Can White allyship contribute to tackling ethnic inequalities in health? Reflections on the experiences of diverse young adults in England

ABSTRACT Ethnic diversity and racism have not featured strongly in English research, policy or practice centred on understanding and addressing health inequalities. However, the COVID-19 pandemic and the Black Lives Matter movement have shone fresh light on deep-rooted ethnic inequalities and mobilised large segments of the population into anti-racist demonstration. These recent developments suggest that, despite strong counterforces within national government and the mainstream media, there could be a shift towards greater public awareness of racism and potentially a willingness to take individual and collective action. This paper addresses these developments, and specifically engages with the contested notion of ‘allyship’. We bring together the experiences of 25 young adults living across England and prior literature to raise questions about whether and how racialized White individuals can play a role in dismantling systemic racism and reducing ethnic inequalities in health. Our analysis reveals a variety of complexities and obstacles to effective and widespread allyship. Findings suggest the need to nurture contingent, responsive and reflexive forms of allyship that can attend to the harms inflicted upon racially minoritized people as well as push for systemic transformation. White allyship will need to take a variety of forms, but it must be underpinned by an understanding of racism as institutional and systemic and a commitment to tackling interlocking systems of oppression through solidarity. The issues addressed are relevant to those occupying public health research, policy and practice roles, as well as members of the public, in England and other multi-racial settings.


Introduction
Despite long-standing policy commitments to address health inequalities at national and local levels in England, there has been disappointing progress. Dominant policy orientations focus primarily on individual behaviour-change interventions rather than wider socio-political causes. There is growing interest in the potential for public opinion to shape more effective action on health inequalities (see for example Popay, 2003;Putland et al., 2011) and prominent calls to increase awareness of the structural determinants of health among the general population (Kane et al., 2022).
However, research on public attitudes towards, and understandings of, health inequalities has not so far given serious attention to ethnicity and racism. This omission mirrors the wider tendency within UK health inequalities research and policy to overlook ethnic inequality; a noticeable contrast to the US (Ingleby, 2012;Salway et al., 2010). Recent events have, however, brought ethnic inequalities and systemic racism more clearly into view within the healthcare and public health arenas, as well as wider society.
The disproportionate impacts of the COVID-19 pandemic on ethnic minority people in England has shone fresh light on deep-rooted social and health inequalities (Ganguli-Mitra et al., 2022). Meanwhile, the Black Lives Matter (BLM) movement succeeded in mobilising large segments of the British population into anti-racist demonstration. These recent developments suggest that, despite prominent and resilient counterforces within national government and the mainstream media denying systemic racism (Commission on Race and Ethnic Disparities, 2021), there could be a shift towards greater public awareness of racial injustice. Important questions are therefore raised about whether and how White people can play a role in furthering the anti-racist agenda and addressing ethnic inequality; a topic of considerable debate within wider scholarship and activism that deserves greater attention within the public health field. Given the plurality of definitions of racism, it is unsurprising that Paradies (2016) argues for the plurality of definitions for anti-racism. Core anti-racism principles however focus on taking action to confront and ameliorate racism and privilege. In public health, an anti-racism approach has been defined as one that seeks to heal and empower racially minoritized people. Anti-racist analyses in public health often focus on the recognition that the world is controlled by systems which benefit or disadvantage the health of some people unfairly based on race (Came & Griffith, 2018). In the field of public health, there are few frameworks which seek to support anti-racism public health praxis (Came & Griffith, 2018) and interest in this area is growing.
