Night-time mobilities and (in)justice in London: C onstructing mobile subjects and the politics of diference in policy-making

The growing interest in urban night-time economies and night-time transport policies presents an important context in which to examine how mobility justice is conceived and operationalised in policy-making. Literature on transport exclusion and transport justice documents the disadvantages experienced by diferent social groups and advances theoretical frameworks for distributive justice and transport accessibility. However, this literature has rarely considered the politics of whether and how mobility diﬀerence is recognised and planned for in transport policy, including issues of deliberative justice (participation) and epistemic justice (knowledge pro-duction). To address these research gaps, this paper engages with Sheller ' s (2018) theorisation of mobility justice and critically analyses the construction of mobile subjects in policy discourse on night-time mobility. We analyse policy documents part of night-time policy for Greater London to examine the extent to which the diferentiated night-time mobilities across social categories (gender, age, ethnicity, income, etc.) are recognised – in other words, how the ‘politics of diference’ play out in transport policy-making. Findings show that the discursive construction of mobile subjects in London ' s night-time policy distinguishes between workers, consumers, and transport users, yet, these broad categories poorly account for diferentiated mobility needs and practices. Publicly available data on diferentiated night-time mobilities in London does not inform current policy discourse, obscuring disadvantages experienced by diferent groups of people moving through the city at night, and thus limits the capacity of existing policy interventions to address mobility injustices. These indings reairm the need for transport research to move beyond distributive justice and accessibility analysis, towards exploring the potential of thinking about distributive and epistemic justice for challenging the status quo of transport policy.


Introduction
Night-time mobilities are currently receiving increased attention within international policy circuits due to the rise of a night-time economy policy agenda and discourses around nightlife and 24-h cities (Hadield, 2014). In London, this policy agenda has sought to celebrate night-time mobility by providing transport infrastructure facilitating continuous movement of people throughout night-time hours to support the functioning and growth of the night-time economy (McArthur et al., 2019), for example through the recently introduced Night Tube. As part of the 24-h city agenda, many cities are reviewing and debating the adequacy and expansion of night-time public transport provision, yet existing research shows that urban night-time policies generally lack an explicit focus on justice with respect to how diferent people move and experience the city at night, and what their related needs might be (Hadield, 2014;Talbot, 2016;Plyushteva, 2018). Both policy and academic research relating to transport equity has predominately focused on day-time travel. Critical thinking on night-time mobilities is thus urgently needed in response to these policy trends. This paper presents an analysis of current London night-time policies focusing on to what extent diference is acknowledged and planned for in relation to night-time mobility, and what this means for mobility justice.
Within transport geography, a small body of work documents differentiated experiences of night-time mobility. Studies have shown that perceptions of safety shape how diferent groups travel at night, including fear-based transport exclusion during night-time hours for lower income groups (Oviedo Hernandez and Titheridge, 2016) and in particular for women (Abenoza et al., 2018;Yavuz and Welch, 2010). Research has also documented how transport disadvantages based on gender, income and race result from insuicient public transport  (Chandra et al., 2017;Delbosc and Currie, 2011;Rogalsky, 2010). Urban geographers have examined the diferentiated experience of groups moving around cities at night more broadly, including exclusion from nightlife venues and public spaces with respect to gender, class and race (Roberts, 2006;Schwanen et al., 2012;van Liempt et al., 2014;Brands et al., 2015;Talbot, 2016). Other social scientists have explored social representations of night-time mobilities (Beaumont, 2015), including the invisibility of workers performing mobile night-time maintenance and care (Macarie, 2017;Smith and Hall, 2013).
These strands of research provide rich insights on how night-time mobilities are diferentiated. We seek to contribute by exploring how the 'politics of diference' -a term irst coined by Young (1990) -that shapes and generates that very diferentiation in mobilities, plays out within urban policy-making: where aspects of diference are negotiated, recognised or excluded in the process of constructing the (im)mobile subjects for whom night-time mobility policy is formulated. This paper thus asks for whom night-time transport policy is made, and to what degree and in what manner do the diferentiated mobilities and needs of individuals and social groups feature in policy-making? In doing so, we address three facets that are unexplored in the current literature. Firstly, existing research on transport equity examines diference in relation to social categories based on case-speciic contextual analysis or national policy guidance Lucas, 2012;Pereira et al., 2017), yet does not acknowledge that these categories are in fact socially constructed and mobilised in particular ways within policy-making. Our approach provides a complement to positivist studies of night-time transport accessibility, to analyse diference with respect to night-time mobilities from a constructivist perspective. Secondly and on a related point, within transport geography the focus has been on individuals travelling from A to B at night, rather than the broader concept of mobility that includes not only the analysis of movement (A to B), but also the representation of mobilities in society and mobilities as embodied practices (Cresswell, 2006). Third, the transport equity literature has focused on exclusion (Church et al., 2000;Hine and Mitchell, 2001;Lucas, 2012;Lucas and Jones, 2012;Titheridge et al., 2014); in this paper, we seek to contribute to recent eforts to engage with more theoretical perspectives on transport justice (Pereira et al., 2017;Sheller, 2018) that have not yet explicitly addressed night-time mobility.
In this paper, we analyse policy discourses in London as a case to examine night-time mobilities and mobility justice. We draw on Cresswell's (2006) conceptualisation of mobility and Sheller's (2018) theory of mobility justice to explore the value of the new mobilities paradigm (Sheller and Urry, 2006) for transport geography, addressing the need to employ more critical perspectives to question the role and practices of government within the sub-discipline (Curl and Davison, 2014). In this vein, our analysis focuses on the politics of diference shaped by the construction of (im)mobile (night-time) subjects in nighttime policy discourses. We ind that Sheller's (2008) theory can help challenge transport research and policy in thinking about the politics of diference: how we conceptualise diference, i.e. diference in relation to whom, and diference in relation to what, going beyond distributive issues of transport accessibility and, crucially for any analysis of transport policy, understanding the politics of diference holistically to include deliberative (decision-making) and epistemic (knowledge) justice. In doing so, we hope to point to a new avenue along which the debate on transport and mobility justice, on how we might think about diference in relation to justice, can be pushed forward.

