Heather Bussing is a returning contributor to our HRExaminer Editorial Advisory Board. Heather has practiced employment and business law for over 20 years. She has represented employers, unions and employees in every aspect of employment and labor law including contract negotiations, discrimination and wage hour issues. While the courtroom is a place she’s very familiar with, her preferred approach to employment law is to prevent problems through early intervention and good policies and agreements. Full bio…
I downloaded a whitepaper from Taleo last week through ERE. It’s called “Social Network Recruiting: Managing Compliance Issues.” If you’d like a copy, go to the Taleo website. It’s under Resources, then articles. (I’d give you a link, but they need your contact information before you can open it.)
The article starts by warning that using social networks for recruiting can increase the risk of attracting a disparate impact discrimination lawsuit because the racial profile of social network users does not exactly match the racial profile for the general US population. There’s a cool pie chart that show that the US population is about 13% African American and a little over 15% Hispanic. But the LinkedIn user population is only about 5% African American and 2% Hispanic. An attorney was quoted saying that she expected more race and age claims based on the idea that using social networks as a source of candidates had a disparate impact on minorities.
Here’s why that doesn’t really make sense. The general population and the number of people qualified for a job are almost never exactly correlated. If the job requires the person to speak Spanish and a sourcer searches LinkedIn for candidates that are bilingual—she is going to get a disproportionate number of Hispanics as candidates – and there’s nothing wrong with that. If a sourcer needs a nurse, he will end up with a disproportionate number of female candidates. It is the search that matters—not simply the pool of resumes in the database. Besides, discrimination based on a legitimate job qualification is legal even if the available pool of candidates with that qualification is limited to a sector of the population that looks nothing like the general US population.
Next, the paper raised a concern about getting information that might not be relevant to the job-qualifications and could include pictures, comments or groups that would reveal that the person is in a protected class. Federal laws prohibit discrimination based on race, color, religion, sex, national origin, pregnancy, age or disability. States like California have broader protections that include marital status, ancestry, medical condition, and sexual orientation.
The warning was: “if some candidates are subsequently eliminated from consideration, some of them are almost certain to conclude it was because they were gay, pregnant, disabled and so on.”
But how will they ever know?
When sourcers research people online, most of those people never, ever know they’ve been “sourced.” And most don’t care. When people put information on LinkedIn or Facebook, they expect to be looked at. That’s the whole point of LinkedIn.
I just don’t see the lawsuit by a class of people who say that no one found them to consider them because they weren’t on LinkedIn. (Although there is probably a lawyer somewhere who would try it.)
Further, even if someone did learn that a company looked at their resume online, how would they ever know that they didn’t get a job they never applied for, were never contacted about and may not even want.
I would never take a case that claims disparate impact based on an employer’s sourcing practices alone. In order to prove a disparate impact claim, the plaintiff must make a threshold case by showing that the challenged policy or practice had a discriminatory effect on a protected class.
That is usually done by showing that the percentage of people hired is statistically different from the pool of potential applicants. The potential applicant pool is generally the people qualified for that job in that labor market.
Sourcers do not keep a list of every person they have ever researched for every job they are sourcing. (If you try to make them, they’ll quit–they like the hunt, not record keeping.) And while there may be a digital record of pages viewed, it is expensive to obtain and there’s no way to know what job related to what page or person. It would cost thousands of dollars just to figure out if there might be a viable claim.
I suppose if a particular sourcing employee only offered 50-year-old white guys for every position she ever researched, there might be a case. But there is no way for a 30 year old Hispanic woman, who did not even know she was being considered to know that.
Even if you can show that there is a statistical disparity that affects a protected class, you still have to show that the very same imbalance actually exists in the workforce. So although one sourcer only proposes 50-year-old white guys, there is no case if the actual people hired are a mixture of race, age and gender based on the qualified people for the job in that area.
In other words there actually has to be discrimination—not just a practice that might result in missing a qualified candidate in a protected class.
Disparate impact cases of actual applicants are very difficult to prove. Disparate impact of sourcing practices will be even harder because there is less evidence and the potential “victims” of the discriminatory practice will probably never know.
So I don’t think it’s worth getting your britches in a bunch over fear of lawsuits from using LinkedIn or Facebook for sourcing. But do look around at who you hire and who you don’t hire. Do actually look at all that data that your cool ATS software collects and see if you need to make any adjustments. And don’t discriminate against someone qualified for the job because of how they look, who they pray to or the ultrasound picture on their Facebook page. Because that’s illegal.
Heather Bussing is a California employment lawyer who has practiced labor and employment law for 23 years.