Post by Stanford University Graduate School of Business

Nonprofits that distribute food and recruit volunteers often encounter a trade-off between fairness and efficiency. However, Associate Professor Daniela Saban finds that with the right data, they may no longer have to choose. Saban partnered with two nonprofits to improve their systems for allocating personnel and donations. “Their data science teams are understaffed and overworked,” she says, so she and her collaborators stepped in to help. For VolunteerMatch, the team redesigned the algorithm that surfaces opportunities to users. Previously, some organizations were flooded with applicants while others got none. The new algorithm moves popular opportunities down as they receive interest, bumping less popular roles to the top. “We were able to distribute the sign-ups more fairly, covering the needs of more opportunities, without decreasing the total number of sign-ups,” Saban explains. Feeding America had a different problem. Smaller food banks often missed ad hoc donations — not because they didn’t want the food, but because they lacked staff to respond in time. The revised system offers donations to multiple agencies at once and breaks ties based on need. In simulations, lost donations decreased while distribution became more equitable. “Nonprofits are amazing organizations with very smart people doing very important work,” Saban says. Her findings show that the tradeoff many accept as a given can be a false choice. “There can be a way of improving equity without much harm to efficiency.” https://brnw.ch/21x460f