By
Impactpool
A negative cycle of underrepresentation
Organization leaders want to hire more diverse talent and tap into underrepresented groups. Donors want the organizations they work with to have fair staff representation.
And ultimately, the communities you help benefit most from support that’s grounded in diverse, representative perspectives. That’s the goal.
But there’s often a wide gap between aspiration and reality. Diversity goals are missed; underrepresentation continues.
This becomes a negative cycle, because low diversity and representation impacts organizational culture.
A non-inclusive working environment then drives-down morale, hurts productivity, and ultimately drives-up turnover, particularly among diverse and underrepresented staff. Accenture finds that the lack of an inclusive culture costs American companies $1.05 trillion annually, for example.
It’s also then harder to attract diverse and underrepresented staff into the organization, so the problem becomes increasingly hard to solve.
All this places a huge burden on humanitarian recruitment teams. Sourcing, engaging, hiring, and then retaining more diverse talent is mission-critical.
But an unfair, biased recruitment process is often a major culprit holding international organizations back. (Not to mention increasing regulatory and reputational risk.)
The unconscious bias burden
The Harvard School of Public Health define unconscious bias as “snap judgements we make about people and situations based upon years of subconscious socialization.”
They emphasize that unconscious bias is implicit. That is, “not deliberately or consciously created, but […] products of our brain’s definition of normal, acceptable or positive.” That definition is “shaped by many factors — from past experiences to our cultural environment to the influence of our social community and media.”
Here are some of the most common unconscious biases recruiters can perpetuate:
Affinity bias – where we unconsciously show preference for people similar to ourselves. For example, that might look like hiring someone who shares the same race, gender, age, or career history as yourself, even when someone else was a better fit.
Confirmation bias – where we unconsciously seek information to support our pre-existing views and reject contradictory information. That might look like disregarding a non-native speaker’s CV on the basis of a spelling mistake, because we assume they aren’t fluent.
Cultural bias – where we unconsciously stereotype individuals based on their country of origin, religion, or ethnic background. That might look like avoiding progressing a skilled candidate based on assumptions about their nationality.
Gender bias – where we unconsciously stereotype based on individuals’ gender. That might look like prioritising a male candidate for a role you believe needs stereotypically masculine qualities.
For example, one study looked at 18,000 evaluations of leadership skills and found no gender differences in objective measures but large differences in subjective evaluations. The specific words used to describe male and female behaviour varied dramatically. These are important considerations given that globally, UN Women say women hold just 28.2% of management positions.
Halo and horns biases – where we unconsciously form a complete view of someone based on a single positive or negative characteristic. That might look like rejecting a skilled applicant because they irritated you by emailing a CV directly rather than applying to your advert, for instance.
Tackling top-funnel sourcing bias
Unconscious bias happens anywhere humans are involved with making decisions. You can mitigate that in various ways, looking across your end-to-end recruitment function through a people, process, and technology lens.
How do you attract? Are you attracting candidates from diverse backgrounds? Or are you demonstrating implicit biases in how, and where, you advertise? For instance, are your job ads indecipherable to anyone outside the UN?
How do you source? Are you sourcing based on objective role requirements? Or are you swayed by subjective biases, like a profile picture?
How do you assess? Are you screening based on objective requirements? Do you have structured, consistent evaluation methods? Are your interview panels inclusive and representative?
How do you onboard? Do all candidates have an equal opportunity to succeed and thrive? Do you make reasonable accommodations to that end?
The whole recruitment funnel matters – but the unconscious bias conversation must start at the top. If you’re not sourcing diverse talent upstream, you won’t hire diverse talent downstream.
AI can play an important role in improving sourcing fairness, helping fill your funnel with diverse talent from the outset. But AI isn’t a clear open goal for recruiters. There are important considerations around how we use AI, and how much control we cede.
An ongoing debate: is AI good or bad for bias?
There’s much debate about the relationship between AI and bias in recruitment. As McKinsey sums up: “AI can help reduce bias, but it can also bake in and scale bias”.
AI models are only as good as the data they’re trained on – and that data is often rife with biases. The risk is that AI perpetuates the biases it learned from, to exacerbate rather than solve the problem.
But AI also has enormous potential to mitigate unconscious bias, by alleviating pressure on human decision-making.
(As one example, we recently tested a human-written versus AI-written job description for a senior position at the UN – and saw AI performing excellently.)