Theory
What perspectives have been employed to think about justice in relation to how people move and travel, and how do these perspectives understand diference? Theorisations of transport justice have begun to emerge (Martens, 2016;Pereira et al., 2017), broadly focusing on access to transport. Coming from the perspective of the new mobilities paradigm (Sheller and Urry, 2006), Sheller (2018) has recently put forward a wide-ranging critique of these approaches, proposing an alternative theory of mobility justice as "an overarching concept for thinking about how power and inequality inform the governance and control of movement" (p. 48) that focuses "attention on the politics of unequal capabilities for movement, as well as on unequal rights to stay or to dwell in a place" (p. 26). We unpack this debate, exploring how diference is treated within transport justice and mobility justice approaches along three axes: i) diference in relation to whom, e.g. individuals, groups and social categories; ii) diference with respect to what, i.e. transport versus mobility as the object of justice and dimensions of justice beyond the distributive; and iii) prescriptions regarding how policy and planning should engage with diference and achieve justice. This is followed by a discussion of the paper's analytical focus on the politics of diference within policy discourse.

Making sense of difference
Literature on transport equity explores how diferent social group characteristics such as age, gender, race or income shape individual travel. In two major reviews of this literature, Lucas (2012) refers to 'socially excluded groups' and 'vulnerable groups', while Lucas and Jones, 2012 discuss 'socio-demographic distributional efects' of transport policy. Diference in relation to whom is understood with reference to a range of social groups (e.g. low-income, the elderly) that are recognised to be systematically transport disadvantaged, as per the corpus of empirical evidence. Diference with respect to what focuses on transport accessibility, but also on other forms of exclusion (e.g. fearbased). Recent work has considered socio-spatially diferentiated transport accessibility in relation to theories of justice (Martens, 2016). For instance, Pereira et al. (2017) draw on Rawls' (1999) diference principle to argue that fairest distribution of transport access should be determined by the "maximin criterion: the distribution that maximises, subject to constraints, the prospects of the least advantaged groups" (Pereira et al., 2017, p. 183). To realise justice in practice, they argue that "policies should prioritise vulnerable groups", citing the elderly, disabled, ethnic minority groups and lower-income families (ibid.). Alike the broader transport equity literature, Pereira et al. (2017) conceptualise diference in relation to contextually-determined disadvantaged social groups (whom) and accessibility (what), although this encompasses a broad range of distributive concerns related to the deinition of accessibility as a combined capability. The prescription for how diference can be acknowledged and addressed in transport policy is to: "go beyond conventional transport surveys to capture other factors that shape interpersonal diferences in individuals' accessibility, including people's cognitive and embodied competencies, cultural norms, time constraints, or whether the social environment is free from any kind of harassment and discrimination" (Pereira et al., 2017, p. 186). Sheller (2018) has critiqued this approach for being limited in its understanding of diference. Speciically, this critique posits that a focus on accessibility obscures the embodied nature of mobility and assumes individuals "more or less inhabit the same bodies" (Sheller, 2018, p. 101). We ind that this is helpful in pointing to the need for conceptual nuance and considering everyday experience, when thinking about diference. Sheller advocates for an intersectional perspective, which considers not just social categories such as race, class, gender, age, disability and sexual orientation, but critically focuses on how these intersect. For example, the focus can be on the diferentiated embodied experiences of women who are also non-Western immigrants within Western society (Uteng, 2009). Originating in black feminist activism and theory (Combahee River Collective, 1977/1982Hill Collins, 1990), intersectionality theory has recently been mobilised in mobilities research (Murray et al., 2016;Büscher et al., 2016). From this perspective, conceptualisations of diference must recognise that social categories are socially constructed, and must question how categories are created in the irst place (for instance, questioning gender categories). Some scholars writing on intersectionality argue that social categories are inherently problematic and often belie the concrete social relations for groups they claim to represent (Anthias, 2012), and instead advocate thinking that rejects categories and seeks to deconstruct them to reveal the complexity of individual lives (McCall, 2005). From these perspectives, diference with respect to mobility can thus be understood and analysed in relation to both social categories (and how they intersect at the level of social groups, e.g. immigrant women) and individuals (categories as they intersect for an individual). Sheller (2018) appears to embrace both analytical lenses, referring to 'marginalised groups', 'categories of people' and 'individual bodies', however particularly emphasising the "the bodily [and thus individual] scale" (p. 136) as an important level of analysis. In summary, the question of diference in relation to whom is thus more complex for Sheller and feminist theorists, and the key point is that these approaches bring a nuanced and political emphasis to thinking about diference.