Within this paper we extend the growing body of work on public understandings of health inequalities, by focusing on ethnic health inequality and racism, and by specifically engaging with the notion of 'allyship'. White allyship has been variously conceptualised and evaluated as a strategy for redressing racial inequality and oppression (Came & Griffith, 2018;Williams & Sharif, 2021). Some promote a model of allyship in which White people acknowledge their race-based privilege and channel their inherent power into actions that challenge racism (Case, 2012). Others, however, view this approach as undesirable, pointing to the ways in which this framing is disempowering and undermining of racially minoritized people's own struggle (Dabiri, 2021). The latter argue that finding common causes and working together in ways that encourage solidarity is advantageous for all exploited groups and more likely to challenge racial inequalities. In this model, rather than foregrounding White privilege, the identification of shared goals and development of coalitions benefit all parties involved. One example of this latter model lies in the Chicago Rainbow Coalition of Revolutionary Solidarity. This 1970s coalition organised Black and other minoritized people, together with disenfranchised Whites, to collectively resist racism, police brutality and the inequalities perpetuated by White supremacist capitalism (Middlebrook, 2019). The contemporary UK Northern Police Monitoring Project (http://npolicemonitor.co.uk/) has many similarities. This model of White allyship is also consistent with intersectional approaches that highlight how there are common sources of oppression (e.g. police brutality) but the immediate causes, and experiences, of the oppression differs. However, rather than viewing these modes of White allyship as necessarily oppositional, contributions that take a more nuanced approach, conceptualising White allyship as contingent, responsive and reflexive are perhaps more useful to those wishing to promote anti-racist action. Ensuring that White allyship actions are rooted in an understanding of racism as institutional, systemic, and historically rooted, and motivated by a commitment to transforming oppressive structures and processes, is key (Nixon, 2019). There are parallels between the growing recognition of the range of forms of health activism more generally (Campbell & Cornish, 2021) and the exploration of White allyship that we present here. The issues addressed are relevant to those occupying public health research, policy and practice roles, as well as members of the public, within and beyond England.

Methods
Here we bring the experiences and understandings of young adults into dialogue with previous empirical and theoretical work to raise questions about the possibilities of allyship as a route to reducing ethnic health inequalities, the forms it might take, and the obstacles and enablers that currently shape the possibilities.
We undertook a series of online focus group discussions with young adults aged between 18 and 30 living in London and the South East, South Yorkshire, the Midlands and the North East of England. We aimed to explore young adults' views of health inequalities and their priorities for reducing them.
The original study design included a series face-to-face group discussion (FGDs). Due to the COVID-19 pandemic, we adapted the study design and used a combination of synchronous and asynchronous online FGDs, using a range of platforms (Zoom, Blackboard Collaborate and WhatsApp) that enabled participants to connect using their laptop, PC or smartphone (Fox et al., 2007;Tuttas, 2015). Advantages to using online methods include increased accessibility for participants, no travel time and some reports of online focus groups allowing participants more time to think about their answers when responding (Zwaanswijk & van Dulmen, 2014). In adapting our study design from face-to-face to online, we considered the ethical and practical implications (Woodrow et al., 2022). This included ice-breaker activities to build rapport at the beginning of group discussions, asking participants to ensure privacy and that each online FGD had three facilitators (one to monitor the verbal exchanges, one to monitor the chat box and one to monitor any technical issues or participant distress). Additional team members acted as observers and notes takers. After each of the first FGDs, one of the facilitators from each group (SR, SEM, VMG) contacted participants individually for a debriefing session to gather feedback.
Four groups were convened, two comprising participants resident in London and the South East (hereafter termed South East) and one each comprising participants resident in the North East and in South Yorkshire and the Midlands. There were 25 participants in total (see Supplementary Material, Table S1). Recruitment to the groups was intended to include a diversity of social identities while creating comfortable spaces within which participants would feel able to speak freely and listen attentively. This approach allowed us to hear the voices of Black and ethnic minority young adults, as well as those of young adults experiencing other subaltern social locations that carry with them deprivation and/or discrimination. To explore potential differences in understandings of health inequalities across groups, we aimed for sampling variation in terms of age, gender, ethnicity, sexual orientation, education, and socioeconomic status.
