Does Representation Matter? Assessing how Representative Bureaucracy Shapes the Relationship Between Automation and Responsiveness to Public Service Requests
With Gregory Porumbescu
April 28, 2025, 11:45 – 1:15pm, Thompson 620
Registration Requested Here
Automated technologies in public service are often portrayed as replacing human discretion and jobs. Advocates argue that substituting technology for government workers can streamline public services and enhance transparency, while critics warn that removing human judgment may undermine responsiveness—especially for clients needing personalized attention.
Despite these debates, empirical research on how workforce demographics shape automation’s adoption and outcomes remains limited. Noting this lack of evidence, we contend the job displacement narrative is overstated, suggesting instead that bureaucrats continue to play a vital role in service delivery—even in automated contexts. Building on representative bureaucracy theory, we propose that demographic congruence between public employees and the communities they serve influences the implementation and outcomes of automated service request systems.
To test this claim, we compiled a novel dataset of over a million service requests from five U.S. cities, merged with American Community Survey data and workforce demographic profiles obtained through FOI requests. Our findings show that representative bureaucracy significantly influences automation outcomes. In cities where workforces more closely reflect the public, government responses to underrepresented communities improve—especially in neighborhoods with historically lower engagement. We also pinpoint a key mechanism: frontline workers actively leverage automated systems to advocate for their own communities’ needs, a phenomenon amplified by workforce representation.
Contrary to the idea that algorithms simply eliminate human discretion, our research shows automation transforms discretion’s role and location in service delivery. In more representative workforces, frontline employees use automated systems to amplify representation of underserved communities. In less representative settings, these technologies often reinforce existing inequalities. We thus move beyond a narrow job replacement narrative to highlight how workforce diversity mediates technology’s impact on service equity, reframing automation as a redistribution of tasks rather than a mere elimination of discretion.
Gregory A. Porumbescu (PhD, Seoul National University) is an associate professor in the Department of Public Administration and Policy at the University of Georgia’s School of Public and International Affairs (SPIA). His research centers on understanding the implications of technology for government transparency and accountability. Dr. Porumbescu’s work has been published in journals such as the Journal of Public Administration Research and Theory, Public Administration Review, Governance, and Social Science & Medicine. Prior to joining SPIA, Dr. Porumbescu served as an associate professor at Rutgers University–Newark. There, he was a co-founding principal investigator for the New Jersey State Policy Lab, an initiative dedicated to enhancing evidence based policy making in state governments. During his time at Rutgers, he was also appointed to serve on the AI, Equity, and Literacy Working Group, contributing to Governor Phil Murphy’s New Jersey AI task force. Dr. Porumbescu’s research has been supported by organizations such as the National Science Foundation, Korean Research Foundation, and the New Jersey Office of the Secretary of Higher Education.
This event is sponsored by the Institute for Social Science Research and so-sponsored by the School of Public Policy.
AI x Democracy Symposium
April 23, 2025, 10:00 – 3:00pm, LGRC Lowrise, A112
Registration Requested Here
One of the most debated topics surrounding artificial intelligence (AI) is whether AI technologies help or hinder the enactment and practice of important democratic processes.
This symposium brings together experts to share insights from their work exploring the intersection of AI and democracy. They will highlight areas of research and development working to create technological tools and approaches that truly advance democracy and democratic processes at local, national, and international levels. They will provide a case study that will help participants think creatively and critically about technology-enabled solutions to entrenched social problems.
Breakfast and lunch will be provided.
See the event website for more information.
Sponsoring partners for this event include the Responsible Technology Coalition (RTC); the Community, Democracy, and Dialogue Initiative (CDD); Manning College of Information & Computer Sciences (CICS) ; Public Interest Technology Initiative at UMass Amherst (PIT@UMass).
Research at the Intersection of Technology, Diversity Science, and Health
March 26, 2025, 11:00 – 1:00am, Computer Science 150/151
- Rae Walker, Associate Professor of Nursing
The “A.I. Nurse”: Tech for Good or Path to Apartheid? - Hari Balasubramanian, Professor of Mechanical and Industrial Engineering
Identifying Associations Between Diseases Using Data Mining and Clustering Methods - Shannon Roberts, Associate Professor of Mechanical and Industrial Engineering
Impact of Driving Automation and ADHD Symptomatology on Young Drivers’ Behavior - Michael Constantino, Professor of Psychological and Brain Sciences
Measurement-Based Matching of Mental Healthcare Patients to Psychotherapists’ Strengths
Hosted by the Institute for Diversity Sciences and the Public Interest Technology Initiative at UMass.
Book Release and Author Talk: Mood Machine The Rise of Spotify and the Costs of the Perfect Playlist by Liz Pelly
February 20 2025, 2:00-3:00pm, Location: Integrated Learning Center (ILC) Hub. Room N301.
Join the Responsible Technology Coalition (RTC) for a discussion with author and journalist Liz Pelly about her new book “Mood Machine The Rise of Spotify and the Costs of the Perfect Playlist“. This work explores Spotify’s origins and influence on music, weaving unprecedented reporting with incisive cultural criticism, illuminating how streaming is reshaping music for listeners and artists alike.
Building on her years of wide-ranging reporting on streaming, music journalist Liz Pelly details the consequences of the Spotify model by examining both sides of what the company calls its two-sided marketplace: the listeners who pay with their dollars and data, and the musicians who provide the material powering it all. The music business is notoriously opaque, but here Pelly lifts the veil on major stories like streaming services filling popular playlists with low-cost stock music and the rise of new payola-like practices.
Additionally, the RTC will be hosting a book club for “Mood Machine” following the talk (exact date TBD) and 10 free copies of the book with be given on a first-come first-serve basis for those who indicate their interest on the RSVP form below. The books will then be given out at the end of the book talk.
Food and drinks will be provided! Please pre-register for the event here: https://forms.gle/cesueX9pLX47dqjE6
Hosted by the Responsible Technology Coalition, a grad student club.