Biased decision-making has real-world oppressive consequences 

Not only does bias affect things like hiring practices, promotions, remuneration and access to opportunities at work, but it’s also baked into everything from our healthcare and the justice systems to the built environment and technology.

One of the reasons for this is that biased decision-making gets baked into data, which is then baked into the world around us. This is something addressed by Caroline Criado Perez’s book “Invisible Women: Data Bias in a World Designed for Men” and the work of MIT media researcher Joy Buolamwini’s work on gender and race algorithmic bias. Both explore the impact that bias has on data and how throughout history marginalised groups have been excluded from huge sets of data leading to bias being embedded into the world around us. 

Here are some real-world examples of data bias:

  • Women in Britain are 50% more likely to be misdiagnosed following a heart attack because heart failure trials are generally conducted on men.
  • An algorithm used widely in US hospitals was less likely to refer Black people than white people who were equally sick to programmes that aim to improve care for patients with complex medical needs.
  • Cars are designed around the body of “Reference Man”, so although men are more likely to crash, women involved in collisions are nearly 50% more likely to be seriously hurt. 

Recommended resource: 

Start with unlearning your own bias

The good thing is it is possible to unpick and unlearn our biases by understanding what they are. The moment you become conscious of and understand your own biases is the moment you can actively begin to challenge and overcome them.

Here is a three-step process to help you to understand and mitigate your own bias: 

  1. Understand your privilege and biases – understand what privilege and power you might have walking into any given situation, take a bias test and spend some time reflecting on where you can improve
  2. Take active steps to unlearn bias – educate yourself about your own biases, pick up a book or watch a documentary (see the recommended reads and watches below). Write down your thoughts and commitments unlearning bias. 
  3. Repeat steps 1 and 2 – unlearning biases is a continual process, as dialogues around topics evolve and shift over time. Also, refreshers are useful for everyone, especially when unpacking complicated social issues, as we forget.

Recommended reads:

Ways to alleviate bias:

However, becoming aware of our biases is only the first step. At work, we have to then embed policies and practices that actively recognise bias and share power. This might look like gender and ethnicity pay equity, maternity and paternity leave or flexible working arrangements.

Below are two real examples of power-sharing techniques that helped to alleviate bias. 

Amplification in the White House 

In her piece for the Washington Post, Juliet Eilperin explained how female aides in Obama’s administration ensured that the women in the administration had their voices heard and their ideas acknowledged. In Obama’s administration, most top aides were men, leaving many of the female staffers feeling sidelined in important meetings. To combat this, they used a strategy called “amplification.” When a woman made an important point, other women would repeat it, ensuring credit was given where due and preventing men from taking credit for their ideas.

Redistribution of decision-making power 

US-based social impact accelerator Village Capital seeks to mitigate bias in the funding process by giving entrepreneurs within its cohorts the power to decide which ideas should receive investment. This approach has proven to be a stronger predictor of future revenue and capital raised than conventional impact investment funds where a small group of people decide which companies receive funding. It also has lower gender and race bias and greater transparency than traditional models, with 42% of Village Capital’s ventures run by female founders and 29% having founders of colour.