Implicit Bias: The New Reality of Racism

“Your data suggest an automatic preference for European Americans over African Americans”

The words glared at me accusingly.

Here I am, a vehement proponent of humanism, with the firm egalitarian belief that regardless of a person’s race they are fundamentally the same – fundamentally human, and therefore deserving of equal treatment and respect. And yet this test – designed and implemented at Harvard University – was telling me, in no uncertain terms, that I’m racist.

In this article, I’m going to explore a quiet, insidious presence that lurks within an overwhelming majority of people – including those, like myself, that might otherwise forcefully deny that they harbor even the tiniest sliver of racism. This presence is what was apprehended by the test I had taken, and held up to the light so that I could see clearly its monstrous shape. It manifests itself in many different forms, as revealed by a number of emerging studies, and its effects are demonstrably harmful; it is known as implicit racial bias.

We are living in strange political times. Thanks to the far left’s vituperative and overbearing crusade of multiculturalism, any opinions that cause even the slightest friction against their own, are liable for swift condemnation.

The recent Evergreen College fiasco[1] is one lurid example of the reactionary and unreasonable nature of this vocal ‘regressive left’, and how, despite the fact that they constitute a relatively small fraction of the population, they are nonetheless a plague on rational discussion.

I originally planned this article to be a discussion of the myopic and fickle willingness of the regressive left to paint the red mark of ‘racist’ or ‘bigot’ onto anyone who dares to express an argument that even vaguely clashes with their rigid ideology.

In particular, I wanted to outline the now almost meaninglessness of the word ‘racist’, since apparently it can be contorted to mean something entirely different from its original definition of ‘a person who is consciously and openly prejudiced or discriminatory against another race’.

You are now a racist, it seems, if you so much as acknowledge real differences between races, whether those differences are biological, or cultural in nature.

You are a racist if you dare to question the righteousness of any aspect of the regressive liberal movement.

You are a racist if you fail to automatically genuflect to the religious beliefs of Muslims.

You are a racist if you are white.

Perhaps I’m exaggerating, but I think you will agree that, though the word has lost none of its sting, it’s now being wielded much more liberally (pun intended) than is warranted by its original definition.

But, in my research for the article, in which I hoped to highlight the difference between actual, run-of-the-mill racism, and an honest attempt to discuss real cultural or ethnic differences (e.g. discussion around the religion of Islam, or a critical assessment of the Black Lives Matter movement), I found evidence that gave me pause.

Perhaps amid the many cloying howls and lamentations for social justice, there were a few voices worth listening to.

Despite my naïve optimism that racial discrimination had been mostly eradicated from our society, I came across incontrovertible evidence to the contrary.

I’m not talking about a small, persistent remnant of bigotry contained within the outspoken KKK, or other pro-white groups; this is a beast of an entirely different nature.

This is a form of damaging racism that is distributed throughout all of society, a type of racism that you, dare reader – yes, even you, might be guilty of.

It is known by several different names, but here I will use the term ‘implicit bias’, and I’ll explain why I’ve chosen to use this term, and why this terminology is important, later in the article.

The social justice warriors, it seems, may have some credibility when they allude to the existence of a ‘systemic racism’ towards blacks, even if they miss the mark somewhat with their own conception of the phenomenon, and therefore with their arguments.

You see, this ‘systemic racism’ is not an institutionalized, or even a consciously perpetuated form of discrimination. Rather, it exists within the system of our subconscious thoughts and behaviours, and as you’ll see, it can manifest in ways that are worthy not only of our attention, but also our collective commitment to quashing it outright.

The ubiquity of implicit bias

Implicit bias occurs when a person is possessed of subconscious preferences for, or aversions to a particular group. A ‘group’ does not necessarily mean race, but simply any collection of people that can be consciously or unconsciously categorized together.

For example, there may be implicit bias towards older people versus younger people, fat versus thin people, gay versus straight people, religious versus non-religious, and so on.

As we’ll see, these biases are well documented, and found to be present in most people – even those who truly consider themselves non-biased; who fundamentally believe in the absolute equality of all individuals.

The bias is primarily subconscious, which is what makes it so pernicious; how can we correct a problem we don’t realize we have?

