Very few people believe they are prejudiced against others. But according to the author of a deeply thoughtful book, most of us are kidding ourselves. By Greg Dixon

The fear came out of nowhere for Jessica Nordell. While working on her new book in a university library not far from her home, the American journalist noticed two young men rise from their seats and walk slowly towards the end of the room.

It was finals week at this Minnesota university, so the library was quiet but also crowded with students as the two men unrolled mats and knelt. As Nordell realised what they were doing, her physical state changed: her heart rate leapt, her hands dampened, her breathing quickened.

To her complete confusion – and in complete contradiction of her conscious thoughts about Islam – she realised that the unexpected sight of these young men praying to their god had induced in her something also unforeseen. Why had this happened? How was it possible she could consciously think one thing about Muslims while her body seemed to be thinking something completely different?

As it happened, Nordell knew exactly why. In no small irony, the book she was working on that winter's day explained that she was experiencing a phenomenon called unconscious bias.

"I had an involuntary fear response, which was shocking to me," she tells the Listener from her home in Minneapolis. "I didn't even realise that I was carrying that kind of automatic reaction to the presence of Muslims. But I felt it physically."

It does seem irreconcilable – that a well-intentioned, fair-minded person, who consciously and mindfully thinks only good things about most people, might still behave in ways that are discriminatory or biased or fearful.

Yet, as Nordell explains in her book The End of Bias, such behaviour is surprisingly common. While most of us move about in the world intending to be fair, and would not deliberately behave in, say, a racist, sexist or homophobic way, we are all still subject to a phenomenon that can cause us to react to others in automatic, spontaneous, often negative ways that conflict with what we believe our values to be.

It might be an unconscious fear at the sight of Muslim men praying; it might be an involuntary shuffle when a person of a different race sits down next to you on a bus; it could be the way you unthinkingly interrupt when women are speaking, but not when men are; it could be how you reflexively talk down to the elderly or the disabled, or that you instinctively judge people who are overweight.

"Unconscious bias can come out in the way we feel, what we say, the way we act, how we judge a person," Nordell says. "It can come out in so many different ways. But I think the essence of it is that it is a kind of spontaneous reaction based on assumptions and stereotypes and associations that we have absorbed over a lifetime."

Here is another irony: The End of Bias could be said to exist because of unconscious bias. The book, which was shortlisted by Britain's Royal Society for its 2021 science book of the year prize, has been five years in the researching and writing, but its genesis goes back much further, to an experience in the mid-2000s.

At the time, Nordell, who holds a physics degree from Harvard University, a master's in poetry from the University of Wisconsin-Madison and was once a gag writer for Garrison Keillor's long-running public radio show A Prairie Home Companion, was struggling to break into journalism. While she had had some success writing for local and regional publications in Minnesota, she was having difficulty freelancing her work to national magazines.

Then, in 2006, she had a brainwave. With a film version of A Prairie Home Companion soon to arrive in cinemas, and having written a "bananas" essay about being an extra in the movie as well as a writer on the show, she decided to conduct an experiment.

"I was pitching the essay and I wasn't getting any response, [probably because] I didn't know any editors. I was just a random person pitching. But I was running out of time [before the film came out], and I had this panicked moment where I was like, 'I've spent so much time working on this essay, I need to do something or it's not going to get published.' I thought, 'Well, maybe if I send this out under a plausibly masculine name, maybe that would make a difference.' I didn't expect it would actually work; it was sort of a Hail Mary. I sent it out as 'JD Nordell'. It was accepted by [online magazine Salon] within a couple of hours."

The JD ploy, which she used for a couple of years, helped her launch her journalism career nationally, and she has since been published by the likes of the New York Times, online magazine Slate and the Atlantic. However, the primary outcome of this experience of apparent unconscious gender discrimination was that it fired a deep professional interest in bias.

"I was so shocked that this [experiment] had worked. I was also quite convinced that the editor of this very progressive online magazine had not intentionally set out to block women from publishing in his magazine. So this situation, this phenomenon of people being well intentioned on the one hand but then being susceptible to sexist or racist reactions on the other, became really fascinating to me."

It's called the "prejudice paradox", this gap between word and deed, and it was first identified by psychologists in the mid-1980s. Their research showed that while a vast majority of white Americans said they opposed racial prejudice and held no racist beliefs, it turned out many still behaved in discriminatory ways. How could this be? The conclusion by researchers was that the white people who said they weren't racist were simply lying.

However, as Nordell reports in The End of Bias, a young researcher called Patricia Devine, now a professor of psychology at the University of Wisconsin-Madison, found it was rather more complicated. Through her research, she concluded it was not conscious lying that created the prejudice paradox, but unconscious bias.

