Managers of businesses have become more aware of the potential for workplace bias due to the Starbucks incident back in April. This prompted Starbucks to close 8,000 of their stores to address an underlying bias issue, which then caused other companies to reevaluate how they solve major bias issues in their own workplaces.
A common approach many firms take is called diversity training – programs devoted to increasing diversity and reducing bias through employee education. These initiatives are generally well-intentioned, and in high profile cases such as Starbucks, can serve to raise awareness for very important issues. There’s only one problem with them: They don’t work.
When it comes to influencing how people act toward one another in the workplace, behavioral science gives us a general rule of thumb: Information doesn’t change behavior. Restaurants may include calorie labels on their menus, yet people eat even more. Despite the millions annually spent on financial literacy programs, people are no better at saving. And sadly, when firms “teach diversity” through corporate education programs, people show no change in attitude, let alone in action.
The cognitive biases our unconscious brains use to make automatic decisions and judgments on our behalf are just too deeply ingrained.
Instead, we humans are much more influenced by our environment, including the programs and policies we use to guide and implement our workplace decisions. To cultivate a culture that embraces diversity, don’t bother appealing to reason or attempting to change implicit bias. Focus on de-biasing the system instead.
Don’t bother appealing to reason or attempting to change implicit bias. Focus on de-biasing the system instead.
With that in mind, here are five behaviorally informed solutions that managers can implement to address workplace bias:
1. Blind yourself — As a hiring manager, check your own biases at the front door, with behaviorally informed hiring policies. Organizations can learn from orchestras, who use blind auditions to eliminate gender bias. By shifting auditions from the open stage to behind a curtain, the orchestra world welcomed a dramatic increase in female musicians, from less than 5% in 1970, to over 40% today. Before reviewing resumes, try removing the names on them. There are several tools available today that will do that. Several go further and remove other elements that could give away gender or ethnicity.
2. Give your gut a break — When it comes to hiring decisions, structure is crucial. To reduce bias in evaluating candidates, start by identifying necessary skills, and an objective scale for each. Immediately after each candidate interview, rate them on each of those skills. When it’s time to make your decision, compare the candidates “horizontally,” one trait at a time, down the list, and make your decision based on an objective overall score. Another way to reduce gut feel bias is to give the candidates a sample work task to complete as a test. Whatever you do, stop going with your gut! Free-flowing chats are useless for evaluation, and unfairly favor memorable story-teller candidates over those most likely to succeed.
3. Mind the self-promoters — Performance evaluations are a notorious platform for bias. For example, women are 1.4 times more likely to receive critical subjective feedback than are men. If you incorporate self-evaluation feedback in your evaluation of employee performance, remember that order matters. Thanks to the anchoring effect, we tend to skew our personal evaluation of something in the direction of the first value we hear. When we hear an employee’s rating of themselves before we make our own decision, we can’t help but skew our review accordingly. One big problem with this is that men are more likely to overrate their own performance than women, so anchoring on self-reviews can create an unfair evaluation ground. The lesson: ask for employee self-reviews after you write yours down – and check their subjective evaluation against more objective measures. Research suggests that conducting more frequent and objective feedback reviews may also help dramatically.
Thanks to the anchoring effect, we tend to skew our personal evaluation of something in the direction of the first (review) value we hear.
4. Establish diverse role models — One of the biggest barriers toward a thriving diverse community is stereotype threat, an unconscious tendency people have to fulfill the prophecy of stereotypes held against them. Combat stereotype threat by showcasing aspirational individuals of diverse identities. Encourage a diverse range of leaders to speak at company events, and make sure that commemorative hallway portraits are broadly representative. One study found that women who were shown a picture of Hillary Clinton or Angela Merkel before giving a public speech performed better than those who were shown a photo of Bill Clinton or no photo at all. Seeing is believing.
5. Lead like a scientist — Don’t just do it – test it! Any new policy or program change should be carefully designed and measured, to know whether it had the intended effect. Identify up front exactly what outcome you intend to influence with the new program, how you plan to measure it and where that metric stands now. Then pilot the concept in a random sample of offices, measure the results and decide whether and how to implement across the whole organization. Even data driven organizations like MIT are finding that in addressing workplace bias, as in all things, measurement is the first step toward progress. The university doubled their number of female faculty in STEM disciplines in 12 years, following a more focused analysis of gender bias. Best of all, share your journey with the world, by publishing your findings.
Business leaders are waking up to the pervasive problem of bias in the workplace. Behavioral science shows us that the greatest levers for change are already in our hands. We can use behavioral insights, and our own company data, to transform our workplace for a more inclusive and thriving future.
This article originally appeared on TLNT.com.