How 360 Degree Performance Reviews Undermine Equity & Inclusion
The results of my recent performance appraisal are disturbing. Most of my reviews (all anonymous) provide constructive feedback, but there are a few that criticize my “failure” to engage and encourage team members because I’m “overly confrontational.” My company does 360 degree reviews to prevent bias but I’m starting to doubt their objectivity. So my question is this: are 360 reviews effective at eliminating bias from the performance evaluation process?
360 degree reviews seem like an effective way to mitigate rater bias from performance evaluations. In practice, however, that’s not always the case.
There’s a saying in data science that applies well to this situation: garbage in, garbage out. It doesn’t matter how robust your model is; if you input the model with bad data, your output will also be bad.
Now, I’m not calling your colleagues’ reviews of you garbage. No. The point is that 360 degree performance reviews do not guarantee good input data. Just because a 360 degree review collects more data points from a diverse set of stakeholders — colleagues, managers, and direct reports — doesn’t make them more reliable. More data does not mean better data.
Key Point #1: Everyone is prone to bias. Having more people evaluate your performance doesn’t make their opinions less biased.
Research published by Harvard Business Review found that only 15% of women managers and 24% of men managers were confident in the employee evaluation process, and most managers viewed the process as ”subjective and highly ambiguous.”
That’s concerning considering how more than half of the 100 large organizations surveyed at a performance management summit said they weren’t taking any steps to reduce bias in performance reviews.
And THAT is concerning because performance evaluations disproportionately disadvantage women. Women not only receive more critical, more negative, and more useless feedback than men, they also receive more feedback on their communication style.
A linguistic analysis found the word “abrasive” appeared 17 times in 13 separate performance reviews of women. The same word appeared zero times in men’s performance reviews.
And THAT is concerning because…
Key Point #2: Biased reviews feed inequity loops because these reviews determine how people get paid and who gets promoted.
Performance evaluations might feel like a check-the-box-for-HR type of activity, but they are much more than that. They play an important role in the employee lifecycle, specifically in influencing promotion and compensation decisions.
Pipeline found through its implementations that men receive promotions at a 21% greater rate than women. For Black women, the gender promotion gap doubles.
One key way to shrink the promotion gap is to de-bias the inputs (i.e. performance evaluations) that determine who is promoted. Remember the equation: biased data in = biased data out = suboptimal decision-making.
Key Point #3: Use advanced technology to remove bias from the employee lifecycle, starting with performance reviews.
So how can you de-bias performance evaluations? As we’ve discussed, it’s not necessarily by collecting more performance evaluations. And as I’ve written about earlier, it’s not necessarily by conducting more implicit bias training, either.
The solution is AI. Currently, 65% of HR professionals believe AI can improve D&I. They’re right. Powerful technologies exist (such as machine learning and NLP) that can read and root out bias from performance reviews. Companies can hardwire these solutions into their evaluation processes to detect, mitigate, and overcome bias. This is the future of work.