Devine's work revealed that someone could in fact simultaneously hold beliefs that they consciously endorsed, while at the same time harbouring negative stereotypes and associations about people that they did not consciously believe were in conflict with their values.

How can it be that these two things can exist together in the mind? "It's a really good question," Nordell says. "I think we all have some kind of combination of beliefs, stereotypes, associations and assumptions. What I've really come to believe is that these sort of narratives of human hierarchy, whether we're talking about male supremacy or white supremacy, are so deep, they're so deeply embedded in our culture that we don't, as individuals, understand the extent to which we have absorbed these ideas."

Having biased associations does not mean you're a bad person — it means that you are a product of your culture. From childhood, we absorb cultural knowledge through our parents, friends, schooling and the media and begin learning to categorise the world and the people in it as we go. Unfortunately, these categories aren't simple mental signposts. Our cultural experience also attaches stereotypes, and both negative and positive associations to these categories.

"When we stereotype a particular group, it is often not a matter of what sort of characteristics are common in that group but rather what characteristics that we've learnt that make that group different from another group," she says.

It is how our brains use these stereotypes and associations that generates unconscious bias, also called implicit bias or unexamined bias.

As Nordell explains in The End of Bias, stereotypes and associations attached to categories are functional. They help us make sense of the world, and predict the world. We unconsciously put them to work at every opportunity.

In other words, even if these stereotypes are at odds with our values, we might find ourselves unconsciously deploying them because they are useful mental shortcuts.

These unthinking, spontaneous, automatic judgments we make about others through stereotypes provide a foundation for both personal and institutional prejudice. Not because we are simply noticing differences between people, Nordell says, but because our culture tells us the differences matter. And for those on the end of those prejudices, it can limit choices and futures, stall careers, ruin health and, in the case of black American men, get them killed.

On a hot July day in 2016, 32-year-old Philando Castile, a school nutrition supervisor, was pulled over by a police officer while driving with his girlfriend and her daughter in St Paul, Minnesota. The suburb they were stopped in was predominantly white. Castile was black.

While reaching for his driver's licence to give to the officer (a 28-year-old called Jeronimo Yanez), Castile informed him he had a firearm on him. It was a weapon he was permitted to carry. He owned it because he lived in a high-crime part of town.

This statement proved fatal. As Castile reached for his wallet, Yanez heard "firearm" and demanded Castile not reach for the gun. Castile, reaching for his wallet, said he wasn't going for his gun. The officer pulled out his firearm anyway and shot Castile seven times.

The medical evidence showed Castile was not reaching for his gun. Yanez was charged with manslaughter and gross negligence. He got off. The jury of 10 whites and two blacks believed Yanez when he said he feared for his life and was not lying about Castile's hand being on his gun. As Nordell notes, the jury extended no such benefit of the doubt to the victim.

Crucially, Yanez's defence argued race wasn't an issue in Castile's death. But Nordell believes differently. "In fact, research shows that race undoubtedly played a critical role, not only in Yanez's decision to stop Castile," she writes in The End of Bias, "but in his perception of the encounter and in the production of fear."

Although a complex web of factors came into play, including US police culture, American history and, quite possibly, the country's problem with overt racism, Nordell believes unconscious bias in the form of fear-based perceptions of black Americans is enormously influential in their frequent deaths at the hands of police.

"I think it is a very significant element," she tells the Listener. "We have so much research that shows that Americans judge black men to be more threatening than white men. And we know, often in situations where police use violence, it is because they have an immediate sense of threat.

"We could have a whole conversation about whether this is implicit bias or whether this is deeply held racism on the part of police officers. I think it is often an unknowable combination of these things. But I do believe, even in the most egalitarian-minded officer, there is the possibility of unconscious racism in the way that they encounter the [black] community."

In other words, the implicit racism that deems all black men threatening made Castile dangerous to Yanez simply because of what he was, not who he was, leading to Yanez's instinctive, unconscious fear response.

"We know Yanez did not see Castile's hand on a weapon," Nordell writes. "Yet Yanez, predicting violence, shot. Mind, body, history and institution: fused together to lethal effect."

For many black Americans, particularly young men, being perceived as a threat by others is part of their everyday experience. However, even when everyday bias doesn't bring with it the threat of being shot by a police officer, the everyday experience of unconscious bias – be it sexism, racism, ageism, homophobia – inevitably has ongoing impacts on lives.

The cumulative effect of unexamined bias over the course of a career or a lifetime is largely an unknown. In preparing The End of Bias, Nordell approached dozens of researchers who were studying prejudice and discovered that no long-term studies of the effects of unconscious bias had been done.

The reason for that is simple enough: the history of the study of prejudice — any kind of prejudice - is not old, and our psychological understanding of prejudice has been subject to change. Until the early 20th century, a lot of what is now thought of as prejudice, psychologists just thought was true. It wasn't until the mid-20th century that the psychology community – which eventually defined implicit bias – started to recognise prejudice and the idea of human hierarchies as problems that needed solving.