Justice beyond transport accessibility
Sheller's (2018) other critique of the transport justice approach (Pereira et al., 2017), in relation to the what of diference, posits that it focuses too narrowly on transport and distributive issues such as accessibility. Instead, Sheller invites us to look at justice and diference in relation to mobility, as a distinct, but related, concept. Sheller (2018) mobilises Cresswell's (2006) prominent conceptualisation of mobility, which includes three aspects: movement, referring to the displacement of bodies and objects from one place to another; representation, referring to the ideas and meanings of mobility as represented in various media (popular media, policy reports, etc.); and practice, referring to individual embodied experiences of mobility. Movement (travel from A to B) has been the traditional focus for transport geography and planning, thus arguably reducing justice to "a geographical problem involving distance, movement and access" (Hine, 2008, p. 47). A mobilities perspective, on the other hand, would consider a broader range of issues relating to justice: beyond access to transport services that facilitate movement, there are important questions of how mobilities are represented in society and how mobility is experienced by diferent individuals.
Mobility, rather than transport, as the object of justice is linked to Sheller's (2018) argument for the need to think beyond the distributive dimension of justice. This requires "a sliding focus of attention that encompasses distributive concerns, including accessibility, but also opens the debate toward wider concepts such as deliberative, procedural, restorative and epistemic justice" (Sheller 2018, p. 134). It is beyond the scope of this paper to discuss all these dimensions of justice; we focus here on deliberative and epistemic dimensions, which are closely interrelated and extend our thinking about justice to diferential power within the politics of mobility. Deliberative justice relates to issues of inclusion/exclusion in decision-making and participatory processes (Sheller, 2018, p. 77). Related to this, the concept of epistemic injustice (Fricker, 2007) emphasises the importance of knowledge production, communication and comprehension to realise justice. It "involves recognizing and creating new forms of knowledge, new facts, and new ways of reconciling seemingly incommensurable ways of knowing" (Sheller, 2018, p. 821) and questioning who/what is or is not represented in dominant knowledge paradigms. While Pereira et al. (2017) propose transport accessibility planning for and on behalf of disadvantaged groups (seemingly by the state), Sheller's (2018) emphasis regarding how mobility justice can be achieved is on demands for the recognition of diference by disadvantaged groups through their mobilisation. 1 Her focus is on "overturning marginalization and disadvantage through intentional inclusion of the excluded in decision making and elimination of unfair privilege" (ibid., p. 75), through the work of both citizens and policy-makers.

