Introduction
This has been stimulated by recent conversations and is based on the NSW public sector’s current practices. The state has an expressed commitment to employing more people with disability, which I do not doubt is sincere. However, experience doesn’t necessarily match.
There is a widespread belief that bias can be mitigated by brief training efforts – in disability awareness, in inclusion and diversity and in combatting bias. There is no evidence I am aware of that confirms such efforts do any good. Indeed, there is evidence the outcome is contrary to the intent. Brief training doesn’t stick as a rule and must be backed up by revisiting and active reflection on the training. This doesn’t usually happen, so participants in such training can be misled that they have been ‘trained’ in bias mitigation.
Under that error many will assume that this brief cognitive input has made them sufficiently aware of bias to ensure they will not be swayed by it. If only it were that simple!
Effective bias mitigation requires two key things – intentional and conscious awareness in company of at least one other person engaged in the same endeavour to mitigate bias and specific practices designed to reduce the risk of bias.
Here I want to focus on the second factor via 3 elements of the selection process – initial assessment of applications, adjustments at interview and the composition of the assessment panel.
The elimination of disadvantage and bias in recruitment isn’t simple, but a more systemised approach will go a long way to reducing risk. Recruitment is often undertaken by people with limited or no experience in selection processes. It’s common for the manager of the work unit to convene the selection panel and determine its members. This is a huge risk area that could be significantly reduced by ensuring all recruitment is undertaken by recruitment professionals. This is, however, not about to happen, so mitigating the consequent risk must become a priority.
I am not entirely disparaging training efforts. In fact, I do think that effective training in bias mitigation has benefits wider than recruitment. However, we must understand the limits of well-intentioned training and ensure its not the only strategy employed.
Initial assessment of applications
Depending on roles applications may comprise different phases. Research into ‘decision-making hygiene’ provides good arguments for each step being assessed separately, and candidates rated for that step. This approach will produce an aggregate score which will rank applicants.
An advertisement for a Senior Project Officer that I saw recently had only 4 stages – a cover letter (a page), responses to 2 questions (a page per question), the CV, and the interview. Assessment is theoretically based on identified core capabilities from the Capability Framework. There can 4-6 capabilities which must be addressed. However, asking a candidate to demonstrate they meet these capabilities required and assessing them against them is a hit and miss affair.
I question the value of the first 2 steps because the requested excessive brevity seemed to me to diminish the value of what was asked for relative to what is being assessed. There are, I believe fundamental problems in asking for elements in a selection process that can’t be usefully assessed toward ranking an applicant.
This example can hence only be used as a thinking model rather than an instance of a workable approach.
Nevertheless, ideally this situation would lead to 4 assessment steps, each rated only against other candidates’ same submissions. For example, all cover letters would be assessed together and ranked. This would be ideally done with names masked.
The advantage of this approach is that the interview tends to be the determining assessment step. This is where applicants with disability are most vulnerable, and where bias can really kick in. A strong candidate may have 3 high scores from the other steps which could off-set a weaker interview performance.
The disadvantage is that it takes discipline and more time to make this approach work and will not be preferred by busy hiring managers, regardless of their obligation to minimise bias. It would have to be a mandatory approach.
Culling applications
The NSW public sector theoretically uses the Capability Framework to identify usually 4-6 focus capabilities which a role is considered to require. This would be a great idea if the intended rationale was uniformly employed.
It would be sensible to cull applications depending on the quality of demonstrated fit against the focus capabilities. These capabilities are skills, not experiences, so a CV isn’t a fair source – unless the applicant has been advised to write a CV highlighting capabilities. This, so far as I am aware does not happen. Ads for roles do not direct applicants to the capability framework or how to prepare a CV. This might disadvantage external applicants if assessment against capabilities was uniformly practised.
Absent the CV in the example considered here the assessment panel is left with the cover letter and the targeted questions. If these are not designed to elicit responses related to the focus capabilities, there is a serious question to be asked about what logic is employed to cull applications.
This question matters because the default rationale for culling may be the CV. CVs are not easy to interpret. For example, a long-term public sector employee may have no awareness of a private sector role, especially in relation to capabilities. In essence, a CV should not be used to cull applications where the focus is on capability rather than experience.
This matters to an applicant with disability because their disability may have disrupted work experience.
What is needed is a standard focus capability based initial written element of sufficient word length to give an applicant with disability a fair shot at making it to interview.
This could be as basic as saying “Here are the focus capabilities for this role. In no more than xxx words please tell us how you meet them. Your response will be used to assess whether you are one of the xx applicants to progress to interview.” This could replace both the cover letter and the 2 questions. A word limit per capability should be set.
Assessment by capability rather than experience is a good approach in my view, but it must be governed by a strong methodology which breaks the habits of past practices.
