November 05 2018

No quick fixes for fair recruitment

Joy Warmington is Chief Executive of BRAP and holds an honorary doctorate from Middlesex University. Roger Kline is research Fellow at Middlesex University Business School. Here they highlight some of the issues with recruitment processes that are preventing organisations from becoming more diverse.

With Ministers announcing targets for ensuring diversity amongst public sector leaderships, organisations are increasingly asking how can they diversify their workforce? A growing number of organisations are insisting on diverse panels as one means of doing so.

We know current recruitment practices repeatedly favour some groups of staff at every stage of recruitment, development and promotion leading to a loss of talent. We know that the evidence shows that diverse workforces and leaderships that are inclusive are likely to be more productive, creative, innovative and engaged. We now know that the reasons women, BME and disabled staff fail to get fair treatment in recruitment is because of multiple forms of bias that influence appointment panels to choose people who ‘fit in” or are “like us” so that too often individuals are hired because of who they are – rather than what they can do.

Understandably, the focus for helping change this lands on the recruitment process. How we shift this pattern depends on two simple things:

  • what we believe are the reasons behind un-diverse appointments?
  • what steps are we prepared to take to challenge the status quo?

The lack of diversity in appointment processes, has been, and often still is seen as the fault of the applicant. There are a range of measures employed to ‘fix’ this – including widening advertisement processes so that it is likely to attract more marginalised communities, using a range of innovative means (open days, tasters, shadowing opportunities) to help demystify roles and encourage applications, and getting further down-stream by reaching into schools to promote opportunities. This “deficit” model may be helpful but doesn’t tackle the core of the problem.

Unconscious bias in interviews

Interviewing has historically been seen as an essentially fair process, but research is beginning to recognise its inherent faults. In our attempt to address these faults, organisations have spent more time ‘tweaking’ the interview process than we have recognising that people work within a system that replicates unfairness, and that they too become part of this system.

Unconscious bias training is one step that organisations often take to address unfairness in recruitment processes. Fundamentally, it can be right to point out that all of us have biases, and that we actively replicate these in our lives – including as part of the recruitment process. This type of training is very varied – it ranges from the application of the implicit bias test (which analyses bias through an algorithm), power point workshops, through to more active training and development opportunities. As we have pointed out previously, and as the recent EHRC review shows, although some experiences are no doubt better than others, all come with the health warning that understanding our adverse biases doesn’t mean that we are capable or indeed willing to change them.

The ‘diverse’ panel

Another increasingly popular strategy is the inclusion of a ‘diverse’ individual as part of the interview process. This has become increasingly common yet the evidence base for it is pretty thin. The impact of ensuring selection panels include women, for example, is mixed. Some studies show that as the numbers of women on a panel increase, the more likely it is that women will be selected but other researchers have found the opposite. One study found that when a woman was the only female member of a high-prestige work group and was asked to vote on another candidate for the group, she is much more likely to choose a male candidate than a female one. In summary the evidence is mixed.

That does not mean diverse panels are a bad idea. Intuitively they can be a positive step, but its significance can easily be over-stated especially when done in isolation. We have unfortunately seen some organisations make this the cornerstone of their approach, yet it can easily risk becoming tokenism.

In one organisation BME staff were invited to be panel members but not be part of the shortlisting process.

In another organisation the main role for BME panel members was to ask “the equality question.”

In a third organisation, BME staff who were significantly more junior than panel members were invited to join panels, but without equal authority on the panel that would make up for their more junior status.

In a fourth organisation BME staff were mandated to join panels without more than token training and irrespective of whether they wanted to join the panel or felt able to contribute substantially.

All these approaches (and there are variants including one where the only panel members required to have unconscious bias training were the additional ones with protected characteristics) have at their core the idea that the responsibility for recruiting diversely is substantially remedied by the inclusion of someone who is diverse.

Interview processes are inherently flawed – even before you include someone in them who is more diverse. There is a tendency for those who interview not to have received any proper training on the specific ways in which bias creeps into the best-intentioned interview and their role on the panel is simply based on their position – rather than their skills in choosing good candidates. Furthermore there is no point in doing this training unless it is put into practice as part of the interview process. How many processes discuss and review biases and their decision making – and recognise this as part of the journey to a fairer decision?

This shortcoming is compounded by the seniority of the line manager who generally chairs and who can ‘trump’ other panel members. If the appointment is specialist in nature, then again, the final say on the appointment may well rest with the ‘specialist’.

There are all sorts of specific ways in which bias can be mitigated within the appointment process – from how the job is described, where it is advertised, what the “essential criteria” are, how shortlisting is done, how the core competencies and behaviours required are tested, and including the interview itself. Without serious attention to these, an additional “diverse” panel member will not make a serious difference.

Setting an expectation

At the heart of successfully building diversity into recruitment processes, including interviews, is accountability. When departments and professions are held to account over patterns of recruitment which show it is much more likely that men will be appointed, or white applicants will be appointed, or staff with disabilities rarely get appointed, then a “comply or explain” challenge – explain the data or change the outcomes, does work. That does not mean telling individual panels which individual to appoint. But it does mean setting an expectation – a performance indicator that says that irrespective of their backgrounds and characteristics, once shortlisted there should be no radical differences between the likelihood of different groups of staff being appointed. There is evidence that targets linked to accountability do work.

In an imperfect system run by imperfect people our willingness to recognise our “faults” can bring us closer to realising the opportunities that often lie right under our nose. Let’s think more critically about the whole process of recruitment rather than just trying to put in quick fixes that have limited fixing ability. And lastly, let’s be clear about the expectation – boards should be concerned about shoddy or unfair decision making – poor recruitment patterns are not ‘accidental’, they replicate the status quo.

Share this post

 

Tags: , , , , , , , , , , , , , ,

Leave a Reply

Latest tweets from Middlesex Minds

Follow MDXminds

MDX Uni Research @MDXminds

Delighted at the announcement of our academic's @ErminiaColucci's new research. https://t.co/AfWVyDbYIE

MDX Uni Research @MDXminds

Last Sunday, 800 candidates stood for 111 seats in Iraqi Kurdistan's parliamentary election. Both major parties say… https://t.co/ffyyp77BPc

MDX Uni Research @MDXminds

Published MDX research reveals that #Airbnb distribution depends on who lives there, not distance to city centre. https://t.co/4poazYay1K

Related blogs