Participants were recruited via community organisations that were supporting young adults in various ways (e.g. support groups and activities for trans young people; training, education, and skill development to help socio-economically disadvantaged young adults secure employment). In the South East and the North East, organisational representatives sent written information to potential participants, who had around 1-2 weeks to consider participating. In South Yorkshire and the Midlands, we contacted community organisations with written information which was then shared with potential participants. FGD moderators (SR, SEM, VMG) arranged a preliminary call with potential participants to provide further information on the study, go through ethics, and answer questions. Interested participants sent written consent by email.
For the two South East groups and the South Yorkshire and the Midlands group, a twopart discussion guide was developed to elicit participants' views of health inequalities and their priorities for reducing them. Each of these three groups met twice between July 2020 and March 2021.The discussions were conducted via synchronous verbal and chat functionbased communication using the online platforms Zoom and BlackBoard Collaborate with two facilitators and lasting approximately 1 hour and 30 minutes. The North East group discussion involved a single facilitator who employed a combination of synchronous and asynchronous WhatsApp exchanges over two days, as a negotiated flexible approach to enable engagement of young men who were working shifts. The North East group followed the discussion guide more flexibly due to the nature of the platform. Discussions in all groups were openended, inviting participants to share experiences and understandings of health inequalities, with each other and with group facilitators.
Recordings were transcribed verbatim, uploaded into NVivo 12 and analysed using thematic analysis (Braun & Clarke, 2006). A coding framework was developed iteratively using a mixture of inductive and theoretically-informed deductive codes. The exploration of allyship was prompted inductively. Coded material was subsequently organised into eight thematic areas (understandings of inequalities; intersectional processes; allyship dimensions; allyship across identities; challenges to allyship across identities; impact of allyship actions; social justice movements; importance of allyship) which were then reviewed and refined to build claims in relation to four sections presented below. We use verbatim verbal and text (from the chat function) quotations below to illustrate the claims being made, identifying the participants by a participant number, their preferred pronouns, their stated ethnicity, and the group they were part of. Ethical approval was obtained from The University of Sheffield (reference number 381) and the London School of Hygiene and Tropical Medicine (17783 -1).

Prospects for allyship as advocacy (mobilising White privilege to disrupt racist systems)
Our participant accounts resonated with the ambivalence outlined above towards a championing form of allyship rooted in White privilege. Their accounts suggested both a need for, and some significant obstacles and uncertainties associated with, this form of allyship.
Some participants highlighted Black and ethnic minority people feeling ground down by racism, exhausted by repeated aggressions, and fearful of the consequences of pushing back at racist systems and processes. Participants voiced their desire for White people in more powerful positions to acknowledge their privilege and take a stand within institutions and broader society.
I feel like it's a White problem to address but because they've got the privilege, they don't need to address it because [racism] is not a problem for them. (P2, he/him, Asian British (Chinese), South East group 1) This respondent sensed a deliberate resistance to addressing racism from White people. This resistance, the lack of addressing racism, was perceived as due to racism not affecting White people. There was also a strong sense of disappointment in the persistence of racism and in White people's complicity with, or active perpetuation of, racist behaviours and structures and their negative impacts on health.
[Quotas for hiring ethnic minority people are] . . . a step in the right direction. But more importantly there shouldn't need to be quotas. Why can't you just be not racist? Do you see what I'm saying? Like why aren't the people that are in HR better trained to look more broadly at the person (P15, she/her, Black British, South Yorkshire and the Midlands).
