Tiptoeing through the merit minefield: blind recruiting and the search for objectivity


The federal government has no interest in forcing all of its agencies to adopt blind recruitment, despite the popular idea winning strong support within the Liberal Party as a way to ensure hiring decisions are merit-based.

Public Service Minister Kelly O’Dwyer avoided saying directly the government agreed or disagreed with the proposal from the ACT branch of the party, which was “overwhelmingly suppor­ted by grassroots members” in a vote over the weekend, according to The Daily Telegraph.

She also chose not to say whether or not the government was even considering the proposal, suggesting it is quite comfortable with agency heads making up their own minds and has bigger fish to fry right now.

“There is no Australian Public Service policy on blind recruitment — agencies are responsible for their own recruitment,” O’Dywer said in a statement to The Mandarin.

“It is at the discretion of agencies as to how they conduct their recruitment and they may choose to use blind recruitment.”

“There is a legal requirement that employment decisions in the Australian Public Service are based on merit. Engagement and promotion decisions are made using competitive selection processes, with the APS providing a reasonable opportunity to eligible members of the community to apply for employment.”

The idea of blind recruiting — making applications anonymous at the first stage, when candidates are chosen for interviews — has become very popular, but mixed results from trials suggest its effects can be uncertain and highly context-dependent.

Interestingly, the Liberals who promoted the idea assured their fellow delegates they were not in favour of “promoting minorities” but saw the policy as a way of making sure the best candidate was selected every time, according to the newspaper report (party officials failed to provide us with the text of the successful motion).

How did we get here?

This is not the usual reason for trying out blind recruiting. Dr Meraiah Foley, who researches these issues at UNSW Canberra, reminds us it originally emerged as a measure that might, in some circumstances, reduce the impact of cognitive biases that disadvantage some candidates, despite the efforts of selection panels to be objective.

“It’s important to understand that the motivation for introducing anonymised recruitment processes originates in decades of studies showing that women and certain minorities are significantly disadvantaged in the initial stage of recruitment in many contexts,” she said.

Even if this is the intention, it’s far from clear that blind recruitment works that way.

“The research shows the effectiveness of blind recruitment is highly contingent on organisational context, and so it can’t be seen as a universal remedy that can prevent any and all forms of discrimination,” in Foley’s view.

The idea that the best candidate should get the job, no matter what, is persuasive and popular but also simplistic. It relies on an assumption that selection panels looking at resumes and responses to standard criteria are able to make objective assessments in most cases. One only need look at workforce demographics for various organisations or sectors to see why a lot of human resources experts, and senior public servants, suspect that is not the case.

Various trials and studies around the world have not provided strong validation of the standard hypothesis, that more women and people from disadvantaged groups would get jobs though blind recruiting, and that’s probably because the roots of this disadvantage go much deeper than unconscious bias.

The concept of merit itself, and how it is measured, also involves biases and personal values.

“Research shows, for example, that some groups have better access to informal networks and mentors, or are more likely to receive better referee reports,” Foley explained. “So the idea that the concept of merit itself is objective or equal is flawed.”

An alternative to the ‘positive discrimination’ hypothesis

Some evidence suggests blind recruitment can in fact compound the disadvantage experienced by women, people with disabilities and candidates from non-English speaking backgrounds or those whose age is significantly different to other candidates, in the initial recruitment stage.

Foley explains this could be because it takes away the agency of the person reviewing the application, reducing their ability to understand the context for attributes that could detract from the application, such as long gaps in experience or unusual qualifications.

When the Department of the Prime Minister and Cabinet’s behavioural economics team ran its own study, its hypothesis that public servants were already favouring the disadvantaged groups and the emotive term “positive discrimination” caused quite a stir, especially among people who disagree with affirmative action policies.

Many took this as evidence supporting their personal view that being a woman or from a minority group actually had flipped around to become an advantage in the APS. But that might not be true.

The alternative hypothesis, Foley explains, is that “some context-specific information … may be interpreted disadvantageously if the candidate’s identity is unknown” while otherwise, the person’s identity might provide the necessary context to give them a fair run against other candidates from less disadvantaged backgrounds.

Some of Foley’s research, published recently in the highly respected journal Public Administration Review, indicated conflicted views on merit and diversity among about 100 public sector middle managers who were interviewed.

Many “acknowledge the existence of implicit biases and their potential to create unequal employment outcomes” but still, few are comfortable with affirmative action as a response, seeing this as “an unacceptable violation of merit” — apparently in line with the view of many within the Liberal Party.

What does ‘merit’ really mean?

The common claim that diverse workforces perform better is generally accepted among APS leaders. Also, there’s a reasonably compelling argument that a public service workforce, or at least certain parts of it, should aim to reflect the society it serves.

Leaving aside the problems with reducing efforts to counter discrimination or unconscious bias to a way to boost financial or organisational performance, meeting diversity targets and building diverse teams would seem to require at least some consideration of candidates in terms of how they would complement existing team members.

Blind recruiting might make that sort of workforce planning harder.

“I think it also comes down to a wider conversation about how we think about merit,” Foley added.

“The concept of merit in the public sector, of course, was originally intended to prevent nepotism and patronage and political influence … and it has since morphed into a sort of religious embrace of the idea of individual skills and capabilities — mine versus yours — without looking at the merit, for example, of the broader public service being representative of the community it serves … or looking at organisational objectives and how the hiring of a particular team of people might serve the organisation’s objectives better.”

She points out that merit, in terms of the legislation governing public sector jobs, does not actually mandate “the objective measuring of an individual’s skills and capabilities over another” but rather is about opening these jobs to all comers and awarding them on the basis of the applications, not on party politics or someone’s desire to help a friend or relative.

“So to me, it’s about this wider question of whether we believe that it’s possible to achieve a sort of pure objectivity in our recruitment,” Foley said.

She points out that the idea of merit comes up quite often when the goal of increasing workforce diversity and affirmative action policies are discussed, but rarely when one questions the clear dominance of tall Anglo-Saxon men in the most senior and best-paid jobs.

The world of behavioural economics, now down under

According to a PM&C spokesperson, the behavioural economics team (BETA) is “exploring opportunities for further research in this space” following last year’s experiment. At that time, BETA’s inaugural head of research, economics professor Michael Hiscox, told The Canberra Times it would difficult but not impossible to run a follow-up study on real job applications.

“There remains benefit in a follow-up trial to further test de-identification in a real world recruitment process in addition to testing other behaviourally-informed interventions at other stages of recruitment,” explained Hiscox, who will soon return to a role with BETA, and will speak at the Behavioural Exchange conference next week.

“A trial involving real job candidates applying for real jobs may be possible, but it would involve overcoming a range of important legal and ethical hurdles. These are not insurmountable, but they do make such a field experiment challenging.”

BETA is hosting the Behavioural Exchange Conference (BX2018) in Sydney next week. The conference, previously held in Boston at Harvard University, London and Singapore, returns to Australia for its fifth year.

A follow-up event will be held in Canberra, with BETA/PM&C partnering with the Institute for Public Administration Australia. Behavioural Insights: Global Perspectives, on Thursday next week, will feature David Halpern along with OECD, Canadian and French panellists.

This article has been updated with a brief statement from PM&C that was provided after initial publication.

About the author
Premium

The essential resource for effective public sector leaders

Can you afford to miss the next briefing from Mandarin Premium? Sign up today.

Get Premium Today