Diversity: Employers must close the gap between intent and action

New research from the Institute of Student Employers (ISE) reports a 12% reduction in the number of graduate vacancies. Fewer vacancies mean application numbers will rise due to increasing levels of unemployment paired with less opportunity. As applicant volumes spike, there’s a temptation for recruiters to resort to blunt screening instruments to reduce the number of candidates in the funnel.

Quick fix tactics such as CV screening or narrowly focused psychometric tests may reduce the number of applications, but they also risk amplifying inequality and damaging diversity efforts. In the bigger scheme of things, taking shortcuts at this stage will further entrench bias into organisational fabric.

Although most businesses didn’t plan to transition to online recruitment so quickly this year, new technology and processes have created more resilient models that can be flexed to any scenario. So, it’s an ideal time to check your graduate recruitment process is fair for all.

Graduates and diverse hiring

The ISE Student Recruitment 2020 report certainly demonstrates commitment to improving diversity, especially in terms of attracting diverse talent and analysing data to monitor diversity.

Amongst the ISE respondents 62% of organisations have formal targets around diversity and many are taking practical action to attract a broader talent population. Tactics include targeting particular groups (58%), changing the universities that they visited (46%) and tailoring their marketing materials (45%).

Engaging diverse talent and getting a broad range of people energised about your company is the first step to improving diversity. However, this great start risks being undermined if the structure of the recruitment process isn’t addressed in tandem with candidate attraction.

For example, do you find you’re filling the funnel with high potential candidates in the early stages but seeing few of them onboarded or progressing in your organisation? If so, you may need to review your recruitment process design.

Intent and action

The ISE report also shows that it’s less common to make structural changes to organisations’ recruitment approaches. Only 19% have removed some pre-entry criteria, for example degree classification, and just 11% have simplified or reordered their selection process. The gap between good intention and action means diversity may be compromised.

The data also shows that 83% of employers set some minimum requirements with the most popular being academic qualifications – a criterion which we know favours candidates from more privileged economic groups.

46% of employers say that using psychometrics is one of the most effective ways to ‘narrow the field’. Psychometrics can certainly yield meaningful results, however, if narrowly focused tests such as numerical reasoning are used up front as a screening tactic, they can allow bias into the process. Because verbal and numerical reasoning are linked to education opportunity, those from more disadvantaged backgrounds will therefore be less well-prepared.

With most of the focus on attraction, and less on structural and process change, graduate recruiters’ good intentions around diversity could be compromised by systematic issues in the recruitment process.

Fairness by design

Sova helps clients build fairness into the design of their hiring processes from the outset. Our new white paper Levelling the Playing Field walks through the process in detail, but here are some of our tips for creating a process that is fair by design.

1. Focus on building a profile of ‘what good looks like’ that does not rely on the perspective of the current majority group, but rather is inclusive. Think carefully about what you wish for and define it clearly and fairly whilst avoiding stereotypes.
2. Collect the right bio data from the outset. This includes inherent characteristics such as gender, age, ethnicity, language, socio-economic background. Be clear to candidates about how and why this data will be used.
3. During design and testing, include a diverse population in order to generate representative data. Actively source diverse candidates to participate to ensure the assessment is interpreted fairly.
4. Identify sources of adverse impact and identify the weak points in the process. Machine learning can flag these areas quickly. The assessment process can be optimised and refreshed on an ongoing basis.
5. Monitor outcomes in real time. Ongoing monitoring, analysis and calibration ensure bias is closed down and unfair decisions at the point of hiring are not transposed into progression.

This article was first published by the Institute of Student Employers (ISE) on the 30th November 2020