Systemic Bias Puts a Stranglehold on Diversity Hiring

You don’t have to look too hard to see the problem with legacy admissions at the more elite universities in the U.S. In fact, this article (June 2018) describes a discriminatory lawsuit Harvard is facing based on a study showing 30% of their recently accepted students have at least one parent who went to the school. 

A comparable problem exists when it comes to hiring. Here’s some proof that most hiring process are designed to hire replicas of people who were previously hired: 

• Companies write job descriptions that require the same academic and experience background as everyone else they’ve already hired. This cloning is a big problem on its own, but worse since diverse candidates tend to have a different mix of skills, experiences and academic backgrounds. As a result, these highly competent people are automatically excluded from consideration. 
• The idea that new hires must fit an existing culture is anti-diversity by definition. But the bigger problem is that the cultural fit assessment is left up to each hiring manager to decide. A better approach might to rank cultural differences higher than similarities. 
• Assessments based on first impression, which is largely driven by cultural bias, determine the types of questions asked and who gets hired. People who don’t meet the standard cultural fit stereotype are assumed less competent and asked harder questions to prove their incompetence.  
• Traditional knockout questions are designed to weed out the weak but in reality filter out diverse candidates who can do the work but have a different skill set. 

It doesn’t take much insight to realize that people naturally want to hire people just like themselves and most hiring processes are designed to ensure this occurs. Some of this happens during the filtering phase and the balance during the interview. By comparing the factors representing a great candidate to those defining a great hire (see image), it’s clear there is no correlation between the two. 

To me it makes no sense to build sophisticated AI machines to be more efficient finding great candidates when the traits don’t correlate with performance. Worse, this flawed preselection criteria eliminates the best people who have a different mix of skills, those who are willing to negotiate the compensation if the job is a great career move and anyone who doesn’t make a perfect first impression. 

Becoming more efficient via AI will not solve this type of systemic problem. It requires disruption and reengineering. Here’s how to get started.  

Huge Disruptive Change: Define the Job Before Defining the Person for the Job

I know this is crazy talk but skills- and experience-laden job descriptions by their very nature are anti-diversity for the above reasons. By defining jobs as lists of 5-6 KPOs (Key Performance Objectives) that describe the major and critical subtasks everyone can be assessed on the same criteria. If people are competent and motivated to do the work they obviously have the appropriate mix of skills, experiences and competences. This will rarely be in the same mix as on the original job description. More important, it removes the lid on quality of hire and broadens the diversity pool. 

While shifting to a performance qualified assessment approach is the essential first step in the process redesign effort, eliminating interviewer bias during the interview is as critical. Here are a few ideas we’ve been using on hundreds of search projects over multiple decades that have been extremely effective:  

• Proactively overcome first impression bias during the interview. Using semi-scripted performance-based questions ensures everyone is evaluated on the same criteria. It also helps to conduct exploratory phone screens before inviting candidates onsite. (More ideas)
• Make it a group decision. Panel interviews are useful for exposing bias as long as roles are assigned and the panel is focused on fully understanding the person’s major accomplishments as they relate to the KPOs of the open job.
• Substitute evidence for emotion. A formal debriefing process with interviewers verbally sharing specific evidence is essential. Not only is bias exposed this way, but comments based on “feelings” are discarded. We’ve found this talent scorecard to be a great tool for collecting the necessary objective evidence.     
• Create a buyer-to-buyer relationship. Too many hiring managers think they’re the buyer and  candidates the sellers. This embedded bias ensures a biased outcome. By recognizing both parties as equal buyers, interviewers treat prospects respectfully, objectively and professionally. 

Companies all talk the good talk describing the importance of hiring great talent and have made expanding their diversity hiring efforts a major goal. Yet after years of trying, little progress has been made. Short-term initiatives and being more efficient doing the wrong things masks activity for progress. The elimination of skills- and experience-laden job descriptions and shifting to an evidence-based performance qualified selection process will begin the disruption. Otherwise the good talk will continue to be only that.  

____________________________________

Lou Adler (@LouA) is the CEO of The Adler Group, a consulting and training firm helping companies implement Performance-based Hiring. He's also a regular columnist for Inc. Magazine, SHRM and BusinessInsider. His new Performance-based Hiring mobile- ready learning app - The Hiring Machine - is now available on your smartphone. His latest book, The Essential Guide for Hiring & Getting Hired (Workbench, 2013) is written for recruiters, hiring managers and job seekers, describing how evidence can be given, gathered and used to make better hiring decisions.