‘Project Implicit’ is a Harvard initiative that seeks to reveal these unconscious intergroup biases, through tests that can be taken by anyone online.[2]

The tests are relatively simple, and they work as follows:

You are shown either a word, with good or bad associations, such as ‘happy’, ‘love’, ‘rotten’ or ‘dirty’, or you are shown a face of someone who is a member of a particular ‘group’, e.g. a black person or a white person.

You are asked to press one of two buttons – the left button for ‘bad words’ and the right button for ‘good words’.

The test then progresses through different stages, where you either press the left button when you see black faces and the right button for white faces, or vice versa.

The theory behind the test is that if you possess an implicit bias – of aversion for one group and/or a preference for another, then your ‘in-group’ will elicit positive associations, and your ‘out-group’ will elicit negative ones.

Therefore, you will find it easier to lump the good words and the in-group together – so your reaction time in pressing the button where they coincide is much quicker than for the opposite situation; where the out-group and good words share the same button.

Conversely, when the out-group and bad words share the same button, your reaction time in pressing it will be quicker.

The data collected from these tests appears damning. An overwhelming bias exists in most people, for one group or another.*1

But though each of these various biases are extremely important to understand and combat, I want to focus exclusively on the racial bias, for this article.

Implicit Racial Bias

The implicit association test described above reveals a strong bias of preference for white people, and aversion to black people.[3]

But for those convinced that the test is too contrived and simplistic to possibly reveal an actual bias – or those who concede that the bias might exist, but deny that it has any real world impact: there is a deluge of disturbing experimental evidence from other studies that reinforces the results of the implicit association tests, and reveals just how consequential this bias really is.

We’ll take a look at several areas where implicit racial bias has been found to rear its ugly head, to show that the issue is undoubtedly real, and far from harmless.

But before we do, it is important to understand something about the nature of the phenomenon.

Since implicit racial bias is largely unconscious, people do not recognize situations where they are under its thrall. These same people hold conscious beliefs in equality, and as such would never be found to overtly favour one race over another.

But where we do see the manifestation of their unconscious bias, is in more subtle situations, where race is hidden amidst other variables.

These other variables are then broadcasted by the racially biased person as the ostensible reasons for a decision which had in fact been, at least partially racially motivated.

But the biased person really believes that their decision had nothing to do with race; the self-deceit is unconscious, and so their bias escapes unnoticed. That is, until it is revealed by experimental data.

Employment and Admission

An experiment by Dovidio and Gaertner[4] gave white American college students an opportunity to recommend candidates for a prestigious campus position. The racial identity of each (fictitious) candidate was clearly advertised, along with their qualifications.

In the instances where a particular candidate had either very strong, or very weak credentials, they were either reliably recommended for the position, or advised against, irrespective of their race – which is consistent with the assumption that American students generally do not possess an overt, conscious bias for one race over another.

However, in cases where a candidate’s credentials were neither strong, nor weak, an implicit racial bias was revealed; white candidates were reliably favoured over black ones.

A similar study by Marianne Bertrand and Sendhil Mullainathan[5] involved the sending of fictitious job applications to various employers that had advertised an open job position. The resumes were randomly assigned either an African American-, or White American- sounding name (Jamal, versus Greg).

It was found that white names garnered a 50 percent higher callback rate than African American ones – indicating that an employers decision was heavily influenced by his or her perception of an applicants race.*2

Legal Decisions

A study by Johnson et al [6] showed that white jurors were more lenient on white defendants than black ones, under certain circumstances.

When the jurors were privy to admissible evidence only, their decisions were not influenced by the race of the defendant – again, consistent with fact that people are not overtly discriminatory.

But when the jurors were exposed to evidence that incriminated the defendant, but was also deemed inadmissible in court, they proceeded to judge black defendants more harshly than white ones, under the same circumstances.

Police Discrimination

 The issue of whether or not racial prejudice exists within the American police force is highly controversial. Rhetoric and narrative spinning aside, the only way to reliably ascertain whether racial bias manifests in police actions is by trawling through the data.

What the data seems to reveal[7], is that implicit bias does indeed permeate the police force – and with what we have already seen above, should that really come as any surprise?

Of course, this by no means suggests that the police force is ‘racist’, or that any given officer is likely to shoot the next black person he deems suspicious.

