New Commentary in Cell Reports Sustainability: The Case for Integrated Assessment to Design Data Centers of the Future
Yong Suk Lee joins nine Notre Dame colleagues in a Cell Reports Sustainability commentary arguing that data center development is governed through fragmented processes, and that a justice-centered integrated assessment approach is needed to co-design the data centers of the future.
Data centers are being built faster than anyone has settled on how to evaluate them. Energy, water, land use, and local fiscal decisions are each handled by a different authority, on a different clock, using a different definition of what counts as a cost — so a set of choices with lasting consequences for host regions is made without any single place where those consequences are weighed together.
A new commentary in Cell Reports Sustainability argues that this fragmentation is the core problem, and that it is fixable. Yong Suk Lee is a co-author, joining nine Notre Dame colleagues across engineering, political science, global affairs, and economics: Joule A. Bergerson, Alexander Dowling, Debra Javeline, Xu Han, Hannah Legatzke, Susan L. Ostermann, Jennifer L. Schaefer, Rafia Zaman, and Arun Agrawal.
As the authors put it, data center development "produces interconnected and increasingly lasting and significant outcomes for society, but current approaches addressing data center development are fragmented." Their proposal is a justice-centered integrated assessment approach: a framework that harmonizes diverse evidence that is spatially resolved — because the burdens and benefits of a data center land in particular places, not in national averages — and temporally explicit, so that commitments of very different durations are not treated as equivalent. The aim is not a single verdict on whether data centers are good or bad, but evidence organized well enough to inform the co-design of future facilities.
For the Future of Labor Lab, the piece connects to a question the lab has pursued across other technologies: when a general-purpose technology arrives quickly, which effects get measured, which get ignored, and who bears the ones nobody counted. Data centers are the physical infrastructure of the AI economy, and the community and labor consequences of building them belong in any honest account of AI's economic footprint.
The commentary appears in Cell Reports Sustainability volume 3, issue 9, article 100865, and is available open access at doi.org/10.1016/j.crsus.2026.100865.