Government behavioural economics ‘nudge unit’ needs a shove in a new direction

Andrew Frain and Randal Tame in The Conversation: “The Behavioural Economics Team of the Australian Government (BETA) or “nudge unit”, founded in 2015, is using behavioural economics in an effort to improve policy outcomes. The problem? Evidence shows it may be the wrong way to address major problems like inequality.

Put simply, behavioural economics is severely limited in its approach to inequality. Fortunately, other psychological approaches are better suited.

Behavioural economics is built on a particular tradition in psychology, sometimes called the American tradition. At its heart is a distinction between rational and irrational psychological processes.

These are often described in terms of two separate cognitive systems. One is a “slow” deliberate system where logic and reasoning prevails, and another is a “fast” automatic system where stereotypes and unconscious biases hold sway.

Behavioural economics assumes that perceptions of groups (e.g. races, genders and nationalities) are driven by irrationality and that we should stop grouping people by stereotypes or labels. Rather, we should view them as individuals.

However, this rules out important inequality-busting techniques like collective protest, quotas, and affirmative action (favouring those who are marginalised in society). All of these rely on perceiving people as members of a group rather than individuals….

Inconvenient results

BETA recently published the findings of its Australian Public Service study into blind recruitment.

In that study, gender and ethnicity information was removed from descriptions of potential job candidates. It was a study designed to interrupt unconscious biases against women and ethnic minorities.

The results were surprising – blind recruitment made things worse for women and members of ethnic minorities. These results illustrate the limits of behavioural economics in action.

In the study, Australian Public Service managers participated in a hypothetical recruitment as selectors. Converse to expectations, when the gender of candidates was unknown (i.e. blind recruitment), the likelihood of being shortlisted decreased for women and increased for men. Indigenous women, in particular, were less likely to be shortlisted.

BETA interprets this as evidence of “subtle affirmative action taking place among reviewers”.

Here lies the challenge. On the one hand, the goal of de-identification was to eliminate the role of unconscious biases in recruitment, removing the influence of characteristics not relevant to potential performance on the job.

On the other hand, BETA tacitly accepts the identified affirmative action for women and ethnic minorities.

This is inconsistent. BETA is left advocating for blind recruitment to mitigate unconscious biases, but not when those biases lead to the outcomes they want. This is the trap of behavioural economics….(More)”.