Report. Available online at: https://www.walkwheelcycletrust.org.uk/media/3706/transport-poverty-england-2012.pdf (Accessed November 7, 2024).

  • Google Scholar

• 107

  The Public Charge Point Regulations (2023). The Public Charge Point Regulations 2023. London,: King's Printer of Acts of Parliament. Available online at: https://www.legislation.gov.uk/uksi/2023/1168/contents/made (Accessed September 8, 2025).

  • Google Scholar

• 108

  Transport &Environment and SKIM (2024). The “Early Majority”: the Next Phase of the EV Transition in the UK. London: T&E United Kingdom. Available online at: https://www.transportenvironment.org/te-united-kingdom/articles/the-early-majority-the-next-phase-of-the-ev-transition-in-the-uk (Accessed September 8, 2025).

  • Google Scholar

• 109

  Transport for London (2017). Transport Classification of Londoners - Presenting the Segments. London. Available online at: https://content.tfl.gov.uk/transport-classification-of-londoners-presenting-the-segments.pdf (Accessed September 2, 2025).

  • Google Scholar

• 110

  Van EenooE. (2025). Car dependence in research: navigating its contemporary relevance. Transp. Rev. 45, 282–300. doi: 10.1080/01441647.2025.2457094

  • CrossRef
  • Google Scholar

• 111

  WedelM.KamakuraW. A. (2000). Market Segmentation, Volume 8 of International Series in Quantitative Marketing. Boston, MA: Springer US.

  • Google Scholar

• 112

  WellsP.XeniasD. (2015). From ‘freedom of the open road' to ‘cocooning': Understanding resistance to change in personal private automobility. Environm. Innovat. Soc. Transit. 16, 106–119. doi: 10.1016/j.eist.2015.02.001

  • CrossRef
  • Google Scholar

• 113

  YurdayE. (2025). Average Cost to Run a Car UK 2025. London: NimbleFins.

  • Google Scholar

• 114

  Zapmap (2025a). EV Charging Statistics. Available online at: https://www.zapmap.com/ev-stats/how-many-charging-points (Accessed September 8, 2025).

  • Google Scholar

• 115

  Zapmap (2025b). EV Market stats 2025. Available online at: https://www.zapmap.com/ev-stats/ev-market (Accessed September 8, 2025).

  • Google Scholar