Exclusion in the interests of inclusion: who should stay offline in the emerging world of online justice?

ABSTRACT England and Wales are in the middle of an ambitious court reform programme, a key element of which is the shift to more online hearings in appropriate cases. This raises a series of new questions for the judiciary, not least of which is whether there are circumstances in which a video hearing is unsuitable because a key participant is not able to engage effectively online. This article considers current thinking about the circumstances in which a case should be excluded from the list of online proceedings and compares judicial approaches to what we know of digital disadvantage from the social science literature. The authors draw on emerging judicial statements about threshold competencies, and original research with court staff, regular participants in court hearings and lay users. It is argued that the complex dynamics of digital disadvantage are frequently misunderstood and underestimated. This article makes clear the need for a more in-depth consideration of the multiple ways in which digital disadvantage manifests itself beyond a lack of equipment or skills. In doing so it raises critical questions about what we mean by user perspectives and how the voices of users are being heard.


Introduction
Video hearings are nothing new in the English and Welsh justice system. In common with a number of other jurisdictions, HM Courts and Tribunals Service (HMCTS) has been experimenting with online hearings for some time and there is a growing body of socio-legal research on the topic (Mulcahy 2008, Terry et al. 2010, Ellison and Munro 2014, McKay 2018, Rowden and Wallace 2018, Fielding et al. 2020, Sanders 2021. In some proceedings such as those involving vulnerable witnesses in rape and abuse cases, or large-scale commercial cases, online participation in 'hybrid' hearings in which some people appear via video link, has even become the norm. The development of a customised video hearing platform for HMCTS, 1 and their plans to expand the use of both fully online and hybrid proceedings, reflects their ongoing commitment to 'dispersed' proceedings (MacDonald, Mr Justice 2020). 2 This article moves away from the focus of much of the existing literature on the dynamics of video hearings and their impact on the outcome of hearings. Instead it considers the related but discrete issue of how decisions are being made about whether an all-party video hearing is suitable in the first place and what factors might need to be taken into account in making these decisions.
Plans for a staggered expansion of the number of video hearings were severely disrupted by the coronavirus pandemic meaning that the cautious approach previously adopted by the court service to the introduction of video hearings had to be abandoned in the interests of avoiding a severe backlog of hearings (Courts and Tribunal Judiciary 2020g, Judiciary of England and Wales 2020, Courts andTribunals Judiciary 2020b, 2021). While some courts stayed open for judges, court staff and some in-person hearings, emergency conditions dictated that a wide range of proceedings in the civil and criminal justice systems were forced to go online using video or telephone links. The conditions in which this happened were far from ideal. 3 However, if any good has come from the pandemic, it is that our evidence base for evaluating video hearings has expanded and HMCTS, the judiciary and the legal profession have gained a vast amount of experience of them in a very short space of time. This adds to existing evidence provided by the evaluation of pilot schemes (see for example Fielding et al. 2020, Rossner andMcCurdy 2020).
Significantly, it now seems unlikely that the administration of justice and attitudes towards it will ever return to what they were before the pandemic. In this article we argue that there is still much work to be done in developing evidence-based policies and practices about video hearings in cases involving the most marginalised and disadvantaged, more specifically the digitally disadvantaged. The authors seek to promote debate about a question which is increasingly important in the post-pandemic justice system: how do we know who to exclude from video hearings in the interests of inclusion and effective participation? The concept of effective participation that we work with is one which sits at the nexus of the concept of the legal notion of due process and the sociolegal idea of procedural justice. In doing so we draw on ideas around creating a level playing field by allowing people access to advice and support; giving them ample opportunity to prepare for a hearing by providing information about the issues in dispute, evidence and format of proceedings; actively encouraging them to engage in discussions at the hearing by presenting, questioning and correcting what is said; and engendering a sense of involvement.
In seeking answers to the question of who should be excluded in the interests of inclusion, it is important to acknowledge that digital poverty remains a significant issue in the UK and elsewhere (Allmann 2022), with national surveys and indices highlighting that there are a considerable number of people with low digital engagement and a lack of digital skills (ONS 2021, Lloyds Bank 2021a. Digital disadvantage is a very complex issue that is often understood in relation to the presence of hardware, skills and connectivity, but is far more complicated than this simple trinity suggests, not least because it often coincides with, and relates to other forms of disadvantage (Ofcom 2022). The issue of how we define digital disadvantage, or understand how multiple forms of disadvantage intersect with each other, is clearly important in an era in which digital by default is increasingly becoming the norm in the provision of government services.
The data on which this article draws was collected in the course of an ESRC funded research project on video hearings which was launched during the pandemic . 4 It was one of a series of projects funded under the UKRI Ideas to Address COVID-19 scheme with a view to encouraging academics to help government find solutions to the unique set of problems they faced during the pandemic. This research drew on the expertise of the authors on researching lay experiences of legal systems and state bureaucracy and its key output was five public information films designed to help the laity prepare for online hearings. 5 In the course of the research, the question of whether there were certain categories of people who were poorly placed to engage effectively online became critical to the design of the films. Understanding our audience was essential for the research team to determine what content should be included, the level of understanding or skills that the films should be targeted at, and the types of solutions which might realistically be suggested. However, we soon discovered that detailed answers from key stakeholders about who should be excluded from online hearings were either not forthcoming or lacking in depth.