Adjustments for interviews
The opportunity to ask for an adjustment at the interview stage is universal these days. But this mostly relates to physical accessibility. Requests for adjustments for other reasons are not yet common. Physical or sensory accessibility requests are usually easy to accommodate. In addition, such disabilities tend to be evident.
There is, however, a wider variety of disabilities which are ‘invisible’, which may not be something a candidate feels comfortable disclosing, or which may not be thought to be a disability.
The standard practice is to give candidates around 20 minutes to preview 4-6 interview questions shortly before the interview. We have been doing this for decades. I can remember when getting the questions in advance was an innovation – and thought be some as going soft on candidates.
It’s a silly practice as it is. At best it’s a quiz under an unnecessary time pressure. It measures nothing useful and significantly disadvantages a wide range of people with disability. This includes people who are neuroatypical, people with anxiety, and people with sensory or motor impairments.
This spectrum of people also includes many who do consider themselves as a person with disability. Some anxiety states are situationally triggered for a variety of reasons. For example, anxiety can be induced by public speaking, singing in public, disclosing personal information in a new group, sitting exams, and job interviews. Impaired performance under a state of anxiety is not indicative of performance in general.
Interview questions serve a purpose in asking candidates to express ideas clearly to show understanding of a role’s dimensions and requirements. They are not intended to be quizzes, as they once were. When we understand that history, we can appreciate that current practices are not only out of date they can be seriously discriminatory.
The solution is to eliminate the pointless hothouse pressure of having a scant 20 minutes to review 4-6 questions and prepare responses. There are 2 further problems with this practice.
Questions should be designed to elicit responses related to the identified capabilities and some questions have 2 elements in them. I reviewed 4 questions from a panel in 2022 and 3 had 2 elements, making a total of 7 questions to be considered in the limited time made available. It was also not clear how the questions related to the capabilities.
I have proposed giving all candidates the questions 3 days in advance. Others, with more experience in contemporary recruitment, think 5 days is better.
I proposed 3 days for all candidates after an applicant was given the questions several days in advance because of their disability. It was the right thing to do. But it also gave them and unintended advantage. The only way to meet their adjustment need and be fair to all is to ensure everyone has the same opportunity. The disadvantage and the unintended advantage are both eliminated. It is a truly inclusive step.
The panel members
I have been an independent on panels that have been skilfully run, and it was a pleasure to have been part of such a high-quality exercise. I have no doubt that well-run panels that reduce the bias risk are real, but they are not as common as we’d like.
As well as being a person with disability and a former DEN Chair I have an extensive background in recruitment. Because of this I feel comfortable taking my role as the ‘independent’ literally. The independent may be the key to ensuring that the bias risk is reduced – if that status is recognised in a more formal way.
The most important change to recruitment practices would be to develop a pool of formally recognised and independent panel members (Inclusion Independents) who are trained in recruitment practice and bias mitigation. Each selection panel would be randomly assigned a pool member who would certify that the recruitment exercise met minimal standards for bias mitigation.
The independent would have standing in the human resources team, so if clear bias was encountered the recruitment could be suspended until a solution was developed.
Identified roles
This is a vexed area that stirs strong emotions that are vented in advocacy, but rarely in opposition. Arguments against are rarely plainly stated or defended. It merits a more substantial consideration than I can accommodate here.
I will simply observe that in sympathy with the principle of ‘nothing about us without us’, where roles concern disability they should be filled only by people with the experience of living with disability – unless specialised qualifications are required.
The theme of identified roles attracts a lot of heat and little honest debate. It is an argument that must be had. But when it comes to roles directly related to disability, I think there is no legitimate argument to be had.
Conclusion
Eradication of bias is a nice target, but likely unattainable. But very significant reduction is realistic. I have addressed only 3 elements of the recruitment process to demonstrate that bias reduction isn’t a serious challenge in itself – so long as a disciplined methodology is developed and implemented. But there are challenges in establishing disciplined uniform methods. These require a commitment by senior management to ensure bias mitigation methods are consistently implemented.
Effective bias mitigation in recruitment benefits everyone, not just people with disability.
I have had over 6 years’ experience in recruitment as a job and many times since as a member of selection panels. My success rate as an applicant is about 50:50.
Effective recruitment is difficult. It is a far more skilled affair than most appreciate. Sometimes the choice is a standout no brainer. But it often comes down to finding reasons to exclude an applicant on a fine point. After all you can offer a role to only one person.
You can exclude applicants through the initial coarse sieve of culling. Here obvious acts of discrimination can eliminate an applicant on grounds of gender, race or assumed religion – to mention a few common excluders.
The finer sieve of assessment through interview is where applicants with disability are at greatest risk. This is where the strongest protections must be located. The most potent is, I believe, the genuine accredited independent. That said, this isn’t a substitute for other bias mitigation efforts, just a bulwark against their ineffectuality until things substantially evolve.