People say “seeing is believing”, but that’s wrong. The truth is, “I will see it when I believe it”.
As an academic psychologist I have spent years, and run dozens of experiments, looking at unconscious or implicit bias and its consequences. I consider factors such as looks, ethnicity, age and gender, to see if they influence world-of-work decisions such as hiring, promotion, salary.
The short answer is that all these factors make a difference, even though they play no real role in the evaluated person’s performance. Beliefs guide the facts we see. They shouldn’t, it’s unfair. But they do. The so-called ‘Beauty Premium’ is real, as are a host of other biases.
Taking decisions this way is not unnatural. Evolution has fashioned us to infer, to fill in knowledge gaps. Is that rustle in the grass the wind, or a snake? Assume, infer, and take the conservative decision. That’s how we survive.
But using inference or stereotypes to guide staffing decisions is not effective because the right candidate may be overlooked and the ‘right-looking’ but wrong candidate selected.
The point is we are very quick to size people up – age, sex, appearance, even height. We fill in the blanks and give them a price tag in a stereotypically consistent way. The problem is that once we decide about something we try to justify it because we don’t like to admit we were wrong.
One study I know asked people to vote on the basis of photos, as if they showed candidates running for public office. Afterwards, the voters were given information about the ‘candidates’ (e.g., political preferences, values, etc.) and then asked to vote again. Despite now having relevant information the voters hardly changed their opinions.
I thought this might be due to past experience – perhaps people have a learned stereotype of what a ‘Leader’ should look like? So I repeated the experiment with small children, too young to have learned bias, showing them pairs of photos and asking who would make the best captain of a boat (a position of responsibility they could understand). I asked some adults to do the same test. The children and the adults chose the same photos. No experiential factor could explain the choices, it had to be nature.
But, perhaps the motivation or education level of the testers played a role? So I did a similar experiment with kids using photos of candidates for elected positions at the Association of Psychological Science (APS). All the voters and candidates were scientific psychologists. But results were the same. When no photo was available in the original ballot material the APS members voted on the basis of publication record (a reasonably good proxy for the knowledge, status, and success of the candidates). However, when there had been photos included in the ballot materials nothing mattered but the face.
Maybe business people would take decisions in a more rational way? So, we asked experimental subjects to look at photos of managers in a large multinational company, and then asked them to judge the managers for competence and personality. We accounted statistically for everything possible – age, qualifications, and so forth. Those managers who rated higher on looks earned more.
Implicit bias is even worse for women. Factors such as being overweight count against women even more than they do for men. And it’s not just appearance. I worked with a Swiss multinational looking at the transcripts of their internal performance evaluations, and statistically controlled for everything possible. Men had a much higher likelihood of being described in a positive way; for example, “he really knows how to put his foot down” compared to a similar woman, who “really knows how to use her elbows”.
Age discrimination was also rife across the board, even though for high-level, cognitively complex jobs there is zero correlation between performance and age. In short, age and being male predicted future job and salary levels.
So women (and anyone else who does not fit role expectations) are walking on eggs. It’s a double bind. They must demonstrate exceptional competence to be seen as equal in ability to men, but must also avoid threatening them with competence and apparent lack of warmth, or behaviour that violates social stereotypes.
An experiment run by a professor at Yale University demonstrated the penalty for violating these social norms. One male and one female actor were each asked to record two versions of the same interview, one where they were calm and one showing some anger. Their answers were the same so rationally, the man and woman should have been ranked the same in the same condition. But it turns out that if a man shows anger it is interpreted completely differently. Men can show their “guts.” Women are not allowed to show anger because they are supposed to be nice, nurturing and kind. When subjects were asked to rank the two actors, the man was seen as higher status and more competent, and offered 50 per cent higher salary. The woman was seen as out of control.
There are ways to reduce bias in the workplace. The first is to be aware of your own biases. Then you can take steps to eliminate them and so reduce discrimination.
Second is accountability. Decisions need to be justified, with objective indicators. Be aware that every piece of information can introduce bias. How the call for applications is made – certain words will attract or discourage women. What information applicants are asked for, including photos, can matter. Who does the initial screening, and is it objective or just personal opinion? Are the screeners different from the interview panel? Are the same interview questions asked of all candidates and is the information aggregated independently? Are validated psychometric tests used (e.g., the most used test in the business world, the MBTI, is actually useless; it has no predictive validity).
Data is also key, it allows us to track what is happening, reveals unconscious bias and creates awareness.
Finally – men. We are part of the problem but also part of the solution. If we champion the cause we can reduce these biases. This is our problem too, not just a problem for women or minorities. Taking decisions correctly is not only the ethical thing to do, in the long run it is the economical and rational thing to do.
Professor Antonakis spoke at an event at the ILO organised by the Enterprise Department’s Gender Focal Point Team.