"So if you think about that, maybe it's not so surprising that it took us a while to understand implicit bias, because prejudice itself wasn't really recognised until 100 years ago. Then, over the following 100 years, there were a lot of different ideas about what the causes of prejudice were, from pathological personality types to institutional problems. It wasn't really until the 1980s and 90s that people started to connect the idea of unconscious, automatic, spontaneous mental processes with this notion of prejudice."

The lack of research into long-term effects gave Nordell an idea for another experiment. With the help of a computer scientist, she created a computer-based trial to simulate the effects of unconscious sexism on a number of women in management and supervisory roles in a company she dubbed "NormCorp".

The experiment was partly influenced by a 2011 US Supreme Court decision, which found against a class-action suit by 1.5 million female Walmart employees who maintained the company denied them promotions. The company records, despite its equal-opportunity policies, showed there were huge disparities in pay and conditions between men and women across its operations. Even so, the Supreme Court found against the plaintiffs because the justices believed it was not possible for such disparities to occur without a co-ordinated master plan for prejudice. The real possibility that individual managers making decisions over a long time had created the problem, without some sort of nefarious corporate plot, was discounted.

The NormCorp computer simulation found that even mathematically tiny differences in the treatment of men and women had an enormous cumulative effect on the number of women in supervisory roles, with the biggest difference in senior management. Nordell was surprised such a small amount of bias resulted in such a male-heavy senior leadership in her fictional company. "But at the same time, the numbers that we ended up with are actually sort of similar to what we see in the real world," she says. "So, in that way, it wasn't a complete shock."

NormCorp simulated just one implicit bias – one resulting in sexist hiring and promotion by individual managers – across one organisation over a limited period of time. In the real world, of course, individuals subjected to unconscious biases can face a lifetime of frequently subtle, and often plausibly deniable, "everyday" prejudice. The long-term effects can only be known to them.

Nordell didn't want to write just about a problem. She has spent the past 15 years doing just that through her journalism about unconscious bias, its impact and how, scientifically, it works. With The End of Bias, she is genuinely in search of the end of implicit bias through real-world solutions.

And her book is filled with examples, primarily though not exclusively American, of how implicit bias is being combated in companies, university departments, medical trauma centres, hospitals, classrooms, law practices and police forces.

Some of the solutions involve carefully designed training workshops (including one created by Patricia Devine, from Wisconsin-Madison).

Some involve culture, policy and structural changes within organisations. Yet others can be as simple as creating a checklist to make sure that unconscious biases are avoided.

There is no silver bullet for putting an end to the phenomenon of unexamined bias, Nordell believes. But an understanding of history seems to be essential in transforming minds and cultures, something suggested by what is called the "Marley Hypothesis".

This takes its lead from two lines – "If you know your history/Then you would know where you're coming from" – in the Bob Marley song Buffalo Soldier.

"The Marley Hypothesis holds that as a person's understanding of history increases, their ability to recognise present-day discrimination also increases. There is some early promising research that suggests that this is true," Nordell says.

Importantly, too, there is a solid business case for minimising unconscious bias in the workplace.

"When people feel they are being treated unfairly, it really negatively affects people's commitment to their jobs. So even a cold-hearted analysis from a business owner would reveal that this is a really important thing to do if you want to keep employees actually engaged and committed to work."

What is also quite clear from her research is that eliminating bias, even within transformed institutions and workplaces, requires individuals to mindfully, actively and continually recognise and confront their own implicit biases – something Nordell has had to do herself. Throughout researching and writing The End of Bias, she had to face her own prejudices, just as she did in that university library.

So what bias did she discover in herself? "Every one in the book," she says, then laughs. "I had to confront how ideas of male supremacy had influenced my ideas of my own worthiness and my ideas about the worthiness of other women. I had to confront how the patriarchy had infected my thinking. I had to confront white supremacy and how the idea of white supremacy had infected my thinking, and my own unexamined notions of racial hierarchy. I had to confront ageism.

"I had to confront all of these biases; they all live inside me. So there was no way to separate myself from the material."

So is reducing unconscious bias the responsibility of individuals, like her and me and you, or should it be done through top-down structural change in organisations, or perhaps through better anti-discrimination laws?

"I don't think laws are the complete answer. We've seen that already. We have had civil-rights legislation, we've had equal-opportunity laws. Laws only put a limit on how bad things can get.

"I think, certainly, it is our individual responsibility – and privilege, honestly – to change and transform. And it is also our role to create new and better structures, too."

• The End of Bias: How we change our minds, by Jessica Nordell (Granta, $36.99).