Course:ELEC300E/CaseStudies/AlgorithmsforJobSeekers
Context: AI and Algorithmic Hiring
| Jobseeker-Centric Algorithmic Tools |
|---|
 |
| Content / Topics |
| Needs and Emotions |
| Engineering Workplaces |
| Labour Rights and Advocacy |
| AI Training |
| Collective Action |
| Relevant Analysis Strategies or Tools |
| Decision Impact Assessment |
| Iceberg Model |
Algorithms play a big role in the job search and in the hiring process: these days the majority of employers use AI to filter, screen and sort applicants.
Clearly these algorithms have a big impact, but how do they work, what values do they reinforce, which power hierarchies do they uphold (or dismantle), and what are their potential adverse impacts?
Stanford researchers analyzed 4 million applications and found:
1) Adverse impact for Asian and Black applicants. The researchers found significant evidence of racial bias in the hiring process, with Black and Asian applicants directed to less advantageous positions.
2) Algorithmic monocultures lead to systemic rejections. Many employers rely on the same algorithms and tools - meaning that if candidates are rejected by those tools, they may be rejected from all positions they apply for.
Case: Job-Seeker Centric Algorithmic Tools
Many hiring algorithms are designed with the employer's interests in mind. After all, the employers purchase and use these technologies, and they stand to benefit from systems that help them hire effective and well qualified candidates.
Job seekers are also heavily impacted by these systems though. What are their needs and priorities, and how might a Jobseeker-Centric tool be different from an employer-centric tool?
European researchers recently looked at this problem, designing an algorithmic decision making system in partnership with job seekers, to match applicants with public sector jobs. Through workshops with jobseekers and job center case workers, they identified a few key needs and priorities for jobseekers:
1) "The Value of Human Contact": Navigating faceless, impersonal online systems can be frustrating, confusing and discouraging.
2) "Seeking Genuine Orientation": Receiving low assessments without constructive feedback or suggestions is demotivating. Rather than just rejecting applicants from jobs, skill assessment tools could help guide job seekers to the education or training they need.
3) "Being Seen as a Whole Human": Having your value reduced to a simplistic score can feel disrespectful, and can overlook valuable skills and attributes that are more difficult to quantify.
Recommended Readings
- 10 minute read, "Algorithmic Tools in Public Employment Services: Towards a Jobseeker-Centric Perspective": https://dl.acm.org/doi/epdf/10.1145/3531146.3534631
- Alternatively, watch the 4 minute video presenting this paper: https://www.youtube.com/watch?v=1etGmI1XHf4
Discussion Questions
Here are a few questions to help you begin analyzing the issues discussed in this case study. The purpose of these questions is to prompt reflection and further consideration: there are no right or wrong answer
- Personal response: Imagine being a job-seeker who is being evaluated by an AI-based algorithm (if you are close to graduating it won't require much imagination!). How would you feel in this scenario? What would be your main concerns or objections?
- Contradictory viewpoints and outcomes: The researchers in this case study identified three main priorities for job-seekers (listed above). How might these priorities conflict or contrast with the priorities of employers?
- Social and Historical Context: How do the issues with AI-based hiring processes relate to what you know of the labor movement and workers rights? You can refer to the Canadian Labor Congress to learn more about Canada's Labour Movement history, and its current priorities. Is there one labour principle, or story from the history of the labour movement that you think relates to this case study?
- Positive action: Consider designing a job-seeker centric hiring system. What is one feature or design decision that stands out?
- Interruptions and invitations: This case study is an example of co-design or participatory design: the researchers and job seekers redesigned the hiring system together collaboratively. How do you think co-design is different from stakeholder consultation, or community engagement? What are the advantages and disadvantages to this approach to design?
- Physical connections: The job seekers in this case study emphasize the "value of human contact". Thinking back to your best previous mentors and work experiences, do they involve some form of human contact? Would interacting with an AI chatbot have helped you in the same way? Why or why not?
- Positive Practices: What is one concrete action that you could take, or a practice that you could adopt in your personal, academic or professional life to address some of the issues, or put into practice one of the approaches discussed in this case study?