‘Diversity Training’ Doesn’t Work. This Might.

This piece is available in audio format on our podcast, "Heterodox Out Loud: the best of the HxA blog." Narration begins at 1:35.

Diversity-related training programs are intended to serve a range of purposes. For instance, they can provide organizations with a signaling device to show that they are ‘with it,’ and ‘doing something’ about inequality, bias or discrimination. This can be useful for PR purposes, and can also help shield organizations from legal liability. But of course, these objectives are largely implicit: the training would not work as a virtue signaling tool if it was explicitly acknowledged as such.

Instead, the explicit objectives of diversity-related training programs include rectifying inequalities, improving the organizational climate and employee morale, increasing collaboration across lines of difference, fostering free exchange of ideas and information, enhancing the hiring, retention, and promotion of diverse candidates, and more.

Put another way, PR and legal purposes aside, there are real organizational needs that diversity-related training is supposed to help meet. It is no small challenge to create an environment where people with different backgrounds, worldviews, priorities and life plans can forge healthy working relationships, learn from (and grow alongside) one another, collaborate to effectively advance organizational goals, subordinate egoistic and nepotistic impulses to prioritize organizational interests and meritocratic decision-making, etc. As I’ve elaborated upon elsewhere, the typical college graduate is not well prepared to succeed at these tasks. On paper, diversity-related training is supposed to help fill these gaps.

Unfortunately, a robust and ever-growing body of empirical literature suggests that diversity-related training typically fails at its stated objectives. It does not seem to meaningfully or durably improve organizational climate or workplace morale; it does not increase collaboration or exchange across lines of difference; it does not improve hiring, retention or promotion of diverse candidates. In fact, the training is often counterproductive with respect to these explicit goals.

But of course, it is not enough to simply demonstrate that the training doesn’t work. Especially in the current moment when organizations are under immense pressure to ‘do something’ (and organizational leaders often feel a deep personal urge to ‘do something’) about inequality, bias and discrimination. Insofar as the apparent choice is between doing nothing or doing ‘something’ – even if that ‘something’ is demonstrably ineffective or counterproductive – most will nonetheless choose to ‘take action’ rather than remain passive.

As social psychologist Jon Haidt recently put it, what is needed most at this moment is not further demonstrations of how diversity-related training goes awry, but a viable alternative — something that better addresses the diversity-related challenges organizations face… ideally, with fewer negative side-effects.

Fortunately, the empirical literature showing how and why diversity-related training fails also points towards ways that the training could be reimagined in order to better serve its stated purposes. Here, I briefly sketch out some of the core lessons:

Too often, diversity-related training is oriented around goals that it could not possibly achieve. For instance, it is impossible to eliminate racism, sexism or inequality in the broader society, or even in the workplace, through diversity-related training programs. This is simply beyond the scope of what an organizational training session could plausibly achieve.

Insofar as training aims for objectives that it could not possibly realize, it is, of course, doomed to fail. And when the goals of the training are extremely broad or vague, it can be hard to speak meaningfully about the efficacy of the training at all, and it can be difficult for those taking part to see ‘the point.’

Training should instead be tightly connected to specific organizational objectives and the specific tasks different team members are responsible for. This helps participants more easily see the relevance and value of the training. It also helps them put the principles into practice more easily.

An additional benefit of this approach is that it helps keep the training focused on advancing organizational goals rather than, for instance, litigating the history or current state of America in the process of trying to cultivate important workplace knowledge and skills — or debating the extent, causes, consequences and possible solutions of various inequalities. Indeed, needlessly (and often, ham-fistedly) wading into these controversial matters can render it almost impossible for people to actually acquire the core competencies that diversity-related training is supposed to impart, and can damage workplace relationships in the process.

One need not, for instance, internalize left-progressive views on inequality and identity issues in order to effectively collaborate with a colleague on a project (not the least because colleagues who are minorities or immigrants often won’t subscribe to such views themselves). Insofar as training seeks to push controversial moral and political ideologies onto participants in addition to (or at the expense of) providing them with practical knowledge or skills, this often lowers employee morale and generates blowback against colleagues who are women, people of color, LGBTQ, etc.

Rather than having standalone sessions on ‘diversity issues,’ which implicitly frame diversity-related competencies as something separate from, or in addition to, one’s general work responsibilities, diversity-related training should be seamlessly integrated with other forms of training.

Consider management training: avoiding nepotism, cultivating diverse teams, leveraging divergent viewpoints, etc. – these are not supplementary competencies; they are part and parcel of the job. Being effective at these tasks is simply what it means to be a good manager. This is precisely how these topics should be addressed in training. They should be part of general development, and again, tightly connected to the specific tasks and competencies participants are expected to master.