. . . there are clear inequalities within the health space dependent on various factors such as ethnicity, race, class, status, language. It's sad and disappointing and makes you feel frustrated and angry . . . .I have for example seen how people are treated in terms of like how they're spoken to. For example, so if somebody speaks English as a second language, the frustration sometimes from health professionals [. . .] I think people are judged. . . . someone who's just arrived or haven't got their language ability to communicate that they need an earlier appointment, they're less confident, you know, to articulate that, then their access to treatment is going to vary, it's going to differ greatly and that's going to have an impact on their health in the long term (P5, she/her, Asian British, South East group 1). Some respondents identified and linked racial health inequalities to long-term effects on the health and well-being of Black and ethnic minority people. Respondents provided clear examples of actions that could be taken by allies, situations where allies could intervene and ways in which systems could be made anti-racist e.g. better HR training (P15). Angela Davis (2016) has written extensively about the balance that racially minoritized people have to find between taking action to disrupt systems of oppression and self-preservation. The exhaustion experienced by ethnic minority people, the higher risks associated with challenging systems, and the potential lack of contextual knowledge and confidence, are reasons why White people may need to adopt an advocacy ally role. [. . .] because of the way that we've [people from ethnic minority backgrounds] experienced racism I feel like we're a bit, we've lost confidence, you know, like so many people, like for me I'm kind of like I'm just here to do my job, I don't want to, you know, I don't want to be a leader [. . .] if anything goes wrong who's going to stand up for you and like back you up, you know? (P18, she/her, Black African, South Yorkshire and the Midlands) P18's question "who's going to stand up for you?" suggests the need for allies to speak up and support colleagues from ethnic minority backgrounds in the workplace. Nevertheless, our respondents' experiences, and the wider literature, highlight various obstacles to White people acting as powerful advocates. Case et al. report would-be allies perceiving their role to 'make other White people listen' (2012, p. 85) but often remaining silent in the face of injustice for fear of jeopardising their own social position or a feeling of powerlessness. DiAngelo's influential US work (2018) further illustrates the unwillingness of Whites to upset the institutions they are part of, while Salway et al. (2016) found a fear of 'getting it wrong' underpinned White silence and inaction in UK health organisations. Our respondents also recognised the discomfort and fear White people have around discussing race and taking action on racism. One participant, who was the only Black person in her workplace, highlighted the complete silence by her colleagues around George Floyd's murder and the BLM movement.
I then spoke to my manager about it [not discussing BLM] and, you know, I said "I was quite disappointed that no-one said anything" and she said "do you know what, I did think about it and I did think I wanted to say something, I just didn't know how to say it and I didn't know what to say. I didn't want to offend you" . . . I think it is [White] people don't want to have those uncomfortable conversations, even though I don't understand because it's way more uncomfortable for me to have that conversation than it should be for you (P21, she/her, Black African, South Yorkshire and the Midlands).
Such perceptions of risk may, however, be well-founded and speak to the importance of recognising the ways in which White privilege and power intersects with other aspects of social location, thereby producing 'shades of Whiteness' (Hartigan, 1999) and differentiated potential for allyship-asadvocacy across space and time. Paul (2017) notes that being an ally who advocates for systemic change -what he terms being an accomplice -will take its toll. This work implies risking your own status, confrontation and loss of relationships, and jeopardising your own privilege. These consequences are likely to be differentiated along the lines of class and gender. One respondent shared an experience of colleagues within her company being actively silenced by management, a clear indication that speaking out on issues of racial justice was disruptive and risky.
I started working for this company only a few weeks ago and I had to sign a clause actually that basically outlines that we're not allowed to share opinions on certain topics [. . .] she said it's from after the BLM protests and stuff that they've made every member of staff sign it because there was a lot of arguments going on. People falling out left, right and centre. People didn't want to have arguments. They've made everyone sign a clause to basically say they're not allowed to discuss it. (P20, she/her, Mixed (Indian and European), South Yorkshire and the Midlands) In addition, some of our respondents expressed concern that White people in positions of power do not fully understand racism. There is therefore a danger in them adopting an advocacy role, seeking to 'resist for' racially minoritized people, in ways that are unhelpful. There is a need to understand racial inequality and racism before allyship-as-advocacy can be usefully enacted. This requires critical reflection and the acceptance of challenge around behaviours, policies and spheres of influence.
