NSW nudge unit weighs in on 'unconscious bias' concerns

article-body

The New South Wales government’s behavioural insights team has weighed in on the growing concern about unconscious bias in recruitment, with a brief guide designed to support public service workforce diversity strategies.

Released on Friday, it looks at the impact of groupthink, confirmation bias, clique behaviour and other cognitive biases that can subconsciously skew decisions on how government agencies select and promote staff.

The premier’s Behavioural Insights Team also suggests the ways agencies advertise could encourage a narrower field of candidates to apply, in the same way the design of various forms and letters can affect how citizens respond to them:

“Like a pipeline, problems at the start will have flow-on effects throughout the process.”

The concise guidebook explains some of the basic academic literature, and is based mainly on a longer report jointly authored by the United Kingdom’s own nudge unit, on which the NSW team was closely modelled:

“It is not an exhaustive summary and focuses only on behavioural interventions, not structural changes or policies such as flexible working arrangements, return-to-work options, cadetships and targeted positions.”

Like most of the biases covered, the in-crowd effect or “affinity bias” only applies to selection and promotion. But the BIU advises that confirmation bias — the tendency to reinforce one’s existing beliefs — is an example of one that also applies to jobseekers themselves:

“Research suggests that interviewers can take as little as four minutes to decide whether or not they want to hire the candidate. Information not consistent with the first impression can be overlooked, which is why it is important 
to use structured interviews (more details below).

“Similarly, candidates with the preconception that they do not ‘fit in’ to a workplace may subconsciously search for information that confirms their preconceptions, which could affect their decision to apply.”

Others include the availability heuristic, which reminds us we are most influenced by the information we can remember the most easily because “events that are recent, emotionally charged, and/or unusual are the easiest to bring to mind”:

“For example, managers might recall a recent time when an employee performed poorly, despite other examples of excellent performance, and in turn, view them as not suitable for a promotion.

“Similarly, when selecting a candidate, recruitment managers might give more weight to performance on the most recent assessment task (usually the interview) over all the other assessments completed by the candidate.”

One piece of advice, which doesn’t mention the value of strong relationships with existing employees, suggests managers “may unduly value the skills and characteristics of their current staff” and be blinded to the superior attributes of others working elsewhere.

Groupthink — “when members of the group favour harmony and conformity over dissent and deliberation” — can erode the purpose of having selection panels, where a robust debate produces a fairer outcome, and “is particularly strong when there are power differentials” according to another passage.

There’s also advice on the halo effect, the endowment effect, the representativeness heuristic, stereotyping, status quo bias and stereotype threat, and how these all relate to recruitment and promotion.

Some of the solutions proffered by the NSW nudge unit to get a more diverse field of applicants include only putting the most essential selection criteria on job advertisements, making the application process as easy as possible, as well as increasing the visibility of senior staff and using more personalised communications to “target groups”.

The guide also has a long list of suggestions for selection, not all unfamiliar, such as using a structured list of questions in job interviews, reducing the weighting placed on interview performance, and removing various personal details from resumes. One interesting idea is to have the final decision made by someone outside the assessment process.