Some three-quarters of workers say they’ve felt excluded at work, according to new global EY research. One way for organizations to make themselves more inclusive is to intentionally cultivate a culture of allyship, argues organizational psychologist Meg Warren, an associate professor of management at Western Washington University who studies workplace allyship and inclusion. But many workers and workplaces, she argues, aren’t going about it the right way.

We spoke with Warren about common myths around what it means to be an ally and how to embed allyship into organizational culture. Here is a transcript of our conversation, lightly edited for length and clarity:

Can you define what it means to be a workplace ally?

The term ‘allyship’ has been used in a variety of different ways in the popular literature, and in the research it is a little narrower. Some of the early research defines allyship as actions by people from dominant social groups who are engaging in supportive or advocacy behaviors towards marginalized groups. It's consciously using your privilege and power as part of a dominant group in support of a marginalized group.

More recent work has expanded this definition to also look at people from one marginalized group who can stand up in support of another marginalized group—for instance, if you're looking at BIPOC folks, you may have folks who are Asian who are looking to do pro-Black allyship. So you're still an out-group member. There are specific advantages to this: When you are not the part of the group that is currently being targeted in a particular situation, when you speak up and say something, it is seen as more credible, because it’s not coming from a place of self-interest. So that voice has an added layer of educational value and added layer of visibility, and people are more likely to hear it better than when a person from their own in-group is speaking about it.

My research focuses on the why, the what, the how, the so-what, all these different aspects of allyship, with an eye to both motivating allies and also looking at interventions—what works and doesn't work, and what might be some myths around allyship work, that are not supported by the research. Those myths can be busted.

What are some of those myths?

Of the most common ways of thinking about how to be an ally seems to be that if you see discrimination occur, the first step is to call it out. But in study after study that we have done—experiments, qualitative research, quantitative research, survey research—across the board, what we keep finding over and over again is that that is the riskiest way to do allyship. It's not necessarily the most helpful way to do allyship from the victim's perspective. If an ally speaks up and confronts the perpetrator’s bias to support the victim, the victim is more likely to get worse backlash from the perpetrator and other observers. Sometimes things happen to the ally too, but the victim may be more injured in the process. So from a victim's perspective, that sort of allyship is not necessarily the best way to be an ally.

And potential allies are less likely to step up as an ally if confronting bias in this way is the only thing that is understood as allyship. It is much more risky and you're more likely to burn bridges with other employees and other folks in the organization. You're more likely to be seen as creating disharmony.

A lot of folks will probably still feel like they must do it, because that's helpful for the organization and for equity. But if it's not helping the victim as much, then they may want to reconsider other ways of being allies. What my research has been doing is looking at, what are other ways of behaving as an ally and engaging allyship interventions? This is not to say ‘Don't call out bias.’ It's just kind of a sharp tool. Be careful when you use it. Use it when it's really important and consider from a variety of different strategies what is the best fit.

What are some other tools in the toolbox?

Of the more proactive allyship strategies that we've seen, one is what we call ‘impression promotion.’ If you see on the horizon that there might be backlash, or there might be a situation that would occur that's going to be biased for a colleague—one example is when you know that a colleague is pregnant and you may probably see people starting to question their competency—you could proactively start to promote the impression of this person within your team. Keep reminding folks about the work that they're doing, the kind of achievements that they have. They've surmounted a lot of huge challenges in the past. Start building up that space, so people already know not to start thinking of this person as someone who will be incompetent in the next few months.

Another really useful strategy that we've been seeing is what we call ‘highlighting strengths,’ pivoting the conversation if someone is being discriminated against or at the receiving end of microaggressions. Let's say their competency’s being questioned. Rather than calling out the bias and saying ‘This is not a fair accusation,’ pivoting is pointing to the strengths that disprove that. Highlighting strengths essentially pivots the conversation to not just saying ‘Where you're coming from is incorrect’ and calling that out, but saying, ‘Here are all the amazing reasons why this person is really strong, so obviously where you're coming from is not accurate.’ We see from our research  that this has long-term effectiveness for changing the situation and starting to move people towards equity. It has less backlash, so it's better for the victim in that way. And people are more likely to see the ally in a more positive light, so from the ally’s perspective, there's less risk involved.

When there is discrimination, it is possible that sometimes a victim may say, ‘I don't really want to get into this, I don't want to call this out, I don't want someone else calling it out. This is one that I want to let go. I have other battles I would like help for.’ But at the same time, from a leader's perspective, a leader may decide that allowing this to pass right now is not good for the culture of the organization. It allows this to be seen as normalized and acceptable. So sometimes what a victim wants may be at odds with what is best for the organization in terms of moving ahead towards more equity. And the leader might decide that rather than making a point about this with this particular victim in mind, they have a more generalized conversation about how not to engage in these sorts of discriminatory behaviors, without pulling attention to this one person. Are there other ways to do more inclusion work and allyship work without it being a focus on this person?

So the idea there would be to address the problem behavior, but to do it outside of the moment where it's happening.