In the sections which follow we start by reviewing guidance from HMCTS and the judiciary and what it says about the types of people and cases that are not suitable for video hearings. We go on to present data from a national survey which asked professional users of the justice system to reflect on the types of people who struggled to engage effectively online. In our discussion of the data we compare the views held by professionals to the insights offered by studies of digital disadvantage undertaken by social scientists. Reflecting on these various sources we make two key arguments. Firstly, much of the judicial and HMCTS guidance reflects a pragmatic response to the rapid changes which took place during the pandemic but is lacking in the sort of detailed appreciation of the nature of digital disadvantage which can help judges to make evidence-based decisions about whether to proceed with a video hearing. Secondly, we argue that many of the insights that emerged from the survey data, especially from those working most closely with disadvantaged users, show a clear awareness of the problems of effective engagement, most notably the intersectional nature of disadvantage. We conclude by reflecting on the limitations of existing data on digital disadvantage in a legal context and argue that understandings of how it operates could be deepened if more attention was paid to the phenomenology of online encounters.

Who makes decisions about online hearings and on what basis? 6
The primary responsibility for ensuring that a trial is fair and that all the participants can engage in proceedings by receiving information, following what is going on, asking questions and making their case, lies with the judge. In Re A [2020], the President of the Family Division was keen to stress three 'cardinal points with the utmost emphasis' in relation to the management of such considerations (para 2). The first of these, and the most relevant in the current context, was that the decision whether to conduct a remote hearing was at the discretion of the judge or magistrate conducting the case, 7 paying attention to the principles of fairness, justice and, in that particular case, the need to promote the welfare of a child. It has also been made clear that party preference should be taken into account by the judge. In a message to all circuit and district judges in April 2020, the Lord Chief Justice, the Master of the Rolls and the President of the Family Division asserted that if all the parties opposed a remotely conducted final hearing then that would be a very powerful, though not decisive factor against holding one (Courts and Tribunals Judiciary 2020a, 2020d), though recent research has cast doubt on the extent to which lay participants are regularly involved in the decision making (Equality and Human Rights Commission 2020).
By far the most detailed national advice available to the judiciary in deciding whether an online hearing is appropriate, can be found in Appendix E of the Judicial College Equal Treatment Bench book produced in the middle of the pandemic in 2021. This acknowledges that litigants are not always willing to disclose the problems they face and that some forms of disadvantage which may impact on the ability to participate effectively may be difficult to identify. This guidance is particularly useful when discussing how to identify cognitive, mental health or neurodiversity issues, but also acknowledges that recent migrants, the domestically abused or the impoverished may have difficulties finding a quiet and private space from which to take part in an online hearing. Disparities in access to the internet are also emphasised including the lack of availability of a stable internet connection, lack of phone credit, the inability to charge devices and limited data allowance (Judicial College 2021).
Elsewhere, doubt has also been cast on the value of such generic guidance in supporting case by case decision making by the judiciary. Guidance issued by Lord Chief Justice, Master of the Rolls and President of the Family Division (Courts and Tribunal Judiciary 2020) argued that national guidance is a blunt instrument which cannot take any account of the strengths or weaknesses of local judicial, staff and technological resources in responding to challenges faced by litigants. In line with this stance, Re A [2020] made clear that guidance issued by the senior judiciary as to the cases that might or might not be suitable for a remote hearing provided no more than illustrations which could be used to support the judge or magistrate in coming to their decision. 8 One way the gap between national guidance and local practices has been bridged is through the production of jurisdiction specific guidance which can take account of the types of litigants, cases or resources within particular parts of the legal system such as family courts, employment tribunals or chancery courts. This was the predominant approach adopted during the pandemic in which the judicial leads published extensive guidance for judges and court staff working in their jurisdiction. A content analysis of this documentation reveals that much of it was perfunctory; limited to giving instructions about such things as how to access an online system, the need to test out equipment in advance and the necessity of muting microphones when not speaking. The level of sophistry apparent in the Equal Treatment Bench Book (2021) does not appear to have percolated down to the level of jurisdiction specific guidance or to have been referenced in it. In the limited situations in which judicial advice considered the circumstances in which an online hearing may be inappropriate three themes emerged. In descending order of the importance attached to them, these are the characteristics of the case, the type of litigants involved and the lack of adequate technology.
Opinions about the type of cases that should remain offline varied considerably. There was a view that there are certain types of cases involving the deprivation of liberty, the compulsory removal of children, the disciplining of litigants or jury trials which require the full power of the state to be exposed, are enhanced by being held in a court building (Marks 2016, MacDonald, Mr Justice 2020, Courts and Tribunals Judiciary 2020a. Guidance has also warned against ambitious remote listing, especially in cases in which there is hotly contested evidence (Clarke 2020, Courts and Tribunals Judiciary 2020a). These cases can be characterised as severity exclusions, though it has also been acknowledged that the seriousness of the case can also be a reason to arrange an online hearing as a matter of urgency (see Re A 2020, MacDonald, Mr Justice 2020, Courts and Tribunals Judiciary 2020a). At the opposite end of the spectrum the judiciary has identified instances in which holding hearings online is to be encouraged because the hearing is likely to be short, interlocutory or involve only case management issues. A more liberal approach to the suitability of an online hearing has also been discussed where a final hearing is being heard on the basis of written submissions only, no evidence, the factual issues are limited or a point of law is being decided (Judiciary of England and Wales 2020, MacDonald, Mr Justice 2020).