Similarly, in most organizational roles, people are expected to be able to collaborate in teams, to deliver and receive constructive feedback, and to amicably manage disagreements. Diversity-related training should not be approached as something separate from helping people fulfill these general expectations. Instead, considerations about working with people of different backgrounds and worldviews — who may possess different communication styles and behavioral expectations or preferences — should simply be part of equipping people to succeed at their regular organizational duties.

Indeed, freestanding diversity-related modules often reinforce the (false) impression that people from historically marginalized or underrepresented groups are looking for special treatment or playing by a different set of rules. In fact, (with some exceptions) most are looking for the same treatment that is due all employees irrespective of their race, gender, sexuality, religion, etc. Folding considerations about diversity into more general guidance about organizational expectations helps to prevent this misimpression.

And again, by tying the training to the specific roles and tasks people are expected to perform within the organization helps participants better see the value and relevance of the training – and to put it into practice.

Perhaps most critically, this model provides people with multiple touch points on diversity-related content: they receive certain training when initially taking on a role, some supplementary training (as relevant) to assist with particular organizational initiatives, other training when they move into a new organizational role, etc. These trainings would reinforce and build upon one-another (without being overly redundant, as the training would be tailored around specific roles, tasks and goals), helping to mitigate an effect often observed in diversity-related training as it is currently practiced – where the lessons learned from the training erode as time goes on, and people inevitably settle back into familiar patterns of action and interaction, to the point where the training leaves no durable positive effects.

Training on issues like racism and sexism should be discussed as examples of more general cognitive trends (instead of unique pathologies)

People generally gravitate towards others who are ‘like them.’ They often hold negative attitudes, and demonstrate a willingness to discriminate, against those who seem very different from themselves. People make snap judgements about others based on how they present themselves, the context of encounter, and their own prior experiences and background information. People prefer data that flatters what they already believe, and are skeptical of (or resistant to) information that challenges their priors.

Critically, these are not tendencies that are unique to whites, or men, or heterosexuals (i.e. members of ‘high status’ groups) – they are general features of human cognition.

Yet, as political scientist Eric Kaufmann explains, many people from ‘high status’ groups perceive there to be a double standard with respect to these behavioral and cognitive tendencies. Specifically, it seems to be not just acceptable, but encouraged, for members of ‘lower status’ groups to organize along the lines of their race, gender or sexuality — explicitly for the purpose of competing against ‘high status’ groups. However, when members of ‘high status’ groups do the same (to preserve or enhance their own social position, just as everyone else seems to be doing), this is condemned and described in terms of pathology (racism, sexism, homophobia, etc.). This perceived double standard, Kaufmann argues, exacerbates resentment against people from ‘lower status’ groups, and renders people more susceptible to reactionary politics.

Rather than reinforcing this perception of double standards, diversity-related training should start by explaining bias, discrimination, nepotism and motivated reasoning as general cognitive tendencies, which all people are susceptible to – and should equip participants to recognize these inclinations, and to understand how they can derail organizational efficacy. Then, the training can drill into how these broad features of human cognition often express themselves along the lines of gender, race, religion, etc. — again, not just among members of ‘high-status’ groups, but often among minorities as well.

That is, in-group favoritism and out-group parochialism along the lines of race, gender, sexuality, etc. would be described as specific instances of broader cognitive phenomena — and participants would be guided in how to apply general frameworks and tools for mitigating these impulses to the specific cases of interactions across the lines of race, gender, sexuality, etc. This approach helps connect diversity-related training in a deeper way to broader skills and competencies (which renders them more likely to be retained and put into practice), and also helps reduce the stigma, defensiveness and polarization associated with discussions around racism, sexism, homophobia, etc. per se.

Indeed, a version of this approach, Bias Habit Breaking Training, has been empirically demonstrated to durably and positively influence outcomes and behaviors in large-scale, behavioral, and longitudinal studies. More research, of course, is to be done. But the results are consistently promising, and the findings do comport well with what would be expected to work better, in light of the empirical literature on what does not work.

Training should be about managing rather than avoiding conflict

Conflict and misunderstanding are virtually inevitable when strangers with different backgrounds and worldviews, different perceived interests and priorities, with different plans and expectations, are folded into an organization together and expected to pursue common goals (often while competing against one-another in various respects as well) – typically with timeline and resource constraints, with a lot riding on the success or failure of individual and collective efforts, etc. Diversity-related training should be built around a recognition of this basic fact (that is, taking conflict for granted).

Rather than trying to avoid conflict or misunderstanding, it should be oriented towards helping people leverage divergent views, constructively resolve misunderstandings, and think through points of commonality or compromise when interests and priorities seem to diverge. That is, the training should be about the pragmatic task of managing conflicts within an organization. More specifically, it should focus on helping employees manage conflict interpersonally — that is, without appeal to formal authorities.