The data simply reveals statistical anomalies – small impurities in the aggregated pool of total recorded incidents, which says only this: the subtle presence of implicit bias may account for a slight statistical lean towards white-over-black favouritism in police-civilian interactions.

This ‘slight statistical lean’ though, is clearly a devastating problem, in terms of the reality it betrays.

The lean is embodied by the gun triggers of a thousand cops in a thousand ambiguous situations, and perhaps implicit racial bias has its finger on one or two of them.

Underlying cause of implicit racial bias

So we have diagnosed an insidious and noxious condition that exists in our society, but in order to treat it, we should first understand its root cause.

There are two possible explanations:

First, it might be the case that intergroup bias is an innate feature of human nature, which evolved in accordance with the tribal structure of early human societies.

Amidst fierce competition for resources, it would have been evolutionarily beneficial to favour one’s genetic kin in the acquisition, distribution and retention of these resources.

To harbor an egalitarian commitment that extended beyond one’s own tribe to include genetic competitors, was to disadvantage one’s own genes. Such an egalitarian-minded human would have been happily exploited by many, who would not return the favour – thus such aberrant behavior should have been selected against.

So it is quite possible that nature could have instilled in us a preference for those we perceive as belonging to the same group as us, and a distrust for those we do not.

The second possibility is that bias towards groups is learned.

Rather than having in-built racist tendencies, humans might simply possess some kind of neural software that allows in-group and out-group models to be formulated by experience, as our cognitive, conceptual associations of different categories of people build up over time.

Despite efforts to quash them, stereotypes still linger in society. Narratives about the inferiority of certain groups persist, even though they may emanate from the relatively small groups of explicit racists, or bigots.

So perhaps racial bias can be explained by these discriminatory residues, combined with our innate proclivity for categorizing and stereotyping into in-groups and out-groups.

One way we might ascertain whether implicit racial bias is innate or learned, is by comparing the bias of whites towards blacks, with the bias of blacks towards whites. If the bias in each case was equal and opposite, then the effect would likely be explained as a purely innate, genetically determined feature.

But what the data from the Implicit association test shows, is that while a majority of white Americans display a preference for other white people; black Americans are split down the middle – with half harbouring a bias in favour of other black people, and the other half against them.[8]

What this seems to indicate is that implicit bias is largely a product of subconsciously instilled attitudes, which are themselves likely a product of undercurrents of racial differentiation in society.

Overcoming implicit racial bias

So what is the solution to the problem?

Must we suppress any and all references in society that illustrate racial differences – ‘light-hearted’ or otherwise – as the new wave of social justice warriors seem intent on? Should we ban the politically incorrect jokes of stand up comedians, and the dissemination of any data that shows racial discrepancies in, for example, crime rates? Should we block all discussion that so much as hints at variation between race, or cultures?

Clearly not. Not only would such measures be impossible to implement, but any decrease in implicit bias they might yield would not be worth the cost of abandoning freedom of speech, and the social and political progress that might otherwise be made.

Fortunately, however, the way forward in curtailing the negative effects of implicit bias has already been illuminated through various studies, and it turns out there are two primary approaches.

The first is to induce a re-structuring of people’s internal models of their in-groups and out-groups.

Specifically what this entails, is encouraging people to formulate a new, common in-group identity that is superordinate to racial group identity; that is, the new group contains the different racial groups as members within it.

In such a group, members of one race would lump themselves together with members of another race – and use the pronoun “we”, rather than “they”.

With this invisible bond of kinship, the prevalence of bias is reduced.

Many studies have provided support for this theory[9][10] so it is abundantly clear that the ‘group models’ with which humans categorize themselves and others are not genetically fixed, but promisingly malleable.

One such study[11] was conducted at a University of Delaware football match. The experiment involved either a black or a white interviewer who would approach white fans as they entered the stadium, and ask whether they would like to participate in a survey of food preferences.

The interviewers would alternate between wearing a hat of the home university or the away university, such that test subjects could mentally categorize them into both racial in- or out-groups, and university in- or out-groups.

The experiment measured the likelihood of whether subjects would agree to an interview, as a function of whether the interviewer was black or white, and what university their hat represented.