The politics of difference and social categories
There are two reasons we ind that these aspects of the mobility justice approach (Sheller, 2018) 2 have value for thinking about diference and disrupting established transport policy. First, we ind ourselves questioning whether the focus of Pereira et al. (2017) on the distributive justice of transport accessibility constitutes a suiciently fundamental shift in approach to disrupt the status quo of transport policy. The practical application of accessibility planning that is sensitive to socio-spatial diferentiation has been limited despite the existence of appropriate planning tools (Papa et al., 2016). Transport policy continues to collapse the diferentiated nature of mobility to cater for 'universalised' subjects (Hine and Mitchell, 2001;Soja, 2010) through a suite of policy-making tools drawing on neoliberal economics and utilitarian philosophy (Kębłowski and Bassens, 2018), which work against recognition of and thus concern for diference. This argument has been made in previous work on night-time transport policy in London, where a focus on minimising Generalized Travel Cost efectively universalised the mobile subjects in question (McArthur et al., 2019). In the UK, a landmark, research-informed report (Social Exclusion Unit, 2003) launched a promising policy agenda on exclusion-focused accessibility planning in the early 2000s, but has since faltered due to national political changes and the onset of economic austerity (Lucas, 2012). Considering these realities, the call of Pereira et al. (2017) for even more sophisticated forms of accessibility planning incorporating new types of data (consistent with the idea of capabilities) seems far from pragmatic. Arguably, the political dynamics of how transport policy is decided and by whom, including the recognition of diference and the mobilisation of social categories in policy-making, has not been signiicantly disrupted despite the growth of research on transport equity. The practical and ideological obduracy of transport policy and planning frames, tools and practices should not be underestimated -especially as these are embedded in broader power relations. As such, it is crucial to think about deliberative and epistemic justice and the more radical potential of participation and knowledge production to afect political change. The distributive justice of transport accessibility should and will continue to be central concern, but we agree with Sheller (2018) that mobility justice must look beyond this. 3 Decision-making and knowledge production is typically dominated by white men in technocratic professions, while it is clear from research cited in our introductory section that there are disadvantages sufered by women, non-white and lower-income people with respect to nighttime mobility.
Secondly, we would argue that although existing work on transport equity/justice does not lack an empirical or normative commitment to thinking about diference, it nonetheless adopts an insuiciently critical perspective on the politics of diference as they relate to social categorisation and the transport policy process. The social groups in relation to which diference is explored in the literature are treated as analytical categories existing 'out there', rather than as socially constructed categories -for example through policy discourse. The social groups that diference is analysed in relation to appear to be based on the case-speciic context of empirical studies, or on national policy guidance Lucas, 2012) 4 -the latter of which is of course political in nature. Pereira et al. (2017) appear to advocate for contextual analysis of diference, with least-advantaged groups considered depending on the aspect of transport accessibility or transport investment in question. The selection of social groups analysed to account for diference is thus lexible in both of these approaches, which has limitations for consistent, systematic analysis of diference that can dismantle institutional racism, sexism, ageism and ableism. Both of these arguments, regarding the need to look beyond distributive justice and the need to pay attention to the 'politics of difference', can be traced back to the feminist theory of justice developed by Iris Marion Young. Young (1990) argues that 'displacing the distributive paradigm' of justice is crucial because a distributive focus tends to "depoliticize public life" (p. 10) and "ignore the social structure and institutional context that often help determine distributive patterns" (p. 15), and because important concerns of oppressed groups in society demonstrably extend to non-distributive issues such as decisionmaking and cultural representation. Our focus on deliberative and epistemic justice falls in line with Young's emphasis on decisionmaking, although we draw on Sheller (2018) conceptualisations of these dimensions. Young (1990) introduced the concept of the 'politics of difference' to describe social movements emerging in 1960s United States, e.g. for Black, feminist and gay liberation, which asserted "a positive sense of group diference" (p. 167) that questioned the ideals of a uniied polity, of assimilation into a society dominated by White straight men, and of "liberation as eliminating diference and treated everyone the same" (p. 164). Here, we discuss the 'politics of diference' in rather diferent terms -less contextually and more narrowly -in referring to how diference features in policy-making and techniques of government. Young's work nevertheless provides antecedents and broader intellectual context to our arguments, as she also juxtaposes the notion of 'universalised' citizens part of a uniied polity and the recognition of (social group) diference, in relation to both the philosophical foundations of liberal theories of justice (the universalising implications of impartial reason, e.g. Rawls, 1971) and the 'universality' of the bureaucratic rules of welfare capitalist states that attempt to be blind to diference.
Returning to the transport justice approaches discusses above, both the whom and how of recognising and planning for diference appears to rely on the notion of a benevolent state. We would argue that whether concerned with equality or not, the work of government policy always involves the discursive construction of subjects to be governed, which involves naming and categorisation (van Hulst and Yanow, 2015). Specific social categories are thus constructed and mobilised in policy documents and processes, which is how the politics of diference plays out in policy-making. This is central to the way in which transport policy can universalise mobile subjects and render diferences invisible. We thus agree with Sheller's (2018) call a focus on the construction of "diferentially enabled [(im)mobile] subjects" (p. 79) in thinking about mobility justice, which in our view provides an important constructivist complement to positivist approaches within transport geography.

Policy discourse and the construction of mobile subjects
The diferences between the transport justice (Pereira et al., 2017) and mobility justice (Sheller (2018)) approach is summarised in Table 1, along with our modiied approach employed in this paper.
Drawing on Sheller's (2018) theorisation of mobility justice, we move away from the idea of a universalised mobile citizen to examine how the movement, representations and practices (Cresswell, 2006) of night-time mobilities feature in policy discourse, and to what extent and how this discourse recognises diference. As we are analysing policy rather than individual lived experience, we analyse diference in relation to the construction of social categories and the extent to which these are understood as intersecting. This does not discount the value of intersectional thinking that rejects social categories; we see our approach as complementary. Our focus on mobilisation of speciic (inherently limited) social categories seeks to reveal how techniques of government shape whose mobility (at night) is valued, rendered explicit and governable, but also whose mobility is rendered invisible in dominant policy discourses. This allows us to unpack for whom nighttime transport is planned. Following this empirical analysis, we discuss justice and the what of diference, by juxtaposing the construction of night-time mobilities in policy discourse with other data available on distributive issues for night-time mobility in London, as well as by relecting on the deliberative and epistemic issues at stake in the production of this discourse.