The available guidance also makes distinctions about the suitability of online trials depending on the type of people taking part. Re A [2020] stressed the need to consider the ability of a lay participant to engage with, and follow, remote proceedings meaningfully; their familiarity with the necessary technology; whether they have instructed lawyers; and their intelligence and personality. In a family law context, it has been suggested that in all cases where the parents and/or other lay witnesses are to be called, the case is unlikely to be suitable for remote hearing (Courts and Tribunals Judiciary 2020a, 2020c). Conversely, it has been suggested that cases involving only expert witnesses are more likely to be suitable for a remote hearing (Judiciary of England and Wales 2020, MacDonald, Mr Justice 2020).
Much less attention has been paid in jurisdiction specific judicial guidance to the particular types of vulnerability that render online justice problematic but there were some exceptions to this. For instance, it has been noted that particularly careful consideration will need to be given to any remote hearings involving litigants in person, or parties and witnesses for whom English is not their first language (Courts and Tribunals Judiciary 2020a). 9 Guidance from Mr Justice Hayden, Vice President of the Court of Protection is possibly the most useful in this context, though it largely relates to the need to be sensitive to problems faced by litigants during proceedings: 'Judges and advocates must be alert to the potential for litigants in person to be left behind in a discourse which may be less apparent to the judge on a remote link than if they were physically present with the litigant in court. At the start of a remote hearing a mechanism by which a litigant in person may indicate to the judge a lack of understanding or need to interrupt ought to be agreed upon and explained.' (Judiciary of England and Wales 2020a p.14) And later in the same guidance when discussing how litigants might be supported he asserted: 'I recognise that there are potential impediments to an intermediary effectively supporting a litigant's understanding and communication when they are not in physical proximity to each other; subtle facial and emotional cues signalling a lack of understanding, increasing distress or fatigue may not be picked up on early enough on remote video and audio feeds. That having been said the vulnerabilities of a litigant ought not to become an impediment to them to receiving a swift determination of their case, particularly so where imaginative considerations can be applied.' (Judiciary of England and Wales 2020a p. 18) 10 In a similar vein, the Lord Chief Justice also observed in March 2020 that it is very unlikely that a telephone hearing would work if a litigant in person is homeless; leads a chaotic life because of alcohol or drug abuse; has learning disabilities; is living with significant mental health issues; or has other needs or disabilities which would militate against effective telephone hearings (Courts and Tribunals Judiciary 2020b). This more nuanced approach to judicial guidance is also apparent in the advice provided by Judge Barry Clarke (2020) in an employment context which refers to the personal circumstances, disability or vulnerability of participants, familiarity with technology and the ability to understand the proceedings. However, most jurisdiction-specific guidance avoids definition or discussion of 'digital disadvantage'. Indeed, the general focus of jurisdiction specific guidance is on supporting the technological capacity of the judiciary and HMCTS staff (see for example HMCTS 2020bb, Judiciary of England and Wales 2020).
The absence of detailed consideration may well come from a lack of experience. The guidance issued by the Lord Chief Justice, Master of the Rolls and President of the Family Division (Courts and Tribunal Judiciary 2020) has acknowledged that prior to the pandemic there was a common understanding that hearings with the judge and all participants attending remotely, were unlikely to be appropriate for many cases. In their words: 'That position was justified by a range of reasons based on fairness, justice, the importance of the issues and the ability of the judge to assess the witnesses and lay parties in any setting other than an oral hearing in a courtroom'(p.2).
They added that the advent of the current COVID crisis did not undermine the soundness of that position. However, recent research by HMCTS has found that many judges had little pre-pandemic experience of remote hearings and that a significant percentage of 974 judges (24%) and 307 court staff (15%) surveyed by them would like additional training in how to make decisions about when to use remote hearings (Clark 2021).
Despite the focus in official guidance on the central role of the judge in making decisions about which cases are suitable for online hearings, recent research suggests that the situation is more complex in practice. There is a suggestion in HMCTS research which evaluated the use of video hearings during the pandemic that, despite its obvious limitations, the jurisdictional guidance discussed above sometimes took precedence over the discretion of an individual judge. In the words of the report: 'In some cases, the decision to use remote hearings was driven by a jurisdictional approach whereas in other jurisdictions the judge made the decision' (Clark 2021, p. 15, see also p.38). Moreover, it may well be the case that judges are not making decisions on their own. It is considered good practice for the listing office, clerks and court officials to support the judiciary by considering as far ahead as possible how future hearings should best be undertaken (Courts and Tribunals Judiciary 2020d). Research by Rossner and McCurdy (2020) indicates that although the ultimate decision is formally one for the judge, an initial triage is conducted by court staff in some jurisdictions based on the responses that litigants make to a pre-hearing questionnaire. The HMCTS evaluation report has also made clear that when making a judgment on a case-by-case basis, judges sometimes consulted with counsel in advance of the hearing. Litigants appear to have been directly consulted much less often.
The recent evaluation of online hearings during the pandemic provides us with important information about the factors that judges are actually taking into account on a case by case basis (Clark 2021). Data collected for this report indicated that the perceived vulnerability of the parties was considered to be significant for 67% of judicial respondents. This compares with 45% who considered the type of hearing to be very significant; 40% the type of case; 36% a party's lack of representation; 33% the length of the hearing; 32% the number of parties involved; 31% the need for cross examination; and 30% the presence of witnesses (Clark 2021). These data reinforce some of the jurisdictional guidance issued, especially in relation to the type of hearing, type of case and procedural complexity. It is significant however that judges at the coalface are much more likely to focus on the perceived vulnerability of the parties than jurisdiction specific guidance. There is however no specific mention of digital disadvantage in the data reported or the ways in which different vulnerabilities intersect with each other. While these data about the use of discretion at local level are an invaluable starting point for future research, there is clearly scope for these issues to be explored in more depth in the interests of promoting research-based decision making. In the words of the Equal Treatment Bench Book (2021): New technology has been developed and acquired at speed, and judges, representatives and parties have had to learn very fast how to access such technology. Judges need to be careful that their understandable focus on managing the technological challenges does not distract them from the equally challenging task of ensuring procedural fairness in this largely new medium. A different judicial skill set is required for remote hearings. (para 6, p. 475) How we do this, is an issue which we begin to turn to in the sections that follow.