It was found that, in the cases where the interviewer was white, the hat they wore had little effect on whether a subject would comply.

However, when the interviewer was black, subjects were significantly more likely to comply when the interviewer wore a hat indicating they were from the same university.

Thus, in these cases, subjects categorized the black interviewers into a new in-group, which superseded the racial one, and their implicit racial bias was reduced.

The problem with this method of creating common in-group identities, is that its long-term effectiveness is questionable.

For a persistent, widespread effect, it would be necessary to deeply inculcate some new common group identity that would be as salient and discernible as skin colour, or a university hat.

Sure, we could all walk around wearing “Team Earth” t-shirts, but unfortunately I don’t really see that catching on.

So it seems that the best solution to mitigate the effects of implicit racial bias lies in the second option: awareness.

It has been demonstrated that when the existence of racial bias is pointed out to someone, the effects of the bias are reduced, or eradicated entirely.

For example, one study[12] examined physicians’ treatment recommendations for black, and white patients.

In the control case, physicians were more likely to recommend aggressive treatment plans for white patients, presumably as a result of implicit racial bias.

However, in the experimental condition where physicians were first made aware of the possibility of an implicit bias affecting their treatment recommendations, the disparity between treatment plans for black and white patients disappeared.

So then, it would appear that the most effective, and implementable means of combating the effects of implicit racial bias is by making people aware that the bias exists, so that they might be more reflective in situations where race might otherwise become a factor in some evaluation.

Discussing implicit racial bias

It is essential that we don’t conflate implicit racial bias with the original concept of ‘racism’ because, though both are egregious encroachments on equality, a conscious, explicit attitude of racism is infinitely more execrable and worthy of harsh condemnation than a subconscious perception we might carry around with us, since that is largely beyond our control.

So we must have compassion for the problem, and not ostracize people for their implicit racial baggage, but at the same time we must not normalize it, or accept it as an immalleable feature of human nature.

One means of maintaining the distinction between implicit bias and racism is simply through the terminology we use. We should continue to use the term ‘implicit racial bias’, with a clear understanding of what it entails, as opposed to lumping it together with ‘racism’.

Yes, I’m aware of my own hypocrisy here – even the title of this article I have used the term. But the fact is, until the distinction is firmly ingrained within our collective consciousness, it is necessary to draw people’s attention to the issue of implicit racial bias with the use of such language – so long as its distinction from racism is made explicit, and the new term of ‘implicit racial bias’ is adopted to illustrate it.

And illustrate it we must – if we hope to fix it.

So spread the news; the fix is in: racial discrimination remains prevalent in our society, and it won’t go away until we acknowledge its existence.


*1 Cautionary notes on over-interpreting results of the IAT

*2 The study addresses the possibility of whether the employers’ bias was actually due to perception of social class, rather than race, by showing that African American applications that also displayed a residence address that indicated higher socio-economic status, were not favoured more than whites.

Also, a similar study, that used common racial last names failed to replicate the findings of the original experiment. But, while the last names used for the Black applicants (Washington and Jefferson) are overwhelmingly common for African Americans, and not white Americans, this may not be common knowledge, and so employers might not have identified racial differences based on these names.

Regardless, further studies are needed to ascertain beyond doubt the presence and extent of racial bias in job employment.


[1] Evergreen College Protests

[2] Implicit Association Tests

[3] Implicit Association Test data

[4] Dovidio. JF, Gaertner. SL. (2000) – “Aversive racism and selection decisions: 1989 and 1999.”

[5] Bertrand. M, Mullainathan. S. (2004) – “Are Emily and Greg More Employable Than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination.”

[6] Johnson. JD, et al (1995) – “Justice is Still Not ColorBlind: Differential Racial Effects of Exposure to Inadmissible Evidence.”

[7] Vanity Affair article with agglomerated links to relevant studies on US police discrimination

[8] Implicit Association Test – Black bias towards other blacks, or whites

[9], [10] Studies that support common ingroup method of reducing implicit bias

[11] Nier. JA. et al. (2001) – “Changing Interracial Evaluations and Behaviour: The Effects of a Common Group Identity.”

[12] Green. AR. et al. (2007) – “Implicit Bias among Physicians and its Prediction of Thrombolysis Decisions for Black and White Patients.”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s