Methodology
Our analysis analyses discourse(s) contained in policy documents related to night-time mobilities in London. Speciically, we examine: 1) reference to particular social categories; 2) how movement is captured, mobilities are represented and mobile practices are featured (Cresswell, 2006); and 3) how, in combination, these are used to construct diferent mobile subjects to be governed.
Our analysis reviews the emerging policy agenda and strategies around transport to support the Night-Time Economy for the Greater London metropolitan area. London's night-time transport governance has several distinctive features. London has an integrated transport authority, Transport for London (TfL), under the direct oversight of the Mayor and Greater London Authority (GLA). The actions of the Mayor are scrutinized by the London Assembly's elected members. Over the past ive years, TfL, the GLA, and the London Assembly have commissioned diferent pieces of evidence, and published strategies and reports highlighting a series of needs related to the extended provision of transport options at night. The report Impact of the Night Tube on London's Night-Time Economy put forward the economic case for introducing weekend Night Tube services. This led to the introduction of the 'Night Tube' services in 2016, under Mayor Sadiq Khan. 5 In addition to the introduction of 24-h rapid transit across several lines on Friday-Saturday, London has an extensive night bus network that was substantially upgraded in the early 2000s (McNeill, 2002). Both night-time transport services are visualised in Fig. 1.
In an attempt to comprehensively analyse how night-time mobilities are constructed in policy, 10 policy documents were identiied based on an online search, snowballing through document inter-referencing and our monitoring of the developing policy agenda. The documents reviewed are presented in Table 2. Our case study thus comprises a set of strategic documents spanning economic, cultural and transport policy, 4 Lucas and Jones (2012, p. 12) cite UK Department for Transport guidance on twelve types of "socio-demographic variable for assessing diferential impacts, based on their degree of relevance and ease of data collection", but do not ofer a theoretical or normative position on what social groups should be recognised and analysed. Lucas (2012) discusses a range of groups but similarly does not ofer prescription, citing the national guidance of the UK Social Exclusion Unit (2003). 5 Although preliminary studies and announcements of its introduction were released under Mayor Boris Johnson's leadership. E. Smeds, et al. J o u r n a l o f T r a n s p o r t G e o g r a p h y 8 2 ( 2 0 2 0 ) 1 0 2 5 6 9 but all relating to night-time mobilities, as well as supporting evidence base documents, cited by the strategic documents in reference to background data. The documents were coded using qualitative content analysis, with movement, representation and practice coded using the working deinitions presented in Table 3. Reference to social categories such as gender, race, income, age, disability, etc. we are also coded using the same approach.

Nocturnal mobile subjects
Our analysis found that overall, London night-time policy discourse tends to universalise mobile subjects, and displays very limited recognition of diferentiated mobility needs. All documents use similar language and universalising descriptors for nocturnal mobile subjects such as 'Londoners', 'people' and 'all'. As is displayed in Table 4, the strategic policy documents focus on four types of nocturnal mobile subjects, which are regularly mentioned across all documents: 'consumers', 'workers', 'the immobile', and 'transport users'. These subjects are also 'universalised' in the sense that they refer to groups of people (e.g. consumers and workers) who in reality are diverse (e.g. the different professional categories grouped together under workers). The Greater London Authority (2018) report that serves as the key evidence base for future night-time policy does not discuss diferentiated mobilities in depth, instead focusing on what 'Londoners' (in general) do and want from the night, either on the consumer or workers side. These four types of nocturnal mobile subjects are referred to across the documents with difering frequency, and discourse regarding their mobilities also difer, as summarised in Table 4 in relation to movement, representation and practice.
Universalised consumers (encompassing visitors, Londoners and young people) are the most regularly mentioned subjects across all policy documents. Their mobility is described with reference to ease and speed of travel and shaped by the willingness and capacity to access to 'all London has to ofer' at night. The description of their movements focuses on lows towards nightlife and consumption hubs, in particular in Central London's West End (movement from home to nightlife centres, or from the airport to Central London). Their embodied experience  of moving across the city is represented with reference to consumption (of alcohol, goods or entertainment) as well as safety and ease of travel. The second most represented subjects are those engaged in productive labour at night, and although references to diferent professions (hospital workers, cleaners, taxi drivers, etc.) are made throughout the documents, they are most frequently referred to simply as 'workers'. Representations of their mobility difer from that of consumers. Their movements are described as being more diicult, less smooth. However, documents also recognise that without a night-time workforce, the city would stop functioning at night. Workers' movement is understood generally as occurring to and from work, often from outer boroughs to central London and back. However, the documents do not contain any precise spatial information on workers' night-time mobility patterns, even though London's night-time policies recognise that lowpaid workers tend to live further away from the city centre, where housing tends to be more afordable. Workers' embodied experience of moving across the city is described in reference to personal safety and interactions with other users (such as drunk consumers). A third nocturnal subject is acknowledged, albeit very rarely, across all documents: those who 'stay put' or the 'immobile' nocturnal subjects. Individuals' motives for being immobile is referred to as voluntary, for instance, if people want to go to bed, stay at home -or forced, in the case of disabled and/or elderly people. In those cases, movement is described as impossible and hindered, and the practices of immobile subjects described as diicult, restricted or simply absent, when people decide to stay put. The fourth type of mobile subject referred to is 'transport user', a term used within three editions of the annual 'Travel in London' report, which presents data from transport user surveys. Only one of these, the 2015 'Travel in London' report, contains data on the needs of diferent night-time transport user groups. It emphasises the diferentiated needs of marginalised and vulnerable populations with respect to using latenight public transport and 24-h bus services, such as black, Asian and minority ethnicities (BAME), women, older people, lesbian, gay, bisexual and transgender (LGBT), disabled people and low-income groups. The mobility of these users is represented as diicult and sometimes unsafe. Their movements vary and the report ofers granular information on travel destinations and individual perceptions of travel by public transport, walking or cycling. Safety is described as an important aspect of users' experience with diferent transport modes. The 2016 and 2017 Travel in London reports reviewed are much more general in tone and provide less information about how diferent user groups experience night-time mobility. The 2016 report provides preliminary data on Night Tube use and the 2017 report looks at overall travel trips, with limited exploration of the diferences in night-time mobilities across diferent transport modes. These two reports refer to broad categories of nocturnal mobile subjects: users (2016, p. 69), residents (2017, p. 39), workers (2016, p. 69), young people (2017, p. 39) and in one instance, women (2017, p. 39).
The policy documents overwhelmingly focus on public transport, i.e. the Night Tube and night bus services, with two exceptions: the 2015 Travel in London report discusses perceptions of safety when walking at night, across categories of gender (men/women) and ethnicity (white/BAME), and the 2016 Late Night Travel Options report discusses women's perceptions of safety with regard to private hire. The reasons for this are unclear from the documents, but can probably be explained by the fact that the Night Tube is the Mayor's lagship nighttime transport policy and thus attracts disproportionate attention, as well as the fact that night-time mobility is a nascent area of transport policy.