Methods
In the course of our empirical work we sought to move beyond official judicial guidance to collect data about the extent to which other repeat players in the legal system were attuned to issues around online participation (see also Finlay 2018, Ryan et al. 2020. Front line workers are for instance, as likely if not more likely than, a judge to interact with the parties before a hearing. Moreover, the need to gauge their views became even more important in light of the research discussed above which indicates that they are involved in initial triaging. Two key methods were used in an attempt to explore the views of repeat players, more details of which can be found in our project report and technical appendix. 11 Firstly, a national survey was distributed to networks of advice workers, clerks, ushers, interpreters, intermediaries, technical support staff and lawyers in the summer of 2021.We contacted a number of national organisations such as the Advice Services Alliance who shared it with their members or uploaded the link to their newsletters or websites. Further circulation was possible with the assistance of our contacts at HMCTS who distributed the survey to four of the working groups they convene such as the Litigant in Person Engagement Group. They also circulated the survey to the suppliers of intermediaries and interpreters, and posted a link to the survey on a Microsoft Teams page for Digital Support Officers. The survey contained a series of questions which were generic, but also asked specific questions for different user types. The questions around litigants' suitability for online hearings was however asked of all respondents who indicated that they had experience of online hearings. Overall the survey received close to 500 responses. Secondly, we conducted a series of focus groups in July and August 2021, with just under 150 members of the public, ushers, technical support staff and workers in the advice sector. These focus groups, comprising up to ten people each, were conducted online with professional users of the court service, and in person, under social distancing conditions, for lay users. 12 Our commitment to exploring the particular needs of the economically, socially and digitally disadvantaged meant that some of focus groups were conducted with users of food banks recruited by the research team (n = 37). All participants in focus groups were asked about a range of issues relating to their perceptions of, and experiences with, online hearings. Feedback was also sought on a prototype version of a public information film produced by the research team for use by lay users of online hearings. This provoked a number of important discussions about the challenges faced by lay users of the justice system.

Understandings of digital disadvantage by other repeat players in the court system
The survey asked a range of questions including whether participants thought that there were any types of people who are not well equipped to participate in video hearings. Of the 356 who responded to this question just over 70% (n = 251) did so in the affirmative. They were invited to write in further details relating to their response and 223 respondents gave 494 separate and detailed reasons why certain types of people and cases posed serious challenges in video hearings. Chart one shows the seven main categories their responses fell into. Overall, the contributions of front line staff reflect a far more nuanced account of the complexities of disadvantage than those that have appeared in judicial guidance. In common with jurisdiction specific directions, seven per cent of survey respondents were concerned that certain types of litigation such as family cases involving safeguarding or the removal of a child, were wholly unsuitable for video hearings. Criminal cases other than 'mention', short plea, sentence hearings and those involving dangerous defendants or long hearings were also cited in this context. Certain types of legal actors such as litigants in person or lay witnesses were listed as being amongst those who would not be able to cope as well in online hearings because of the paucity of support. In addition several mentions were made of children because of their lack of access to private space from which they could engage in a hearing and the lack of technical skills in the very young. As one intermediary argued: 'I work with preschool and primary aged children. I think it's very hard for young children to build a rapport with the judge and the advocates if they are only 'meeting' them over the live-link. I think young children focus and process language better when they are cross examined face to face.' Significantly, solicitors, barristers and judges without the technical know-how to use computer-based systems for online trials were also cited in this context. 13 As Chart One indicates the second most frequently cited category of people who were not considered suitable candidates for online hearings were 'the disabled'. This is clearly an extremely broad label which needs further unpacking. 14 Of the 135 responses in this category the most frequent subcategory related to people with learning difficulties and special educational needs (33%) such as autism or attention deficit hyperactivity disorder. Reference was also made to the ways in which the neurodiverse rely on non-verbal cues that are more difficult to convey on a screen showing numerous faces. This category was followed by references to those who were deaf or hard of hearing (28%). By way of example, one interpreter noted that people who use BSL interpreters would have problems with video hearings because sign language is a 3D language and video works in 2D. Others mentioned that it was difficult for people who needed to lip read to do so efficiently when the screen was split into multiple small faces. Finally, respondents referred to people with mental health or psychiatric problems (22%) who might be more prone to becoming confused or suffer from paranoia linked to technology.
