OPINION: Algorithms used to vet job candidates allow biases to slip into hiring process

Two hands holding a resumé. The hands are patterned with gears.
Graphic by Natalie Bauer

Is my application or resume ever going to reach the hands of an actual person? How can I best tailor my application to match the criteria of a computer program going through thousands of applications that, for the most part, are pretty much the same? Why does a computer get to decide whether I get a job?

As I’ve been submitting application after application for summer internships, these questions have all crossed my mind.

With the high number of applications received for single positions nowadays,  many companies are turning to automated applicant tracking systems (ATS) and artificial intelligence to vet candidates in the hiring process.

It’s been strange navigating the technology-infiltrated hiring process. To increase my chances of getting my resume past computer programs, I’ve been told to submit my resume only in the format of a Word document, use keywords matching job descriptions and get rid of any fancy formatting (which, by the way, I love). 

I’ve done a recorded video interview with HireVue, a software that allows companies to have applicants record themselves responding to questions on video, then analyzes participants’ speech and facial expressions.

As a first-year, I have limited experience with the hiring process, and wanted to know more about how I was being evaluated as a job candidate, so I did some research online. What I found shocked me, and showed me that computer programs are really not yet sophisticated enough to serve as fair evaluators of job applicants.

Since most computer programs that sort through resumes rely on machine learning and base their scoring systems on patterns seen from past applicants, they can often reinforce prejudices. 

For instance, last year, Amazon was caught up in a huge scandal over its gender-biased application-sorting algorithms. Since most past applicants were male, Amazon’s algorithms favored patterns like keywords related to males more so than those relating to females. 

In fact, they penalized applicants for having the word “women’s” on their resume, and even blacklisted applicants who had graduated from two women’s colleges. Having learned its lesson, Amazon no longer uses computer programs to sort through job applications.

And in December 2018, Upturn conducted a study titled “Help Wanted: An Examination of Hiring Algorithms, Equity and Bias” that concluded that “predictive hiring tools can reflect institutional and systemic biases.”

Additionally, algorithms are subjective even in their targeting of potential job applicants. Researchers at Carnegie Mellon University found that Google’s advertising system shows high-paying jobs more often to men than to women.

Companies like Hilton and Unilever now use HireVue, the software that I had to use for my recorded video interview, and one that analyzes candidates’ speech and facial expressions. Problems are abound with the use of such technologies as well. 

For example, people from different cultural backgrounds do not have the same perception of facial expressions, and HireVue’s speech analysis technology may discriminate against people with different accents.

Human beings are not infallible, and computers can make mistakes, too. The biases of computer programs that are designed by subjective human beings have made me question the fairness of using algorithms to vet job applicants. 

We shouldn’t use algorithms in the hiring process until technology becomes sophisticated enough to the point where it can rate job candidates objectively, lest we allow for discrimination to perpetuate in the hiring process.

Technology is a wonderful thing. Every day, it awes me with the things that it can do — address society’s most pressing problems, cure the incurable and even solve the unsolvable.

But we must be more cautious in exercising our use of technology. This is especially pertinent to many of my fellow students at Harvey Mudd College who are pursuing careers in science and technology.

I guess that’s the whole reason HMC’s mission statement reads: “Harvey Mudd College seeks to educate engineers, scientists and mathematicians well versed in all of these areas and in the humanities and the social sciences so that they may assume leadership in their fields with a clear understanding of the impact of their work on society.”

The impact of our work on society is something to keep in mind as we navigate the complex world of technology.

Michelle Lum HM ’23 is from San Jose, California, and just wants a job. To the computers out there that are currently busy scanning her resume, she pleads for them to extend some sympathy to her.

Facebook Comments