Tensions are increasing between laws that say certain types of information should not be used in hiring — age, sex, religion — and employers’ use of the Internet to screen candidates.
An experiment conducted by two Carnegie Mellon University researchers found evidence that sharing personal information on online social networks can lead to hiring discrimination.
Researchers Alessandro Acquisti, associate professor of information technology at CMU’s H. John Heinz III College, and Christina Fong, senior research scientist at CMU’s Dietrich College of Humanities and Social Sciences, tested the impact that information posted on social networking sites such as Facebook by job candidates can have on employers’ hiring behavior.
“While it appears that a relatively small portion of U.S. employers regularly searches for candidates online, we found robust evidence of discrimination among certain types of employers,” Fong said.
Results of their experiment, involving more than 1,000 survey participants and more than 4,000 employers, were released on Thursday.
They tested responses of employers to a Muslim job candidate and a Christian candidate, and to a gay candidate and a straight candidate. The results found significant discrimination against the Muslim candidate compared with the Christian candidate among employers in Republican-leaning states. In those states, 17.3 percent of Christian candidates received an interview opportunity versus 2.3 percent of Muslim candidates.
No evidence of discrimination against the gay candidate relative to the straight candidate was found.
Political values were determined using Gallup Organization lists of the 10 most strongly Republican and 10 most strongly Democratic states.
“Our survey and field experiments show statistically significant evidence of hiring bias originating from information candidates shared on their online profiles,” Fong said.
Douglas G. Smith, managing partner at Jackson Lewis LLP, a Pittsburgh law firm that specializes in workplace issues, said lawsuits have been filed nationally claiming hiring discrimination based on information obtained from online social networks. “The best practice is that employers should not be accessing social media sites to get information on candidates. Our advice is to stay away.”
Discrimination laws prohibit asking about religion, national origin, age, disability — a whole list of categories, Smith said. “The risk you run on a social media site is an applicant may discuss their religion or a disability. Then the employer has the information, and even if they make a legitimate decision, they face an uphill battle in court in making a defense on an employment decision.”
Brian Balonick, an employment attorney and shareholder at Buchanan Ingersoll in Pittsburgh, said there’s “too much of a risk for employers of someone claiming, ‘I have a characteristic that was discovered on Facebook.’ ”
“Social media in human resources has become a hot topic in a lot of ways,” Balonick said. “There’s always this tension: Does the employer need to know that or not? Social media has put that in our lap.”
Last year, Facebook warned users to protect passwords and their privacy after an increase in reports of employers or others seeking to gain access to people’s Facebook profiles or private information. “This practice undermines the privacy expectations and the security of both the user and the user’s friends,” Facebook said.
The experiment used real information found online from actual members of popular social networking and job search sites such as Monster, LinkedIn and CareerBuilder. Acquisti and Fong then used that data to form job candidate resumes, profiles and names for the experiment.
Acquisti said they manipulated personal traits for the fictitious candidates regarding religion and sexual orientation but held constant information on professional ability and work ethic.
“We wanted to give candidates the same level of professional backgrounds so that if we found significant differences in interview opportunities, it had to be due to employers independently searching online to access profiles on their own. We did not supply links to the profiles,” Acquisti said.
A first step was an online survey involving more than 1,000 people reacting to the fictitious candidates’ resumes and online profiles to test whether they appeared realistic. About half of the participants said they had some experience in human resources, Acquisti said.
Then in the field experiment, the researchers submitted applications on behalf of the candidates to real job openings at more than 4,000 employers.
They collected data that helped them get a sense of how many employers searched for job candidates online.
Acquisti said it is not possible to track visits to a social media profile by a specific person or account. But it is possible — by using Google’s advertising search capabilities — to track how many searches are done for a specific term or name in a given period of time.
“We were tracking over time whether the names of our candidates were being searched for or not, and to make some estimates of the frequency they were being searched,” Acquisti said.
John D. Oravecz is a staff writer for Trib Total Media. He can be reached at 412-320-7882 or [email protected].