Some of the views expressed are supported by social science data on digital disadvantage. Current statistics suggest that those who are disabled make up a significant proportion of people who are internet non-users (ONS 2019). But while some of the data provided by survey respondents were grounded in credible examples of how disability impacted on participation, other assumptions about 'disability' are deserving of being challenged or considered in more depth. One example in this category was the assertion that those who rely on lip-reading for communication would inevitably experience difficulties. This might be a problem where everyone appears on screen in the same sized window, but it is not necessarily an issue with software programmes that provide larger on-screen windows for those who are speaking. People living with hearing impairments who are comfortable with subtitles could also be accommodated in online trials if the resources were made available. Issues around partial deafness can also be lessened by using volume controls and headphones. In other instances involving a physical disability, it may be that litigants would prefer to attend a hearing online because it would avoid the inconvenience of travelling to a courthouse on public transport or navigating a building once there. It is significant in this context that judges from the Special Educational Needs and Disability tribunal reported that there has been a significant upturn in the young people appearing at online hearings about their needs during the pandemic. They suggested that being able to take part in the hearing from the comfort of their own home caused significantly less anxiety for some neurodiverse individuals than attending an unfamiliar hearing centre (see also Judicial College 2021). It is clear from these examples that, although the respondents to our survey seemed more attuned to issues experienced by the differently abled or disadvantaged, there is still a need to move beyond broad categories such as 'the disabled' to look at the detail of particular conditions.

The particular problem of digital disadvantage
Survey participants were much more likely than judicial guidance to demonstrate an awareness of digital disadvantage and the way it manifests itself. This is an issue which came to the fore of public debate during the pandemic as access to the internet became vital for communicating, ordering products, and accessing services but it remains a pressing issue post pandemic. There is little or no publicly available data on the number of instances in which a remote hearing was declined during the pandemic, but data produced by HMCTS suggests that public users that attended in-person were slightly more likely (16%) than those who attended remotely (11%) to have additional support needs. This suggests that some hearings were changed to in-person if support needs were identified (Clark 2021). However, there is evidence that in some jurisdictions online or telephone hearings were conducted during the pandemic involving vulnerable litigants without the necessary equipment or skills to engage effectively (Ryan et al. 2020).
According to the Office of National Statistics' most recent statistical bulletin on internet users, 6.3% of UK adults have never used the internet (ONS 2021). The widely cited Lloyds Bank Consumer Digital Index (Lloyds Bank 2021) has reported that almost a third (29%) of the UK population has low digital engagement and that 2.6 million people in the UK population of over 67 million are still offline. The Lloyds Essential Digital Skills Report (Lloyds Bank 2021aa, p.10) also calculates that 19% of the population lack even the foundation level essential digital skills. 15 This suggests that the proportion of people identified by the courts during the pandemic as being digitally disadvantaged is low in comparison with the national profile. Low pick up rates are even more evident when one takes into account that vulnerable people such as the poor, people with learning difficulties and mental health difficulties are over represented in some sections of the legal system such as the criminal courts and social security and child support tribunals. It is also worthy of note that the Lloyd's Digital Skills Report relies on data drawn from Lloyd's Bank, Halifax and Bank of Scotland customers, who may not be representative of the most vulnerable who do not use banking services. This suggests that the index may in turn underestimate the proportion of people in the general population with a digital deficit.
The importance of digital poverty in determining whether someone could engage effectively with an online hearing was well understood by survey participants. The detailed responses of 223 survey respondents who wrote about digital disadvantage fell into four main categories. The most frequently cited problem (41%) was 'digital literacy' of the kind examined in the Lloyd's Bank Digital Skills Report, with one front line advice worker arguing that this was often 'close to zero' amongst her clients. Research into digital skills development has illustrated how achieving digital tasks commonly requires complex abstract steps that must be broken down to achieve the final end goal such as setting up an email, or having a digital document in the right format ready to be uploaded (Allmann and Blank 2021, p. 264). Existing guidance on joining an online hearing remain challenging in ways that will exacerbate existing inequalities. It requires litigants to instal or test their ability to participate in the hearing; to ensure that they have a stable connection and a minimum bandwidth of 1.5 Mbps; to share files using links to a Dropbox, One Drive or Google Drive account; to use a side panel for messaging; to mute their microphones when not speaking; to have a 'suitable' device with camera, speaker and microphone; to appear from an environment that is 'suitable' for the provision of remote evidence; and to have a second device on which to navigate the bundle (Judiciary of England and Wales 2020b, Courts and Tribunal Judiciary 2020, MacDonald, Mr Justice 2020, Clarke 2020, Johnston, Judge Sarah 2020, Clements 2021, HMCTS 2020d). Moreover the minimum recommended bandwidth for a successful remote video hearing is 1.5 Mbps in both directions which is a higher standard than upload speeds in the Government's Universal Service Obligation which stipulates a minimum 10 Mbps download/1 Mbps minimum upload (Hutton 2022).
These requirements have the capacity to further alienate those who are already at the margins of technical competence and confidence. Problems cited by survey respondents included relatively simple tasks such as people not being able to log in or not being able to navigate the controls once a hearing had started. Several interpreters drew attention to the ways in which poor digital skills meant that clients in the same room as them would frequently use the mute function when it was their turn to speak. Not having access to electronic devices (32%) also featured strongly in the survey data with several respondents drawing attention to the fact that people may not have access to a desktop, laptop, tablet or even a smart phone. The Adults' Media Use and Attitudes Report (Ofcom 2022) details that an increasing proportion (now 21%) of internet users in the UK now access the internet exclusively via smartphones. One advice worker in our survey mentioned that most of her clients only had a mobile phone to connect to video hearings and that many individuals may be using outdated technology. More recent research undertaken by HMCTS showed that the top three devices used by public users attending remote hearings were mobile phone (61%), laptop (21%) and landline phone (10%) (Clark 2021).