[. . .] my chief exec is a middle-aged White woman that . . . is middle class. We don't have anything in common, but you make decisions for us . . . . You will never go through the same experiences, you'll never know what it's like to be a Black woman, what it's [like], you'll never understand (P21, she/her, Black African, South Yorkshire and the Midlands).

Prospects for allyship as solidary coalition (resisting with groups from ethnic minority backgrounds)
Critics of advocacy allyship rooted in White privilege argue that it can detract from ethnic minority people's agency and assets, rather than empowering these communities to flourish (Dabiri, 2021). Meanwhile, intersectional analyses alert us to the differentiation of White privilege and power along other axes of oppression, including gender, socioeconomic status and age. Both sets of arguments suggest the value of building coalitions that pursue an anti-racism agenda alongside broader based social justice work. Our discussions with young adults suggested significant current barriers to such coalition allyship, but also some opportunities to nurture solidarity across diverse ethnic social locations.
This form of allyship implies Whites taking a supporting role in the fight for racial justice and may involve taking a step back from work within racially minoritized spaces but stepping up to lead antiracist work within White communities (Leonard & Misumi, 2016). Our ethnic minority participants recognised the importance of spaces for them to come together and speak honestly about race and racism.
[. . .] we just complain to each other, other Brown and Black faces that look like us, [. . .] it's like we understand each other on a very core basic level but those conversations are being had between us and not with those that call themselves our allies or our White British counterparts because [. . .] when they have those conversations with us it becomes a "oh something similar happened to me" and then it becomes like a competition of, you know, it's not just you, you're not special, or you're taking it too seriously or you're being sensitive or, so your feelings are basically not valid [. . .] Your thoughts and process is just [a] piece of shit on the wall (P16, she/her, Pakistani, South Yorkshire and the Midlands).
In the context of this respondent's words, the role of an ally would be to support the creation of safe spaces for people-of-colour, in some cases without White people being present.
Furthermore, this participant's contribution suggests a gulf between what she considers to be supportive interracial discussions and her real-world experiences. This resonates strongly with Reid's (2021) call for aspiring White allies to avoid the common tendency to invalidate Black and ethnic minority experiences, and questions White people's readiness to operate in supportive solidary formations.
Understandings of health and social inequality are shaped by our own experiences and position in society. Our tendency to interact with those who are like ourselves can result in limited understanding of the challenges faced by those occupying dissimilar social locations (Bottero, 2005;Irwin, 2018). This was evident within our groups, with participant discussions tending to focus on their own experiences of inequality. Further than flagging White individuals' inadequate grasp of the nature and impact of racism, the broader South Yorkshire and the Midlands group discussion suggested a lack of consideration among the participants of the possibility of shared experiences and struggles with White people. The narrative was framed in terms of difference and disconnection from White people. Meanwhile, the young White men from the North East were entirely silent on experiences of ethnic minorities (despite being prompted to think about how Covid may have impacted some people worse than others). And, while none of our White respondents expressed racist ideas, earlier work shows how right wing political elites have fuelled anti-migration sentiment among White working class people by presenting migrants as the source of their socioeconomic disadvantage (Gillborn & Kirton, 2000). This sentiment could potentially influence feelings towards support for racially minoritized groups. It was, however, encouraging to find that some White participants in the South East groups, which had a mix of White and ethnic minority participants, did discuss racial inequalities, particularly in relation to access to health care. These findings suggest the need to build greater appreciation of common sources of oppression across different marginalised social locations -the bedrock of solidary allyship -as well as increased understanding of particular experiences.
Importantly, the accounts shared did suggest commonalities in the young people's lives across ethnicity. Shared experiences included: feeling 'othered'; stigmatising and discriminatory behaviour from professionals (e.g. GPs and the police) and members of the public; being constrained by structural factors beyond their control; and challenges of poor health. For example I'm friends with some of like the White girls here and which is why I have no problem with that, but if we walk into a shop one after the other you can see they'll be polite, for example in [Supermarket] my friend will be in front of me in the queue, they'll speak to her nicely, speak to her politely, but when it's me you're treating me differently, there's no way of sugar-coating it, I can tell that you're treating me differently to the way that you treated her, which is not fair (P15, she/her, Black British, South Yorkshire and the Midlands.).