If it doesn't look like it's appropriate for the moment, it's not a bad idea to wait and have a much more multi-pronged strategy to work on this in the long term. And of course don’t use that as a way to never do anything, or just push it away. But you may want to be strategic on how you approach this.

With an ongoing backlash to and deprioritization of diversity, equity, and inclusion programs in the workplace, what role do allies have to play in ensuring equitable outcomes?

My sense is that a lot of this attention to DEI work, social-justice work, goes in up and down in waves based on what's in the broader social and political context. If we can think about allyship as rooted more in what's very basic to us as humans working in organizations, ‘This is the kind of world we want to live in, this is what we want to do, this is who I want to be, how I connect allyship to my moral self’—I think that it becomes much more resilient to these waves. If we can connect allyship to who we want to be as people, what kind of organizations we want to see in general, then I think we can continue working on developing allies through programming, through training, where it's resilient to these waves.

How should workplaces be thinking about teaching allyship?

A lot of allyship work training is looking at, ‘Do this, not that,’ and a lot of specific behaviors. We can argue quite a bit at length about those specific behaviors and whether we're comfortable with them or not and get lost in the intricacies. I think what is more useful is to take a step back and look at principles of what good allyship would look like. What are our core principles of where we should be as allies? What kind of psychological mechanisms within us are we trying to activate? Are we coming from a space of justice? Are we thinking about this from a space of compassion? We want to be pro-social, and that is part of our organizational values, our principles. We want to create healthy workplaces. If that is a core principle, then allyship should be tied to that core principle.

Going back to those principles as foundational to allyship programming might be useful. From there, if we can look at, what strengths do organizations have that they can bring to the table? Some organizations are naturally service-oriented. Let's take nonprofits—they're already focused on serving the community, and their compassion is already activated in specific ways. If we can turn that inward and look at how we can use that channel, that same compassion towards our own employees, does that allow us to have more rich conversations? So look at existing strengths within your organization as foundational to allyship.

Another important principle is looking at gateways. Perhaps in a particular organization, we are far away from actually doing allyship around, let's say, disability or LGBTQ issues. We don't do more than the bare minimum, but we've made a lot of progress on gender. That's a great starting point. Let's see what progress we've made around gender. What are some principles, what are the things that we can have consensus around what we have already done really well at? And then let's see if we can connect that and use those strengths now to build into these other areas that we've not done so well.

So almost like creating an allyship culture document or mission statement?

Yes, allyship principles. And then from there, specific behaviors.

What would be an example of an allyship principle?

For instance, if we can look at  the organization’s strategic mission or strategic vision of who we want to be, what kind of employees we are, use that as a baseline to then develop an allyship mission statement that closely parallels that.

A lot of your research has kind of explored the gap between intention or self-perception as an ally and actual effectiveness. Can you talk about that?

It's interesting because we see about 70% of folks say that they intend to be an ally, but we don't necessarily see action at that level. A first gut reaction to that is either, ‘They're lying, it's just socially desirable,’ but it's not necessarily socially desirable everywhere to say that you want to be an ally, either. It really depends on what kind of organization and where you're placed. So that's one possibility. The other possibility is that there are a number of different barriers that prevent you from being an ally—so even though there's intention, it's not turning into action.

Our research has shown that there are barriers at a variety of different levels that do exist. People talk about barriers that exist at the organizational level—for instance, they don't have the sorts of support systems within the organization to be allies. They may not have the psychological safety to be an ally, or to speak up without at least socially being reprimanded, if not actually having material consequences to their job. There are folks who feel like they are so overburdened with work and time pressure that if they had to, let's say, take on voluntary activities or programming on allyship, then it would come at a cost of their work, and their managers would not be okay with that. So there are a variety of things that prevent them from the organizational context to actually be doing more work and learning more about allyship so that they can be empowered to be allies.

The other is a relational piece, where there may be barriers in terms of your colleagues not understanding your interests about being an ally, and there’s tension and a cost to interpersonal relationships to being an ally. It takes a lot more cognitive energy to think about how you can translate your intention into allyship while still maintaining relationships so that you can continue to work well and be effective in your organization.

Finally, that third level is internal—personality variables, emotions, fear, all of those can get in the way of being an ally in very specific ways. One example here— and we've done this research now across a bunch of different countries, so we know that this does translate across beyond the United States and North America—is that when folks believe that confronting bias, calling out bias, is one of the main ways in which you can be an ally, folks feel, ‘If I am too agreeable a person, I can't be an ally. If I'm too introverted as a personality, I can't be an ally.’ But what we've seen is that there are other allyship strategies that actually are very suitable for introverted folks. They're very suitable for agreeable folks to be allies. But if those are not part of our repertoire, then we think we can either be an ally by calling our bias or not at all. There's no other option.

What are some of those allyship strategies particularly suited towards those who might be more reserved?

Highlighting strengths, the pivoting and highlighting, works really well for folks who are highly agreeable, because you're not necessarily burning any bridge. You're just speaking about someone's strengths. You're not worried about hurting anyone's feelings. Folks who are more introverted might prefer to start with first talking to the victims privately and checking in with them, seeing how they're doing, strategizing with them to do something more concrete and visible later on, versus starting out in a public meeting or public space. Not just one conversation, but really investing that time and energy to build that rapport, thinking in terms of how you can address this, especially if it's an ongoing issue. ‘How we can address this together? What can I do to help?’ Being there in the long term can really be very helpful. It's almost like being a sounding board, a reflection partner, a thought partner to making some real change happen. And it works really well for introverted folks. You don't have to be extroverted to do that.