Survey respondents indicated that for a variety of reasons, many court users do not have access to a fast and stable broadband connection but the problems faced by the digitally disadvantaged go beyond not having access to a device or to the availability of adequate technology and bandwidth. Comments about having no internet or poor bandwidth (20%) and limited data (4%) linked the possession of hardware to the capacity to actually connect with it. In recognition of this more general problem there have been calls to ensure more 'meaningful connectivity' (A4AI 2020), though standards about what is meaningful constantly change, as new technologies and systems rapidly outpace older ones. Focus groups with professional users indicated that a lack of familiarity with technology often compounded situations that were already stressful for lay users. With the closing of many face-to-face public services including public libraries and businesses offering free Wi-Fi, the pandemic has demonstrated just how fragile access to the online world can be for some. The post-pandemic cost of living crisis in the UK is likely to further exacerbate these problems at a time when online hearings are on the increase.

Compound disadvantage
The identification of discrete problems associated with not having the necessary equipment, bandwidth, skills or confidence is important, but it is the cumulative effect of compound disadvantage which is most often considered in contemporary discussions of digital disadvantage. Digital disadvantage is likely to be associated with other characteristics commonly connected to structural disadvantage or poverty. Research suggests that the digitally impoverished are also more likely to be older, less well educated, unemployed, disabled and more socially isolated than the norm (Helsper and Reisdorf 2017). For example, the Ofcom report discussed above found that those least likely to have the internet at home were 'those aged 75+ (26%), those in DE households (14%) and those who are most financially vulnerable (10%)' (2022 p.7). In light of this, Helsper devised the term 'socio-digital' specifically to highlight these intersections between the social and the digital (Helsper 2017(Helsper , 2021. Furthermore, research has shown how multiple forms of disadvantage can come together, to create a 'a multiplicative, rather than an additive, effect' (Liu 2021, p. 3091).
Court staff and professionals who engaged with our project clearly appreciated the intersectional nature of disadvantage. For instance, joining the hearing using only a smartphone was often mentioned in relation to a poor and unstable internet connection, which in turn was frequently caused by using a pay as you go service. One advice worker reported that 14% of her clients said they had no access to the internet, and 20% of those with no access also reported that they were not confident using the internet. The multi-faceted nature of the problems experienced by people, and the impact of the recent severe reduction in legal aid in England and Wales, was summed up by one advice worker in the following terms: 'People without the technology to take part i.e. no computer access, or smart phone access, or access but insuff [sic] data for a video hearing, or insufficient literacy or understanding to use the technology i.e. people with vulnerabilities like learning disabilities, brain injuries, mental health problems, sensory problems, or other health problems that limit their understanding or ability to use technology or to engage with HMCTS. Those without access to representation will be particularly vulnerable to the challenges.' Issues around the many-faceted impact of economic poverty also featured strongly in our food bank focus groups where discussions often focussed on the inability of people living in crowded conditions to buy suitable equipment, get online and find a quiet or private space. Those living in modest accommodation or overcrowded shared accommodation such as migrants, asylum seekers, women living in refuges or those without a permanent home have already been highlighted in discussion above but what is significant about the survey responses is that they were seen as being particularly vulnerable to multiple or compound disadvantage including digital poverty. As one advice worker indicated: 'The client group I work with are not well equipped to cope with video hearings. These are those with multiple and complex needs, including homelessness, substance addictions, mental health and offending behaviour.' These findings draw attention to the problems of talking about a digital 'divide', which evokes an image of the haves and the have nots. But social science research is increasingly revealing a multifarious understanding of digital disadvantage behind statistics on whether people have access to the internet, a computer or basic know how. Having access to a device or strong secure bandwidth may mean that online worlds are technically open to people but may not result in extensive or effective online travel. By way of example, Denvir et al.'s (2021) research into digital skills in the legal arena demonstrated that the disadvantaged who have the skills to go online are not as effective as the advantaged in using the internet to find relevant legal information. It is also the case that even among frequent users of the internet we are beginning to see examples of narrow use and device limited literacy (Allmann 2022).
Situational disadvantage is particularly relevant in this context. This concept remains underdeveloped within the academic literature 16 but can be defined as a set of temporary contextual problems which impact on people's ability to use technology effectively in a particular situation. This means that even with the necessary hardware, skills and connection a litigant may still be unable to participate effectively in a hearing. This may be because of the stress caused by the high stakes involved in the outcome; unfamiliarity with legal rules, rituals, language and procedures or the authority figures present. Rossner's (2021) work on the use of online hearings in the Tax Chamber suggested that these factors could be mitigated by the presence of video hearing officers who participated in the tax video hearing pilot scheme. These officers helped litigants test out their equipment in advance and took them through a trial run of the system before the day of the hearing. Unfortunately, Rossner and McCurdy have reported that video hearing officers were not expected to be on hand when the pilot scheme was rolled out more extensively.
In their discussion of the sort of digital support currently offered to users of e-government services, the Good Things Foundation (2020) has argued that the public faces complex barriers to accessing services online and to effective participation. They have argued that they require a package of support that goes beyond the digital aspects of connecting, to also cover their emotional, procedural and legal needs. Their research demonstrated that those seeking advice when accessing HMCTS digital support services often fell outside the anticipated target groups. Some who appeared on the surface to have the digital access, digital skills, digital confidence and capability to access an online service independently, often needed support for other reasons such as previous negative experiences with government services or the stress caused by the situation they needed help with, such as debt, divorce or an application for probate. This brings to the fore the importance of recognising situational disadvantage and suggests that many of the skills required are 'simultaneously basic and advanced' (Allmann and Blank 2021, p. 264).

What are the implications of these data?
