The Unconscious Bias Approach to Equality

Unconscious Bias refers to the learned stereotypes about groups of people that are formed outside of conscious awareness. They are automatic, unintentional, deeply ingrained beliefs that have the ability to affect our behaviour.

Scientists believe that stereotypes in general serve a purpose because clustering people into groups with expected traits help us navigate the world without being overwhelmed by information.

Whether we realize it or not, our unconscious biases influence our professional lives, from the way we think to the way we interact with colleagues. Unconscious or implicit bias can lead to instinctive assumptions that a nurse must be a woman, or an engineer must be a man. We have all been exposed to thousands of instances of stereotypes that have become embedded in our unconscious minds. The evidence is overwhelming that unconscious bias seeps into decisions that affect recruitment, access to healthcare and outcomes in criminal justice. This is a very useful outcome of the psychology on unconscious bias.

A common example of bias is ‘Affinity Bias’, our instinctive reflex to more easily trust and like people we see as similar to ourselves, such as preferring someone who went to the same school. 

More and more organisations are now looking to train their staff on unconscious bias awareness to offset the risk of discrimination. But does it work?

The unconscious bias approach seeks to raise awareness around the mental shortcuts that cause us to quickly make (often misguided) assumptions about a person’s ability or character. However, current understandings of how to approach unconscious bias in practice are limited. Much of the controversy centres on the unreliability of the standard tests of bias, and how little they predict behaviour.  

Research shows that awareness alone doesn’t lead to behavioural change. There is more success in changing behaviour in approaches that focused on long-term actions, such as habit-breaking, than in simply raising awarness. Attempts to combat bias benefit from more evidence-based exercises that increase participants’ self-reflection such as perspective-taking through writing about challenges faced by other groups, alongside concrete steps for improvement. Explicit demonstration of next steps are needed to take us from awareness to action. The unconscious bias awareness approach is limited in this regard. It generally does not identify the actions to take or provide support for doing the work that needs to follow. and those people who do want to take action can find themselves experiencing higher levels of anxiety as a result of their new awareness – and we know that anxiety drives bias. 

The mandatory nature of most corporate unconscious bias training, without explicit actions to follow, can also leave participants doubtful of what an organisation wants from them, and resentful that their personal emotions and opinions seem to be usurped by their employers’ needs. Feeling confronted about biases undermines the reflective work necessary for an awareness-led approach. Confronting social differences and ingrained ways of thinking is challenging, even frightening, and the easy reaction is to defensively reject challenging new information or changes. 

Some people will feel that because they can identify discrimination they are less likely to be bias, therefore ignoring their own biased behaviour. Others can become complacent about their own biases, avoiding responsibility for avoiding discrimination following training, perhaps because of a belief that the workplace has been made free of bias.

Many researchers now believe that unconscious bias awareness and training can be actively harmful (see links below) if it is promoted as the primary answer to discrimination. It centres the vulnerability of the bias perpetrator to anxiety, instead of the people who experience discrimination in a wide variety of ways. It takes the focus off organisations with high levels of resources, and puts it on individuals with relatively little influence within the organisation. This is particularly problematic when those individuals are under constant pressure to make decisions fast, with limited information, and limited evidence about the impact of the organistaions’ decisions as a whole.

The key is in the impact that our biases have on our behaviours, collectively, and how these are embedded in widespread patterns across society.

How does bias produce discriminatory outcomes?

In terms of our institutions, the police, doctors, politicians, teachers, they all hold unconscious bias. 

Those assumptions might be set out in policy, learned through accepted practice or be part of a culture of an organisation. And of course people bring their assumptions to work, from what they hear in the media, or learn at home.

That’s how biases get ingrained into how we work. It’s why doctors turn away certain patients quickly, why police arrest some kinds of people quicker than others, why teachers are more likely to diagnose disability in some children than others, and why politicians are more likely to listen to some groups than others. They rely on a key set of assumptions about the people they interact with (or don’t) which influences their decisionmaking. For example, when  software companies first used AI in gaming, they did not train or test it on groups who weren’t mainly male, adult and white. 

Our beliefs drive what we do and how we do it. Often of course those are at the level of conscious bias – we know we are using those biases, but we wouldn’t admit it openly. 

In a moment where people have to make quick decisions, moments where they are stressed, or moments they are not thinking about their actions, those biases can completely override any rational thought, and this can be dangerous. This has been proven for example in the way biases arise in the work of tired doctors on call. But biases can also become part of our working practices – for example in how women are frequently given lower pain relief, and Black patients even lower. 

Unconscious bias itself is not the main problem – it’s how fast that turns into a factor that determines our behaviour and our choices, how reliant we are on our biases, and how deeply ingrained they become within our institutions and practices, so we no longer recognise them as biased.

The structures that reproduce bias

Structural racism describes the way public policies, institutional practices, cultural representations, and other norms work in various, often reinforcing ways to perpetuate racial inequalities. It identifies dimensions of our history and culture that have allowed privileges for white people and exclusion for racialised people to endure and adapt over time. Structural racism is not something that a few people or institutions choose to practice. Instead it has been and continues to be a feature of the social, economic and political systems in which we all exist.

Disproportionality – usually showing the under-representation of a particular group in an area or profession compared with their number in the population – is now a very popular way of ‘proving’ racism. But it actually gives is a snapshot of something at one stage in a system. And this makes us focus on one point, at which unconscious bias can be assumed, rather than how discrimination is produced and reproduced at multiple stages in any institution.

If we are only aware of the role of unconscious bias, and not how it interacts with systems and processes to reproduce bias, we only get half the answer we need.