Participants from all backgrounds also shared common experiences of poor mental health and frequently located the origins in aspects of their social or economic context beyond their control. This was one area where some participants in all three groups explicitly acknowledged the struggles of those occupying other social locations, though they did not go so far as to recognise common structural causes I guess the pandemic made me think about how much some people have to think about health when they go about their daily lives. And the more you have to think about health the more it kind of strains on your, I guess, general mental wellbeing if you have to go into every shop being wary (P9, she/her, White American, South East group 2). Campbell and Cornish (2021) usefully draw attention to the transformative potential of 'quiet' forms of health activism that are more subtle than the stereotypical noisy and oppositional campaigns. Our racially minoritized participants identified acts that can be considered 'quiet' White allyship. For instance, some talked about positive experiences of everyday kindness, care and support from White colleagues and friends. Others mentioned White people taking actions within their immediate sphere of influence, such as a GP practice where steps were taken to improve translated materials and interpreting services, and a case where a colleague advocated on behalf of an ethnic minority member of staff with asthma who was being denied the opportunity to work from home during the Covid pandemic. These small-scale acts towards redistribution were reported positively by participants.

Loud and quiet allyship
However, quieter interpersonal acts can mask a lack of systemic challenge and enable the status quo to persist. The excerpt below, from a participant who works within the NHS, illustrates how the White person's attempt at supporting her career advancement was overshadowed by her enduring negative experiences within a racist organisation. She spoke of her fear of not receiving support if errors were made, reflecting the wider reality that people from ethnic minority backgrounds are more likely to enter formal disciplinary processes than their White NHS counterparts (NHS Workforce Race Equality Standard, 2021).
[. . .] I remember one of my consultants saying that "you are really good at organising things, you should really apply for like, you know, to become a senior physician associate" and I was like, I am not going to be doing that, I'm not going to be leading a group of people from, you know, I don't mind leading people from Black and ethnic minorities but, you know [. . .].like a group of people with like White and, you know, like I just don't feel confident in myself to do that because that's the way that I was just raised, that's the way, that just do your job and keep quiet, you know (P18, she/her, Black African, South Yorkshire and the Midlands).
In the case above, the White manager might believe they had acted as a good ally and that the obstacle was our Black respondent's lack of ambition. However, the action seems to be framed within a mentality of 'helping' the Black colleague (Nixon, 2019), rather than actively disrupting systemic barriers, as increasingly demanded (Dabiri, 2021;Kutlaca et al., 2020;Paul, 2017).
Many of our respondents referred to the BLM protests, and some were also quick to question whether these forms of visible, 'loud' allyship would lead to systemic change, or rather merely serve the interests of White people seeking to perform their allyship.
It [BLM protests] should be like an everyday thing. People should still be talking about injustice but it's over now and we're back to the football now. It's this, it's that and just a real lack of care, like a real desire to actually sort out this problem. (P2, he/him, Asian British (Chinese), South East group 1) While the potential for social media to amplify and inform was recognised, participants also expressed particular concern that online activity, while superficially loud and oppositional, does not necessarily translate into tangible actions offline.
On social media [BLM], it's become like it was a trend. But, I like to think that people are doing things off social media and it has inspired people to read up, read books and not necessarily have to post about it. I'm definitely one that does post about it a lot, so I don't see why people wouldn't post about it simply because social media is really, you know, a powerful platform to use for these things (P1, he/him, Black African, South East group 1) Other examples of visible apparent allyship were also called out by respondents as empty performative acts.