In this final section we reflect on whether enough attention is being paid to the needs of the digitally disadvantaged in a justice system that increasingly relies on technology to mediate communications between the participants in a hearing. Modernisation and digitalisation are often heralded as progressive reforms that allow the legal system to catch up with developments in society, but the discussion above illustrates the many ways in which it can make existing inequality worse and fast track a process which leaves people behind. There are clear dangers that 'progressive' discourses of this kind assume that hardware and software are neutral or benign media. Fricker's (2007) concept of epistemic injustice reminds us that, like any other way of organising and prioritising knowledge, software adopts the perspective, prejudices and understandings of the engineers who mould and use it. Epistemic injustice is rendered possible, or is reinforced, when people or their perspectives are misunderstood, misrepresented, marginalised or excluded from debate about and design of systems. Given the frequently symbiotic relationship between structural inequality and digital disadvantage this issue is of particular significance in the context of online hearings (see further Anderson 2012, Fricker 2013. In their discussion of the court reform programme key stakeholders have pledged to place the needs of lay users at the heart of the design decisions now being made about online hearings. There is some evidence of this taking place. Officials within HMCTS who are responsible for developing the new customised Video Hearing (VH) system, 17 have conducted a formal evaluation of all the lessons learnt from the pandemic (Clark 2021). The VH system includes a new user interface system which is easier to navigate and tries to reproduce the concept of a journey to the hearing. There is also new online guidance produced in EasyRead which provides people with advance information about how to check whether the VH system will work on their device before the day of the hearing. The implementation team has expressed the desire to continue to develop products with users in mind, to listen to the sorts of complaints being managed by the technical support team and to capture the right sort of information for the judge to make decisions about whether a video hearing is appropriate (HMCTS 2020c). 18 HMCTS has also recognised that that some users may need some help in accessing digital services and has co-designed, and been piloting, a Digital Support service which is available face to face and on the phone to address these needs (Good Things Foundation 2020).
Despite these many achievements, historical precedents suggest that reforms are often subject to professional capture. In their account of the evolution of courthouse design guides by HMCTS and their predecessors 1970, Mulcahy and Rowden (2019 have demonstrated that those responsible for steering discussions about the layout of physical courts and the facilities and services they should contain never once consulted with the public; preferring instead to prioritise the needs of engineers, security personnel, judges and barristers. The result was that the needs of marginalised professional groups such as interpreters, low ranking judges, and lay users were frequently marginalised. This raises important questions about what methods are being employed to capture the voice of users in the video hearing reform programme or how social science research can be used to inform judicial decision making and the operation of triage procedures. It is clear from the material presented in this article that we need to go beyond the rather simplistic label of 'user' to find a language that adequately reflects how the needs of lay participants differ from those of professional users, and how these needs can be identified and distinguished. When reflecting on the category of user, it is also worthy of note that it is increasingly being recognised that there is a need for more research to be undertaken not just on lay users but the role and needs of proxy users (Harvey et al. 2021, Reisdorf et al. 2021) who frequently complete digital tasks on behalf of others.
Methodological issues also arise about how the voices of lay users are actively sought out; what Fricker (2013) has called the ability to give testimony in the production of knowledge systems. The dangers of a blanket approach which groups users together and privileges some over others is all too obvious. By way of example, in the first stage of an 'implementation review' conducted by HMCTS May-August 2020, interviews were undertaken with 167 individuals from 'key user groups' which included 59 court staff, 50 judges, 44 legal professionals, three support staff and just 11 public users (HMCTS 2020c). Stage two of the HMCTS implementation review took more seriously the need to seek out the views of the public and more members of the public took part in their survey (4,808) or were interviewed (78) than any other category (Clark 2021). 19 The focus on mixed methods is to be welcomed in this context. Although important work has been undertaken on digital poverty within the social science community, to date it has tended to be quantitative and survey based, leaving a lack of qualitative research, including work that investigates lived experience or that is co-produced with those experiencing digital poverty (Allmann 2022).
However, there is less evidence that the research into user perspectives has taken into account the needs of the vulnerable or digitally disadvantaged. Despite having some important things to say about how decisions about the suitability of an online hearing were made, stage two of the HMCTS evaluation used a sample frame for in-depth interviews which only included lay users who had taken part in a video hearing between May and October 2020. It did not include those litigants that the judiciary felt were not able to participate online who would have had a valuable contribution to make to discussions of the accessibility of the system. We learn little about access to justice by focussing our research on those who have successfully navigated a system. Moreover, the push-to-web approach HMCTS adopted for the survey of all users they conducted is unlikely to have been conducive to encouraging the digitally disadvantage to take part. It is worthy of note that the option of a postal survey was withdrawn after a pilot survey because of the low response rate of two per cent (HMCTS 2021aa); a decision that could well have allowed the views of the digitally disadvantaged to be captured. 20 These issues are of direct relevance to debates about epistemic justice because they deny vulnerable groups the opportunity to to testify from a different perspective, challenge prevalent assumptions in system design and influence change. The decision of the authors to conduct focus groups with food bank users was not straightforward (Mulcahy and Tsalapatanis 2021), but it did reflect an attempt to avoid methods that used the very forms of engagement that many who are digitally disadvantaged are unable or unwilling to exploit. 21 Significantly, it also resulted in more offers to take part in focus groups than we had places. The strategy adopted by the authors came from scepticism about the increasingly common practice of recruiting focus groups participants through market research companies. These research organisations are extremely adept at attracting members of the public to take part in research by appointing them to 'panels' that reflect the characteristics required by a sample frame, but are unlikely to include users who lead complex lives, do not have an email address, are not mobile, do not have access to the internet or a permanent residence. It is noticeable in this context that the HMCTS evaluation admitted that the research company sub-contracted to undertake qualitative interviews with victims and witnesses attending hearings remotely was unsuccessful in identifying anyone prepared to be interviewed (Clark 2021).