Even with trained recruitment panels, standardised processes, quotas, and so on, humans are inherently inconsistent and subjective. Plus human decisions are hard to authentically review because unconscious biases are, by nature, unconscious. So it’s challenging to get your arms around the problem.
On the other hand, AI has the potential to bring consistency, objectivity, and transparency. All qualities that contribute to a fairer, more inclusive process.
But that’s only the case if AI is used sensitively and ethically.
That’s been a major priority for Impactpool in developing our own AI model.
How Impactpool uses AI to mitigate unconscious bias and improve fairness
Impactpool’s talent ecosystem connects mission-driven organizations with specialized talent for high-impact roles. Advanced AI-matching is a core part of how we achieve that.
We help impact sector organizations reduce unconscious bias and build fairer processes in three main ways:
1 – Reducing sourcing bias
We’ve built an engaged community of 900,000+ impact-sector specialist professionals across 195 countries—54% of whom are women. With our powerful AI screening console, you can search this entire database using nuanced, granular, sector-specific filters.
Read more: Using Affirmative Action to achieve gender parity
Plus, you can easily add your own candidates – either through our API integrations or by migrating to our ATS – so you can search them in the same way too.
For example, international organizations often need to screen for nationality, university degree, languages, gender, past work location, and years of experience.
You might need to source Afghan women with 2-3 years of experience, fluent English and Dari or Pashtu for a gender-based violence position in Afghanistan, say.
Impactpool’s screening filters allow billions of different filtering combinations, so you can get straight to those needles in the haystack. Then contact them straight from Impactpool and move them through your process.
Critically, we’ve limited our AI screening to hard skills and facts, to avoid subjectivity. We also avoid any sort of applicant ranking, which is a common entry-point for bias in AI models. Our AI doesn’t make value judgements. Rather, it surfaces all possible talent against objective criteria.
Where other AI recruitment models tend to be black-box and obscure, our AI is totally transparent around hard skills requirements. It’s a fact-based, data-driven way to source and screen candidates. To ensure you’re filling your funnel with diverse, representative talent from the top-down.
2 – Mitigating attraction bias
Our AI-matching also works the other way around, connecting the right people to the right jobs. Using Impactpool you can create and post jobs easily, and we’ll promote them to the right audiences based on granular sector-specific criteria. (Read more about our Booster campaigns).
We’ve spent years training our AI algorithms using over 1M job descriptions from thousands of impact organizations, so it’s the most advanced tooling you’ll find.
This helps ensure you attract diverse, representative talent to your jobs – a stark contrast from traditional generic tools. How many Afghan women with 2-3 years of experience, fluent English and Dari or Pashtu are on generic job boards and professional networks?
If you aren’t advertising in the right places, it’s little wonder you’re struggling to source people with the right specialized skills, expertise and backgrounds. It becomes its own type of confirmation bias, perpetuating the myth that “we can’t hire more diverse talent, because there’s a limited diverse talent pool”.
3 – Empowering the visibility to drive change
The third way our talent ecosystem helps improve diversity and representation is through robust analytics across your talent database. With Impactpool you get quick reports and easy-to-understand analytics segmented by the fact-based criteria we’ve been talking about.
It’s a far cry from how many humanitarian organizations operate right now. Manual spreadsheets and outdated data exact an enormous administrative burden – for little clarity in return.
Instead, Impactpool gives you a simple bird’s eye view of your talent demographics, so you can identify strategic focus areas. For example, you could see instantly that you’ve got a shortfall of water and sanitation specialists in the Middle East.
Insights like this empower more proactive recruitment – so you can get ahead of your hiring requirements and make progress towards your diversity goals. This long-term vision is essential to improving diversity and representation.
Read more: how to create an effective diversity Action Plan
Improve diversity with Impactpool
Underrepresentation and slow diversity progress are major challenges for most mission-based organizations.
Impactpool helps you tackle those issues, to reduce unconscious bias and build a fairer, more objective recruitment process. To help you connect and engage with the specialized global world-changers who drive your organization.
Start a free trial instantly to see for yourself.
Impactpool is a talent ecosystem for the global impact sector. We help mission-driven organizations attract, source, engage, hire, and deploy sector-specific talent faster. Impactpool gives you everything you need to build teams for the world’s most critical challenges.