Social categories and the politics of difference
The social categories referred to across the policy documents analysed are summarised in Table 5. As discussed, the policy documents focus on constructed nocturnal mobile subjects. Social categories only feature in relation to and within the context of these subjects, rather than as a central focus, as they would feature in the case of policymaking that would seek to recognise diference by starting with a systematic analysis of the diferentiated mobilities of diverse Londoners at  Table 3 Working deinitions used to analyse the policy documents.

Aspect of mobility Deinition
Movement Displacement or travel from A to B (Cresswell, 2006) Representation Ideas and meanings about mobility (Cresswell, 2006), e.g. freedom, progress, economic growth, right to mobility, seamless, lexible Practice Embodied practice, i.e. individual experiences and feelings, emotions, bodily sensations (Cresswell, 2006) Nocturnal mobile subject A person travelling during night-time hours (between 6 pm to 6 am according to London's policy documents), including immobility (lack of/inability to travel) and dwelling as part of this travel (e.g. waiting at bus stop).

E. Smeds, et al.
J o u r n a l o f T r a n s p o r t G e o g r a p h y 8 2 ( 2 0 2 0 ) 1 0 2 5 6 9 night. The strategic policy documents refer to four social categories: women, young people, low-income individuals, and the elderly. For consumers, documents discuss how women and young people might be more worried for their personal safety when going out at night. In relation to workers, the documents recognise that low-income workers might ind it diicult to commute at night. Rarely, the documents also describe younger and elderly people as more likely to stay put (be immobile) due to various challenges. Strategic policy visions for nighttime mobility in London thus does recognise uneven mobility, e.g. in relation to speed, safety and diiculty. While this is an encouraging direction for policy, recognition of diference remains relatively marginal within the overall policy discourse. Among evidence base documents, the 2015 Travel in London report is the only document that substantively recognises diference, referring to a wide range of social categories spanning gender, ethnicity, income, sexual orientation, age and disability. In addition, the 2015 Travel in London report recognises that diferent individual characteristics shaping transport access and experience are likely to overlap in determining how people move and experience movement: "age, ethnicity, income and whether a person is disabled are all likely to be interrelated" (p. 21), illustrating an (implicitly) intersectional approach. None of the other policy documents recognises diferentiated mobility as shaped by intersections of social categories.
Surprisingly, the 2015 Travel in London report is never explicitly referred to in the strategic night-time policy documents (e.g. 24-h vision for London). The report reveals that TfL has access to and has regularly been collecting granular data on diferentiated night-time mobility (in this case, in relation to public transport) across a broad range of social categories; since 2015 for instance, TfL has generated and analysed data regarding the travel habits and lived experiences of LGBT groups at night. Yet for some reason, this report and data is not mentioned in the strategic night-time policy documents reviewed, i.e. appears not to have been used in policy-making. Our analysis thus shows that even when such data on mobility diference exists, it may not be mobilised in policy discourse. The politics of diference is thus not only limited to what social categories are referred to in policy, and thus the extent to which diference is recognised, but also extends to questions regarding the politics of data. The latest Greater London Authority report (GLA, 2018) indicates that tracing people's movement at night is a key challenge for policy, compared to the ease with which such data can be collected during the day. The report anticipates that future access to big data will allow the metropolitan government to gather information on "what sort of things people are doing, and where they're travelling to and from" (p. 11). Interestingly, the Greater London Authority report thus cites poor data availability on nigh-time movement as a problem, despite the fact that the 2015 Travel in London report provides plenty of such data.

Mobility within the night-time economy
Our analysis shows that mobility at night is understood through the prism of broad categories used to refer to people's function as part of the Night-Time Economy (consumer/workers) and their ability/willingness to access public transport (transport users/immobile people). There is a clear emphasis on the mobility of consumers and, to a lesser degree, that of workers. The GLA (2018) report focuses on the need for consumers and workers to move across the city to access opportunities along the 24-h cycle, lacking consideration of the diversity within these categories themselves, and lacking detailed analysis of and data on the actual movements of consumers and workers (e.g. beyond references to movements 'to/from work' in the case of workers). Where night-time movements are described, these are understood in relation to the concentration of activities to nightlife 'hubs', and strategies focus on enhancing their connectivity in order to facilitate consumer access. For instance, supporting nightlife hubs in peripheral London Boroughs is presented as key to broaden access to the night-time economy, in line with the understanding of domestic and international visitors coming to enjoy London's 'cultural ofer' as key drivers and features of overall nocturnal movement (GLA, 2018, p. 19). Our review thus shows that discourse relating to night-time mobilities is predominately framed in relation to the economic function of individuals.

