2873. Can justice be profitable? The benefits of balancing equity and equality in optimization models
Invited abstract in session MB-20: Just and ethical sustainability transitions, stream OR and Ethics.
Monday, 10:30-12:00Room: 45 (building: 116)
Authors (first author is the speaker)
| 1. | David Rea
|
| DATA, Lehigh University | |
| 2. | Leonardo Lozano
|
| University of Cincinnati | |
| 3. | Craig Froehle
|
| Operations, Business Analytics, and Information Systems, University of Cincinnati |
Abstract
Employees care about being treated fairly and will react strongly to perceived unfairness. As such, organizations are increasingly seeking ways to incorporate employee-centric principles in their decision-making processes. This presents the challenge of balancing two sets of trade-offs: the organizations well-being versus their employees, and their employees' desire for equality versus equity. Ensuring these tradeoffs are consistently balanced according to the organization’s values at the scale and speed of modern business requires an algorithmic approach. The increasing use of algorithms to guide such value-driven decision-making necessitates a better understanding of how these elements should be integrated within decision-making models. This research explores these complex trade-offs computationally by modeling two representative case studies: a synthetic gig-economy package delivery problem and an empirical interhospital patient transport problem, both of which are NP-hard. One key observation is that certain combinations of functions in the model objective lead to opportunities for organizations to perform better while also representing fairness more precisely. We also find that the effects of trade-off parameters are partially determined by both their relative levels and the functions used in the objective. These and other observations offer guidance for practice while also motivating future research to help fully explore how these concepts shape decision-making.
Keywords
- Ethics
- Optimization Modeling
- Vehicle Routing
Status: accepted
Back to the list of papers