Did the tech companies fail to test their AI products on women and Black people because of unconscious bias? Or because this is an industry that has systematically excluded both from employment for decades and promoted a culture where affinity hiring was rewarded, and where online abuse and harassment – and even overpolicing – of women and racialised people was seen as unimportant consequences of tech products? Seeing these points of inequality within the larger institution is important, whether that’s a company, industry, or sector.

What can we do individually?

First – understand that the world is made up of institutions and structures (organisations, states, ways of doing things) that shape how we make decisions and what information we use to do that. That’s as simple as what information we hear on the radio. How even a joke about disabled people or a throwaway comment on refugees can become a negative belief we use. 

Racism, for example, is a structural issue. Most institutions that structure how we live our lives were built at a time when the world was oriented towards an idea that there was a racial hierarchy of intelligence and moral superiority. Many of our institutions have outwardly changed, but not lost aspects of that approach. It’s in who is considered valuable in their work, whether that’s the employee or the service user. All institutions reproduce a society’s biases, but they differ in the extent to which those are embedded in policy and practice. 

Second, give yourself the opportunity to make decisions on evidence: 

Slow decision-making is proven to be better because it relies more on evidence than instinct. Slow down and review your personal or professional decisions. The point is learning to change, not adopting guilt and becoming paralysed by it. Try not to be defensive with yourself about your choices. 


  • how we decide everyday things
  • what we find threatening or anxiety-inducing
  • affinity bias (halo effect) and people around us
  • confirmation bias and how we treat information we don’t agree with 

If you think about the things you can do individually, you may get stuck at the points we just discussed, where you’re reflecting on your own biases. But probably you’ll do that once and forget about it. Actually, it’s important to remember we are constantly creating biases out of received information. So, building in prompts and practices that allow us to consciously make better decisions is a really helpful way of approaching this for the longer-term. We do it when we want to go on a diet, build an exercise habit, get our kids used to going to bed or school at a certain time. Prompts and new practices can take time to bed in, and work best when you involve the other people who you would like to use them too, at home, in your community or in your workplace. 

Third, remember you have a lot more power than you think. 

The main problem comes when we think that (a) we can’t really change ourselves, and more importantly (b) we can’t change the society around us, because everyone else is biased. Plus we tend to think of other people as more biased than ourselves. This means that there is no point in doing more than acknowledging its existence. But the biases that are in society now were created through the institutions and systems we have, and so they can be undone too. It is entirely possible to remake the world. Alana Lentin’s work on antiracism captures this elegantly – she points to the creation of racial hierarchies on purpose as proof that we can undo them. 

Today we don’t always see explicit racism in institutions like we would have 20 or 30 years ago, but we do see it in the outcomes of education, employment, policing, health and community. We see the impact of racism – as a founding principle of those institutions – still determining who gets the same chances as who does not.

So do you have to tear down those institutions? Not necessarily – there’s never before now been a concerted effort in mainstream society to create anti-racist institutions, and collective effort and investment in that could prove remarkably positive. What’s interesting right now is how much companies don’t want to be seen as racist and there’s a lot of PR work towards that, but not a lot of work behind the scenes to change the impact of their service or product on different groups. 

What are the things we can do collectively? 

At the level of society, we are interested in how institutions can be changed and reshaped to eliminate not just bias in today’s decisions, but the ways that bias has shaped the institution as it developed. That means thinking about how our schools, workplaces, health systems, policing, the arts, and how we build our neighbourhoods affect groups differently. How they exclude or minimise opportunities for people because they are female, older, disabled, LGBT, a migrant, or from a certain ethnic or religious background. That means paying attention to how the state treats people differently in practice, even where there are laws that say we are equal. The treatment of autistic children in education is a good example, or the way the law makes it difficult to prove discrimination at work. These are mainly examples of conscious bias, but the reason we can accept them is because they work with our unconscious biases about who is valuable in our society. So as a society, we need to ask more questions about how fairly are people treated now who have historically been treated as less. 

You can bring what you now know to the spaces where you have a role – any role – you don’t have to be the boss or the leader. If you are the mum who organises coffee or play dates, if you are the churchgoer who organises events, if you are the artist or musician who looks for collaborations, if you are in a community organisation or a member of a sports club, if you have a job (any job), then you are part of a collective that is making biased decisions all the time.

You can pay attention to what that looks like, and start to ask questions about the way people are invited to take part, who is included, what criteria there are for membership, or certain roles, and you can start to think what that collective or organisation would look like without those biases.

There are lots of examples of ways that hiring managers can debias applications or interviews, for example, or how sports or music clubs become more inclusive. Many organisations don’t realise how their policies and practices disadvantage certain people. But you might have a role in deciding funding, or planning, or how a service runs. If so, you will want to think about the evidence that gets put in front of you about the impact on different groups, and what information you are not getting. Often information isn’t collected on the things we don’t want to know – because our biases tell us its not priority.

So (1) what does your group do that impacts people differently, and (2) what evidence is that based on? Then thirdly and lastly, how do we keep on top of things once we get the ball rolling? We build in prompts to review what we are doing, and if you’re in a formal organisation, you can set targets to hit and maintain. If you were saving money for something, you’d check your balance and progress regularly – the same goes here. 

Learn more:

For background on the psychology of unconscious bias, we recommend Pragya Arwal’s book Sway 

For more on institutional racism, see and

For courses on addressing racism, try the Irish Institute of Antiracism and Black Studies who do online courses in Ireland on Black history and studies, or Rachel Cargle’s website The Great Unlearn. 

For more on critiques of the unconscious bias approach, see

Briefing authored by Daniel Reynolds and Lucy Michael