[During Covid 19] Our Trust put out some meetings on like doing risk assessment for us, people from Black and ethnic minority. However, they made those meetings during the times that we were working, that most Black and ethnic minority staff would be working. I remember asking them "can I have the videos from the meeting?" and up to now I still have not got them, you know. The meeting was held at 9 o'clock on a Tuesday morning, where are most of the Black and ethnic staff? They are in work, or they are finishing a night shift. So, I think from that point of view it was not really, it's like they were just putting it in the broadcast to say "oh we're working on this" but they were not really acting on it (P18, she/her, Black African, South Yorkshire and the Midlands).
In the scenario above, providing the recording would have meant better involvement of the Black and ethnic minority staff in the COVID-19 risk assessments. It would have been an act of allyship to follow up and ensure that the recording was provided. This scenario highlights the importance of taking actions, even small ones, towards inclusion.
Thus, pleas for White allies to move from passive sympathy to political action resonated with many of our racially minoritized respondents, but they were alert to the importance of seeing beyond superficial form to evaluate the transformative impact of actions, whether 'loud' or 'quiet'.

Reflexive and contingent allyship
The discussion above highlights varied forms that allyship can take, the diverse contexts in which action is needed, and the tensions and complexities that can undermine impact. Our respondent experiences and understandings and the wider literature remind us that: well-intentioned White people get things wrong; opportunities for allyship are overlooked or ignored; significant groundwork is needed to build coalitions between variously oppressed groups; the extent of an individual's White power and privilege is inflected by other social identities and context dependent; and allyship work can be alienating and exhausting. These findings suggest the need to nurture a reflexive and contingent notion of allyship, that attends to core anti-racist principles, but can accommodate a variety of modes. Such an approach may offer promise in disrupting racism and its impacts on health across varied contexts and scales.
For would-be White allies, opportunities for action are likely to be various, and determining when and how to act effectively across contexts is not easy. Recent proposed principles of White allyship foreground reflexivity and humble reflection on one's own complicity in the reproduction of racism (Came & Griffith, 2018;DiAngelo, 2018;Nixon, 2019;Reid, 2021). Our Black respondents reiterated this need, and some highlighted how far many White people must go before they are likely to contribute to effective allyship. I do believe a lot of people are unconsciously prejudice and they don't realise that they're being it because, they've probably never been pulled up on it, or if they have, then they've seen the person to be, I quote, an angry Black woman or an aggressive Black man (P15, she/her, Black British, South Yorkshire and the Midlands).
I'm pretty sure people have all heard about that [George Floyd] story and it just brought to light something that's obviously been happening for as long as we've all existed and isn't spoken about, you know, learnt about, isn't sort of fought for enough at all. I think it's just been an awakening to some people. I think some people already are doing, you know, what's in their power and trying to fight for change. However, others I think are waking up to the systemic racism around the world in education, in health, in at home, you know, everywhere basically (P1, she/her, Black African British, South East group 1).
The need to build awareness, skills and confidence to act among a much broader base of White people also speaks to the need for a nuanced and flexible understanding of allyship. Recognising and appropriately promoting the range of allyship tools at our disposal may help to ensure attention to the well-being of those around us who are at the raw end of racial discrimination now, as well as pushing for radical change. Table 1 suggests actions that can be taken as colleagues, community members and public health professionals across the range of forms of allyship identified above, some of which come directly from the respondent narratives.

Conclusion
Before concluding, some limitations should be acknowledged. Our analysis drew on a relatively small body of empirical data. Furthermore, the groups varied in their degree of homogeneity, with the North East group being a particularly homogenous group of young White men living in disadvantaged socioeconomic circumstances. There would be value in exposing participants to more diversity, providing an opportunity to share stories and potentially identifying commonalities of experience and potential for solidarity; aspects that were limited in the group discussions here. The discussions covered a lot of ground, exploring varied aspects of intersectional health inequalities, so that we did not delve as deeply into understandings and experiences of allyship as we might have done in a more narrowly focused project. We also acknowledge that the inductive nature of the analysis meant that opportunities for further probing were not taken up. Therefore, our findings are necessarily suggestive, and we encourage more detailed investigation. Nevertheless, bringing these experiences into dialogue with wider literature has served to raise some important issues for wider reflection within the public health community, which has to-date given limited attention to this area.