Establishing formal grievance procedures to help stakeholders resolve their issues through intermediaries has been shown to increase workplace hostility and turnover. Trying to prevent conflicts from occurring among diverse teams is a completely unrealistic goal — the pursuit of which often produces severe negative side-effects.

For instance, presenting participants with a list of taboos and prohibitions can erroneously lead people into thinking that colleagues from historically marginalized or underrepresented groups are fragile and easily offended (often leading them to reduce interactions with ‘those people’ and to chafe under ‘P.C.’ rules). Training should instead focus on helping people have frank and authentic conversations across difference (i.e. not conversations where people from ‘high status’ groups are expected to bite their tongues, and simply acknowledge and validate whatever colleagues from ‘lower status’ groups have to say; instead mutual dialog which recognizes both ’subaltern’ and ‘dominant’ perspectives as useful) while preserving and building relationships — and even having fun.

Indeed, in an environment of trust, openness and collegiality it is far easier to correct misunderstandings and resolve disagreements than it is in an environment where people are constantly worried about causing offense, or about facing social and professional sanctions for any misunderstanding or inadvertent slight, where any colleague can (and is often encouraged to) report on their peers or supervisors for any perceived misstep, at any time. Equipping people with the skills and tools to constructively work through conflict can help drive innovation and bring people closer together.

Diversity-related training, then, should not be focused on avoiding and policing misunderstandings or conflict, but on helping people build relationships and collaborate despite eventual (essentially, inevitable) disagreements — and on leveraging divergent perspectives in order to advance organizational goals.

In order for people to benefit from diversity-related training — to actually internalize the information and skills they are supposed to be gaining — participants have to enter into the training in the right state of mind. If people don’t want to be there, if they feel as though their arms are being twisted, if they don’t see a point or value to the training, then they are unlikely to learn much – and in fact, may (and often do) leave the training with lower morale and higher resentment towards colleagues from historically marginalized or underrepresented groups than they had before.

Again, diversity-related training tends to be more effective, and less problematic, when integrated into more general development or indexed to particular initiatives and priorities. However, insofar as organizations continue to offer freestanding (and more general) diversity-related training, it should be presented as a resource that people can choose to opt into, rather than a requirement that people must complete or face sanctions.

Insofar as leaders wish to encourage higher levels of participation, they can offer slight incentives for those who take part (rather than leveling threats or punishments against those who decline). If people are in a session because they chose to be there, they tend to be far more open to learning than they otherwise would be.

It should simply be acknowledged: the reforms described here would essentially end diversity-related training as it is currently understood — as freestanding sessions focused specifically on issues like racism, sexism, homophobia, etc. (particularly among members of ‘high status’ groups).

In its place would be training on general cognitive distortions which all people are susceptible to – equipping participants with practical tools and resources to identify and account for these tendencies, and to recognize how these general inclinations often play out along the lines of gender, race, sexuality, etc.

The training would take for granted that disagreements and misunderstandings will occur among diverse teams. These conflicts are not signs of pathology, wrongdoing or an unhealthy organizational culture; they are more-or-less inevitable consequences of diversity. But they do not have to be toxic or damaging; they can even be productive.

The training would not be one-size-fits-all, but would instead be built around specific organizational tasks, roles and initiatives – helping people more easily see the relevance and value of the training, and to put it into practice. The training would be seamlessly integrated into general employee development, rather than being its own ‘thing.’

Based on the available empirical research, this approach seems likely to be more effective at achieving the stated goals of diversity-related training. However, it should be emphasized that most of the components of this alternative paradigm have not, themselves, been empirically validated or widely implemented to date. Only after more rigorous testing and implementation can we be confident that this alternative approach will work better in practice (some promising attempts at empirically validating alternatives here, here, here). But in theory, the empirical literature on how and why the prevailing approach to diversity-related training typically goes awry seems to provide us with a promising roadmap of what more effective training might look like.

However, it should be underscored that there are limits to what even the best-designed training may be able to do. Fortunately, there are other tools that are available to organizations to help improve outcomes such as increasing diversity and retention. Formal mentorship programs, making training programs available to all employees to help cultivate skills useful for advancing to the next stage — these can help break up cliques that would otherwise form, providing opportunities to some and not others (typically those who share demographic attributes with people who are already in leadership positions). Properly structured, diversity task forces can also be effective at helping understand how organizations can better recruit, retain and support members of underrepresented groups — and in mobilizing organizational leadership to address those challenges. That is, there are many other evidence-based tools available to help promote DEI goals beyond training; organizations may be well-served by exploring these other avenues as well or instead. Yet to the extent organizations do wish to offer diversity-related training — the approach detailed above is likely to be more effective, or at the very least, less harmful than the methods that currently prevail.