In one of your studies, most men thought they were great allies for their female colleagues, but the women they worked with disagreed. Is there any sort of workplace training that can help with that self-awareness gap?

We are most aware of what's going on on the inside, but we are not aware of how we are perceived. The real self-awareness gap is about how you're being perceived. I don't want to negate any internal work that people may be doing, but until it actually is visible on the outside, it's, well, invisible. So self-awareness about the extent to which your internal work is actually making any change is useful. Having more honest conversations about all the work that is necessary for change to happen—the starting point being the internal work, but also having honest conversations about where that has to translate to visibility—is really important. If we are too quick to judge folks as not doing anything at all when they might be doing internal work, we may shut them down before we can have those more honest conversations about, ‘Great, you're doing this much, but we need more.’

Going back to when you mentioned that an effective alternative to calling out bias in the moment may be to address the problem more widely later on—wouldn’t the self-awareness gap also be a problem here? How do you make sure that these organization-wide messages or interventions reach the right people, or that people are open to hearing them?

Let me take a quick step back here. I think there are times where it's very valuable to confront bias in the moment. If the comment or the microaggression or discrimination is absolutely egregious, it makes sense to call it out in the moment. If there is a risk that a lot of other people who are important stakeholders are listening and are going to take home the message that this is what the organization stands for, it is important to call it out right in that moment. If it is relatively private—so you don't necessarily have folks who are going to take home a wrong message, or if there are other ways to talk to other people about the problem here—then it might make sense to deal with this privately with the perpetrator, the person who made the biased comment or biased act, and to help them really see where the problem lied, what happened, what could they do differently. Because confronting bias doesn't necessarily allow people enough time to really reflect and see what happened. It shuts down the behavior. It says, ‘Don't do this. It doesn't matter whether you understood why or not. Just don't do this again.’ But you need more unpacking if you really want it to be educational. So sit down with that person and help them see what's happening and understand this more deeply, and also get at, ‘Maybe this is a learned behavior. Maybe this is not necessarily just a reflection of who you are. We all have bad habits and we all need to figure those out.’ Having that more honest conversation later might work just as well.

Someone once shared with us that their organization uses the phrase “purple flag” as an internal language to call attention to moments of bias—so if I said “purple flag” in a meeting in response to someone’s offensive comment, we would all understand it as a safe space to talk about why that is a behavior to correct. I'm curious your thoughts on systems like that.

I can see that working if the purple flag is truly viewed by everyone as a safe space to talk. If people are reading it as a slap on the wrist, then it still has the same psychological effects of shutting down, creating guilt and shame. That can be useful to prevent harm in the moment, but over the long term does not create change. I think if those who call out bias or call out the purple flag or whatever, if they also acknowledge, ‘I've done this too, I've made mistakes too,’ coming from that place of humility might take away a little bit of the defensiveness and allow there to be more psychological safety.

What else should leaders take from your work in terms of creating workplaces that foster allyship?

There are a couple of things closely related here. Look at allyship as a developmental process, as a skill that we develop. It's not a yes or a no, you're an ally or not. We don't necessarily think in these black and white binary ways about leadership and customer service—we think of them as things where you develop the skills to do well. Before we start, we are all pretty bad at it, and we all go through this really awkward adolescence phase. We go through a phase of having these ‘aha’ moments and insights that surprise us if we are not punished for every mistake that we make, but shown what those mistakes are and given a chance to learn. If we can take that learning orientation, that developmental orientation, to allyship, that would be a nice basis for allyship programming and training. It's a long-term commitment. We can't just say, ‘We'll have one allyship workshop for two hours once a year, and then we're done.’

Other research on workplace allies has identified an allyship “power paradox”—the idea that someone with more power in an organization is more well placed to be an ally or an advocate for marginalized workers, but at the same time their offers to do so are more likely to come off as condescending or patronizing rather than genuine partners that center the needs of these marginalized colleagues. What are your thoughts about navigating that?

I can absolutely see that. Some of our language around sponsorship and mentorship does assume that people in power are the ones who are meant to take the first responsibility towards allyship. And of course if we do want broad change, systemic change, that is absolutely important. But if we can make allyship accessible to all, so that every single person is stepping up as an ally in an organization, some of these paradoxes will just become less—not necessarily visible, but they just won't hurt our eyes so much, because everyone is engaged.

That's my hope, that we can move toward allyship as being a mass movement. It's not that people in leadership are the only ones who are responsible for everything, and if they fail, the whole organization's diversity and allyship plan has failed. Because they're going to go through their infancy and adolescence phases, especially if that's not what they were trained for. Get it into culture and expect every single person to learn to be an ally, and be ready to deploy allyship whenever the situation calls for it. Just make it part of what it means to have healthy relationships in the organization.