Engaging with the digitally disadvantaged may well involve the adoption of other approaches which go beyond traditional methodologies or are simply more expensive or time consuming. 'Human' centred design, co-production and participatory research undertaken in collaboration with the vulnerable or marginalised is increasingly seen as being valuable in this context (IDRC 19932021-2021, Powell-Lawton 2001, Ringaert 2001, Kotamraju and van der Geest 2012, Patrick and Hollenbeck 2021. While this approach reflects a broad church and is sometimes in danger of over stating the extent to which such research is genuinely collaborative and empowering, it does signal a fundamental shift in attitude away from doing research on people, towards doing it with them (Cornwall and Jewkes 1995). The focus on co-production of questions, iterative design and inclusion serve to address many of the inequities of knowledge production highlighted in contemporary debate about epistemic injustice. Ignoring the importance of such methodological considerations runs the risk of the digitalisation process remaining insensitive to the ways in which online interactions inhibit participation in the justice process or can be more responsive to the needs of the digitally disadvantaged. Failing to listen to the perspective of this group also increases the possibility of the reform programme becoming implicated in the production, reproduction and exacerbation of existing structural inequalities.

Conclusion
Returning to the question of who needs to be excluded in the interests of inclusion, we have shown that not only are there a range of factors that make people unsuitable for online hearings, but that the concept of disadvantage can rarely be understood by reference to just one characteristic. The guidance issued during the pandemic has failed to capture the breadth of users who should attend face to face. Judicial guidance issued during the pandemic largely failed to address the issue of how digital disadvantage could be identified, focussing instead on the characteristics of cases and procedural complexity. More particularly, very little of this guidance reflects a nuanced understanding of the way in which vulnerabilities intersect with each other. By contrast, the research we conducted with other professional and lay users of the justice system placed these concerns front and centre; reflecting concerns which were often corroborated by the emerging literature on the topic. Of particular importance here is the suggestion that lack of awareness of the many ways in which people can be digitally disadvantaged is in danger of creating and exacerbating existing issues around effective participation.
There is no simple answer to the questions posed by this article but as digital by default or push to digital become the norm in the provision of government services, the issues of how digital disadvantage is defined and intersects with other vulnerabilities, needs to be made a key consideration in the design of software programmes, public engagement or information initiatives, and judicial and court staff training. For some, recognising and training citizens in digital competencies has been identified as one of the key challenges for the knowledge society (Morte-Nadal and Esteban-Navarro 2022) but we argue that the responsibilities of democratic e-government or e-justice need to go far beyond ensuring that the populace are trained and equipped to participate. The research reported here makes clear that more sophisticated guidance needs to be provided to judges in the aftermath of the pandemic and that there is evidence of demand for this. The HMCTS evaluation of remote hearings during the pandemic found in a survey of over 8,000 court users that only around six in ten judicial respondents (62%) and HMCTS staff (57%) recalled receiving training and guidance on remote hearings during a period in which more hearings went online than ever before (Clark 2021). There is clearly a danger that current debate places too much emphasis on addressing problems associated with a discrete trinity of hardware, skills, and connectivity, and designing triage systems that only focus on this trinity. The issues discussed in this article are of mounting importance post-pandemic as HMCTS continues to roll out its digitalisation programme in the absence of legal aid for most claims, if access to justice issues are not to burgeon and create a two tier justice system Notes 1. This builds on the lessons learned from the Cloud Video Platform (CVP). 2. For the purposes of this article we are defining hybrid as a hearing in which some of the participants, usually the judge and court staff, are in one room together but interacting with other participants via a video link. 3. These included BT conference call, Skype for Business, court video link, BT MeetMe, Zoom, (Courts and Tribunals Judiciary 2020d). For problems see (Tomlinson et al. 2020, Ryan et al. 2020). 4. The project team comprises the authors, Dr Emma Rowden from Oxford Brookes University and graphic designer Dr Lucy Klippan. HMCTS were project partners and the research benefited from the expert advice of an Advisory panel chaired by Sir Ernest Ryder. 5. See https://www.law.ox.ac.uk/supporting-online-justice 6. Much of the guidance on which this section draws can be found at: https://www.judiciary. uk/coronavirus-covid-19-advice-and-guidance/ 7. On this point see also https://www.gov.uk/government/news/priority-courts-to-make-surejustice-is-served 8. It seems likely that the issues to be taken into account also changed during the pandemic when a decision not to hold a hearing online, however inadequate the technology available, could lead to considerable delays in a case being decided or urgent safeguarding decisions not being made. 9. Other guidance suggests that a telephone or video conference involving only litigants in person is perfectly permissible (Judiciary of England and Wales 2020aa). 10. See also the extensive guidance in the Equal Treatment Bench Book on this point. 11. For further details see https://www.law.ox.ac.uk/supporting-online-justice 12. Social distancing and lockdown created extensive challenges during this phase of the fieldwork. We had originally anticipated conducting focus groups in hotels and community centres across the UK but travel restrictions meant that fieldwork was limited to London and Oxford. When the government restrictions extended into July 2020, which was the last month in which focus groups could feasibly be undertaken, funding set aside for room hire was used to purchase two gazebos which we used to conduct focus groups in a car park in Oxford and a community garden in inner city London. 13. See also Clarke (2020) on this point.