Discussion
We have discussed the politics of diference in night-time policy for Greater London, showing that current policy discourse does not recognise diference in night-time mobility to any signiicant degree, instead focusing on a series of constructed mobile subjects. Despite indications that Transport for London collects and has access to data on diference, social categories are not used to inform night-time policy. Analysis of policy documents alone does not shed light on why certain categories, representations and evidence sources were selected. However, the institutional separation of knowledge production and policymaking is relevant: night-time policy was developed by the Greater London Authority, while the majority of travel data is held by Transport for London. Institutional co-ordination and data-sharing between the two bodies may be required to better inform London's transport policy. The inding that nocturnal mobile subjects are constructed predominantly in relation to the city's economy is not surprising given the dominance of the Night-Time Economy imaginary in London policy. In line with the GLA's strategic objectives (2017 report), night-time policy is focused on supporting the growth of particular spaces of nightlife and consumption. The accessibility or embodied experiences of social groups that are excluded from this policy imaginary are minimised or omitted completely. While there is an absence of systematic analysis of night-time mobility patterns, needs and experiences across social categories, London's policy does recognise some politics of night-time mobility (Cresswell, 2010) with respect to women, young people, the elderly and low-income workers. This suggests that night-time policy in London is broadening beyond the boosterist branding exercises exhibited in other cities (Shaw, 2010). However, night-time policy does not deliver actual transport investment to match this (limited) recognition of diferentiated needs in policy discourse, and fails to deliver on distributive justice in practice. Current provision of night-time transport mainly focuses on the extension of weekend London Underground services, while existing day and night bus services are being reduced due to funding cuts (Walker, 2018). It is questionable to what extent the Night Tube serves disadvantaged groups, such as night-time workers who rely on bus services for connectivity and afordability reasons (McArthur et al., 2019).

Table 5
Social categories referred to in policy documents. E. Smeds, et al. J o u r n a l o f T r a n s p o r t G e o g r a p h y 8 2 ( 2 0 2 0 ) 1 0 2 5 6 9 Existing research on night-time mobilities, summarised in the Introduction of this paper, documents uneven mobility in relation to a multiplicity of social categories, e.g. gender, income and race. What changes to the design of infrastructure, streets and public spaces, routing and scheduling of services and provision of information would be undertaken if night-time mobility policy took diference seriously? It would require exploring what moving at night would mean for immigrant women, wheelchair users, the elderly still on their feet, young BAME men and shift workers. Considering existing evidence on fearbased transport exclusion and inadequate infrastructures (e.g. lighting) with respect to walking and cycling at night, it is also problematic that these modes are neglected in London policy compared to public transport. Considering our indings, although GLA cites a lack of data on nighttime movement, it appears far from certain that policy would change should more data become available, as what is currently available on diferentiated night-time mobility is not being used. The question is whether the overarching imaginary of the London Night-Time Economy prevents addressing diference adequately in transport design and investment. The transport justice approach prescribes new forms of data collection to feed into accessibility planning (Pereira et al., 2017). We agree that accessibility planning needs a solid base in quantitative data on accessibility for diferent social categories. However, as per our argument regarding the limits of a distributive focus, we would link the emphasis on new types of data collection to epistemic justice, rather than solely to government planning. Data on movement needs to be combined with a robust analysis of how intersecting social categories shape the embodied experiences of individuals and groups. For example, existing ethnographic work (Macarie, 2017) and civil society-led research (Norman, 2011) highlights how night-time shift workers in London are often disadvantaged as a result of multiple, intersecting characteristics such as low pay, precarious employment and migrant background with weak English language abilities and/or precarious immigration status. Not all knowledge needs to be produced by government: policy-makers can productively use knowledge produced by other actors.
Indeed, this is central to realising epistemic justice: the evidence base for night-time mobility policy should seek to capture the lived experience and knowledge of disadvantaged individuals themselves, which includes qualitative research to complement quantitative data on accessibility (Kwan, 1999). For example, a prominent theme emerging from our analysis was experiences of fear in relation to night-time mobilities, with references to women feeling especially unsafe and afraid to move at night, and issues such as the safety of minicabs, Uber and the risk of sexual violence. Realising epistemic justice requires ensuring that these representations of gendered safety in policy discourses aligns with the lived realities of people with diferent gender identities. The GLA (2018) has taken signiicant steps to address this issue with the production of a Women's Night Safety Charter and a broader Violence Against Women and Girls Strategy, which appear to have been produced through a relatively participatory process. Such eforts need to be more strongly integrated with night-time and transport policy in the future and translate into concrete policy interventions. Participatory knowledge production can inform better policy analysis, for example Plyushteva (2018, p. 10) has highlighted "the need to complicate rigid categories of journey purpose" typically used in transport policy to account for the diferentiated experiences of those moving at night, and to move beyond the consumer/worker dichotomy speciically.
Beyond planning and knowledge production, realising mobility justice is also an issue of participation. We hope we have illustrated how these issues are all interlinked, and lastly, we would like to drawn attention to the broader politics of diference in relation to deliberative justice, which in Sheller's (2018) deinition starts with recognition of the legitimacy of particular groups to participate in decision-making. The overarching policy strategy for London at night has been coordinated and formulated by the London Night Time Commission (LNTC, 2019), a consultative body set-up by the Mayor to provide evidence and advice on night-time policy. However, night-time industries such as hospitality, retail and nightlife venues are over-represented in the composition of the Commission, illustrating a bias towards the consumption spaces of the night-time economy. Realising deliberative justice would require meaningful representation and participation by actors representing a broader range of social interests. Healthcare workers are for example not currently recognised as a relevant stakeholder, despite representing one of the largest sectors of night-time work in London (McArthur et al., 2019). Yet our own engagement with the Royal College of Nursing (RCN) has made it clear that nurses and healthcare workers in London struggle with transport access (RCN, 2016), which is highly relevant to thinking about diference as BAME and women professionals are vastly overrepresented in the sector (RCN, 2017). In a promising change of direction, the latest LNTC report (2019) recommends the establishment of a Late Night Transport Working Group, 6 and the RCN (2019) is now asking for a seat at the table, along with the implementation of the Commission's idea for a multimodal, subsidised fare for night-time workers. The idea of nurses as an irrelevant social category in relation to night-time mobility at irst glance, points to the need to constantly relect on how to capture important facets of diference in relation to mobility (in)justice in the city, and remain open to political mobilisation and participation by groups representing diferent interests. Relecting on social categories, we thus ind that having a rigid list of categories that matter for a speciic issue can also be counterproductive.
In summary, we have argued that to realise mobility justice with respect to night-time mobilities: distributive justice requires not only access to data on movement and new types of data on embodied practices, but also the disruption of policy imaginaries; epistemic justice requires the recognition and use of knowledge produced outside the state, which relects people's lived realities and extends beyond the quantitative; and deliberative justice requires recognising the legitimacy of a range of social interests and groups, beyond preordained categories of relevance. We hope we have illustrated how attention to epistemic and deliberative justice hold potential for disrupting the status quo of mobility policy-making.

