Unconscious bias and its impact on the gender salary gap

Biases can often be unconscious. Image sourced from www.shutterstock.com

Chief executives who don’t care about the gender pay gap or believe that closing it will only increase costs, might want to consider some of the benefits of audits and strategies for identifying and closing the gap.

The recent “Daughter Water” campaign launched by the Workplace Gender Equality Agency was a light-hearted attempt to highlight that the male-female wage gap was lower in organisations where the CEO has a daughter.

The WGEA is trying to promote more strident auditing processes by focusing the campaign on the 3000 or so Australian companies whose annual reports to the agency indicate that they do not conduct gender pay audits.

Of course equitable treatment of men and women is valuable in its own right, but if that isn’t convincing enough, research studies have shown that perceived inequity in pay is related to lower engagement, higher turnover and greater cynicism.

Cynicism, disengagement and a willingness to leave are probably greater for women in organisations that champion values of transparency, collaboration and merit, but where actual pay gaps are evidence that their human resource decision processes do not reflect the company’s values.

There are many contributing factors to the gender pay gap, with unconscious bias believed to be a major cause. Of course unconscious bias is not limited to human resource decisions; it leads to suboptimal decisions in finance, marketing and operational processes.

What is unconscious bias?

When we say judgements are unconscious, biased or otherwise, it means they are fast, we are not consciously aware of making them and we are not aware of how they influence our decisions, feelings or behaviour. We do not directly control them. Bias is not removed by good intentions.

Unconscious bias in decisions is also not, necessarily, evidence that a person is sexist. For some people, being confronted with evidence of unconscious bias challenges their self-beliefs of being “a fair minded person”, “not sexist” or “rational”, which can lead to defensiveness. Unconscious bias is the product of life experiences, including the societies we grow up in, and as I will explain shortly, can be increased or diminished by the processes used to make decisions.


Salary decisions about men and women can be affected by several different types of bias. First is stereotyping. When it comes to evaluating men and women at work, men have a much more positive stereotype than women. Society has taught us to associate valued attributes such as leadership, decisiveness and success more strongly with men than with women.

Stereotypes have subtle effects when we are evaluating other people’s work. For example, when evaluating job performance we are more stringent and more challenging in our evaluations of the successes of women than the same successes of men. We are more likely to question how much the success was due to them.

Men who are not performing well are more likely to get a break, such as “it was a tough market” or “there were a lot problems with the production system that quarter”. Alternative explanations for poor performance are less likely to be raised for women.

“People like us”

Secondly, affinity bias refers to the tendency to be more attracted to and more supportive of people who are like us and with whom we have a shared understanding. Leaders have “go-to people” they are comfortable dealing with. The mutual trust and understanding, responsiveness and support from a “go-to” person will often manifest as advocacy in the leaders ratings of performance discussions and salary recommendations.


A third type of bias is anchoring. Once we have a baseline, in this case the existing salary of a woman, recommendations for increases are strongly anchored or constrained by the existing salary. The use of percentage increases is an example of anchoring. However, if women are systematically underpaid relative to men, anchoring will perpetuate the pay gap or cause it to close very slowly under existing salary decision processes. Men and women should both be anchored on the higher salary and plans put in place to close the gap over time.

Confirmatory bias

Two other biases come into pay during the process of evaluation and salary recommendations. The first is confirmatory bias. Once judgements or recommendations are made, people are highly motivated to find or produce the evidence that support the initial judgement.

Once recommendations are made, any attempts to minimise bias will be much more difficult. Leaders and their colleagues will bolster their decisions believing that they are providing rational justification, but in many cases they and their colleagues are protecting their belief that they are a “fair minded person”.

Shifting standards

Finally, there is the problem of shifting standards. When committees discuss a number of staff, it is not uncommon for the standards applied to the person who is discussed first, to be different to the standards applied to the person who is discussed, say fifth or sixth. This is, in part, due to the other biases.

Even when their judgements are influenced by stereotyping, affinity, anchoring and confirmatory biases, people will want to be rational. So when confronted with hard evidence of success by women people unconsciously raise the bar and shift the standards applied to men and women throughout the meeting.

What can be done?

Audits, as recommended by the WGEA, are the best place to start. It is harder to motivate people to embrace change without some clearly identified need and a goal. An audit will also help to pinpoint where the gaps are, if they exist, and provide guidance on where to target strategies.

Once gaps are identified, systems and processes that generate pay outcomes, including appraisals, moderation committees and salary reviews, should be reviewed for potential biases of the kind described above.

Redesign of the systems and processes to minimise bias will usually include improvements in the analyses and organisation of the data considered and discussions of the data plus the development of clear guidelines for the chair and members of any committees involved in the process.

Want to carry out an audit? Use the WGEA tool, found in Resources and Downloads.

Read the other pieces in our Gender equality at work series here.

Found this article useful? A tax-deductible gift of $30/month helps deliver knowledge-based, ethical journalism.