Nixon describes the practice of allyship as "fraught, messy, ongoing, and laden with missteps" (2019, p. 11), a depiction that is mirrored in the findings presented above. Further, our analysis supports the assertion that would-be White allies frequently do more harm than good through empty performative gestures or invaliding the experience of Black people. Additionally, the literature also warns of the dangers of White saviourism (Nixon, 2019;Oppong, 2023). Work from the US also highlights the danger of focussing on diversity and alliances of Black and White people masking ongoing experiences of racism and marginalisation that racially minoritised people face (Curry & Curry, 2018). However, there are potentially opportunities for all people to identify common oppressive factors and opportunities for action. Whether we refer to acting in allyship, being an accomplice, engaging in activism or being a member of a coalition, action must be underpinned by an understanding of racism as institutional, systemic, and historically rooted (Dabiri, 2021;Reid, 2021). Racism must also be recognised as interlocking with other systems of exploitation and oppression, being "intimately tied to capitalism, as well as heteropatriarchy and ableism" so that "racism cannot be tackled without reckoning simultaneously with these other systems" (Connelly & Joseph-Salisbury, 2021, p. 206).
Further, our analysis confirms the varied opportunities for action and need for flexible and reflexive responses. White people disproportionately occupy positions of power within State and wider institutions that perpetuate systemic racism, and opportunities for action are frequently overlooked. These people -including public health researchers and practitionerscan and must do much more, through both quiet and loud actions, to redistribute power and transform racist structures and processes. At the same time, there is clearly a need to engage a much wider range of White people -many of whom do not feel powerful or privileged in their everyday lives -in anti-racism work. We need to find better ways to build coalitions that are organised around common exploitation, such as workplace discrimination and poverty, whilst attending to particular experiences of racial oppression. Evidence from previous US civil rights movements shows that the practice of building coalitions was problematic with White people holding negative and stereotypical views of Black Americans (Bonilla-Silva, 2001). However, it is worth considering that the process of building these coalitions across shared issues may lead to increased opportunities for listening to and sharing space with people from different ethnic backgrounds. This may in turn lead to fewer negative and stereotypical views to be held by White people as well as encouraging progress on shared issues. Hope can also be found in examples of alliances of White and Black people from across the world. For example, in Australia where poor Whites, Blacks and migrants forged a strong and disciplined enough team to deliver the first ever defeat of the local government who wanted to close a significant number of schools (Foley, 2000) or the recent Fees Must Fall movement in South Africa where students of all backgrounds mobilised to protest institutional racism (Albertus, 2019). The research evidence base on how to achieve these coalitions remains limited, but is growing (Kutlaca et al., 2020;Selvanathan et al., 2020). Experiential knowledge from those involved in grass-roots community and workplace organising should also be drawn on. Future research that takes a longitudinal approach will be helpful in tracing the processes and impacts of organising attempts on ethnic and other health inequalities.
The current moment seems to present both opportunity and challenge to strengthen White allyship. Calls for anti-racist work in the public health areas have also been growing in recent years ; (Came & Griffith, 2018;Essien & Ufomata, 2021). We are faced simultaneously with increased awareness of racial injustice but a growing preoccupation for many White people with their own day-to-day survival. The dominance of right-wing political ideas, social media and mainstream media in many countries is also fuelling racism and division. Therefore, public health professionals must take a lead in building the solidarity needed to tackle ethnic health inequalities alongside broader social injustice. In public health scholarship and practice, this work may include the elimination of racist practice and policy in the workplace, diversification of the healthcare and education workforce (at all levels) and the delivery of anti-racist public health training. These are actions that can take place alongside the evaluation of progress towards improved health of ethnic minority people and social transformation.