Conclusion
This paper sought to expand the debate on transport justice by exploring the value of Sheller's (2018) theory of mobility justice for thinking about diference. We have contributed both with empirical material on night-time mobilities, as well as a conceptual focus on the construction of mobile subjects through analysis of policy discourse. It can be concluded that while the exact drivers determining how the politics of diference has played out in London night-time policy remain opaque, the extent to which London metropolitan governance recognises and addresses diference does not amount to mobility fairness. In this case, government knowledge production obscures diferentiated mobility experiences and needs. Looking at policy-making in other cities is fruitful for thinking about what kind of approaches are possible. Nocturnal policy-making in Paris has included a set of experimental interventions to improve night-time cleaners' working conditions (Mairie de Paris, 2010), whereas the City Council of Barcelona has funded "a research project-cum-participative feminist initiative… to analyse how urban planning in Barcelona's metropolitan area impacts on the everyday lives of women who work at night" (Ortiz Escalante, 2018). These examples, as well as our case of London, illustrates that the politics of diference plays out diferently in diferent cities, and can be highly contextualised at the urban scale of governance. Imaginaries and discourses regarding night-time mobility produced by policy-6 See Recommendation 9, p. 71 making elites can have wide-ranging impacts: for example, Mexico City's recent restructuring of its Nochebús night bus service cites London's night-time policy strategies and the positive economic impact of the Night Tube, while it unclear whether this included any socially diferentiated analysis (Laboratorio para la Ciudad, n.d.). 7 Planning for diference in urban mobility is of course a critical agenda also for a rapidly urbanising Global South(s) (Levy, 2013), and indeed it must be underlined that context-sensitive and geographically-speciic analysis of how mobility is shaped by intersections of income, race, gender, age etc. is critical. The ways in which the construction of mobile subjects in policy-making, and the policy imaginary surrounding the urban night, are associated with ideas of urban competitiveness -or modernity, progress, sustainability -and what this means for the mobility of those in the city who are not white, economically privileged men is an important research agenda across the Global North(s) and South(s).
Relecting on our analysis, a theorisation of justice restricted to physical movement and transport accessibility -in other words, the predominant approach within existing transport literature -would not have allowed us to account for mobility (in)justice to the full extent that a mobilities approach has. We thus ind that Sheller's (2018) emphasis on nuanced thinking about diference and mobility justice beyond distributive issues has much value to ofer, in the vein of foundational feminist theorisations of justice (Young, 1990). Compared to existing approaches, the challenge with mobility justice will be to ind ways to translate aspects of this novel approach into actionable knowledge for policy. Having said this, in resonance with our experiences of engaging with London policy-makers on night-time mobility, we have argued that there is much value to constructive academic critique in spurring government action towards mobility justice. Transport research can beneit from more critical policy analysis, as well as partnerships between researchers and social movements to mobilise as part of broadbased campaigns for the right to mobility (Verlinghieri and Venturini, 2018).

Declaration of Competing Interests
None.