Fit for What Purpose? Evaluating Data Centers on Local Terms
Fit for What Purpose? Evaluating Data Centers on Local Terms
by Sara Kassir, LCSSP Research Policy Lead
September 1, 2026
LCSSP's new white paper grew out of a workshop we convened earlier this summer on the practical challenges communities face in governing data center development. Our main takeaway was that the question of whether data centers are "good" or "bad" in the abstract has no answer. What matters is how well a particular project aligns with what a community is already trying to achieve, and whether local conditions allow those goals to be met.
Much of the challenge of working at the intersection of science and public policy comes down to identifying the right questions. Research is more likely to be useful when it accounts for the situation decision-makers are faced with and helps them make sense of the available options. Technical knowledge can tell us a great deal about how systems behave, which tradeoffs matter, and what consequences different choices might produce. For policymakers, the more important question is how to apply those insights under the specific conditions they govern.
Part of LCSSP's work is to look for places where the application of Caltech research may serve as a valuable input into public decision-making. In recent months, data centers have emerged as one clear opportunity. As these facilities have become caught up in broader political arguments about AI, economic growth, energy, and corporate power, the debate has increasingly centered on whether data centers are universally good or bad — leaving many of the practical questions communities face when evaluating actual proposals unanswered.
It was in this context that we decided to convene a workshop earlier this summer with the Resnick Sustainability Institute and the Caltech Science Exchange. We brought together people involved in local data center decisions — elected officials and staff, city planners, utility representatives, water resource managers — with experts working on land use and environmental review, local economic impacts, grid and cooling engineering, and public opinion. The goal was not to steer participants toward a common position, but to put different kinds of knowledge in conversation with the people already confronting these decisions and see what emerged.
The individual presentations offered different kinds of guidance, but one of the most interesting takeaways came from synthesizing across them. Whether the topic was mitigating water strain, devising contingencies for grid upgrades, or bolstering local tax revenue, what counted as a sensible response depended heavily on the conditions and priorities of the community being asked to host the project. In other words, there was no meaningful way to evaluate a data center in the abstract; the relevant question was how a particular project fit a particular place.
Here, an obvious question follows: fit against what? If there's one thing all of the pushback against data centers has demonstrated, it's that communities expect to have a meaningful say in what gets built there and on what terms. Assessing the context-specific appropriateness of a project therefore means looking first at the community's existing priorities — the goals, constraints, and tradeoffs that were already present before the project entered the picture. Some places are looking for productive uses for idle land or infrastructure, new public revenue, stronger energy infrastructure, or progress toward clean-energy goals. Others are primarily trying to protect household affordability, shared local resources, public health, or environmental quality. A single project may further some of those objectives while working against others.
This need to evaluate a proposal in context — against both a community's existing priorities and the conditions under which the project would operate — is a far cry from most of the arguments being put forth about these facilities. It is also the organizing idea behind our new white paper, A Question of Fit: Data Centers and Community Priorities. Building on themes raised during our workshop, the paper examines seven priorities that commonly shape local data center decisions, framed as four community "wants" and three "worries." For each, we ask what it means to evaluate a proposed data center against that priority — where a project might advance it, where it might conflict with it, and where pursuing one objective may require tradeoffs with another. The table below illustrates that approach.
For those looking for a clean answer to the ongoing debate over whether data centers are good or bad, a framework built around seven different priorities will probably be a disappointment. For the decision-makers actually tasked with evaluating these projects, that is precisely the point. The framework offers a way to identify which questions matter for the outcomes a community is trying to achieve, and which features of a proposed project will determine the answers.
The national debate over data centers is unlikely to become less polarized anytime soon. But cities and towns still have to decide what particular projects would mean for the places they govern. The better they can understand the demands, benefits, and tradeoffs involved, the better positioned they will be to decide whether a project belongs there — and on what terms.
Interested in subscribing to receive LCSSP's latest updates?
If you are external to Caltech, fill out this form.
If you are on the Caltech network, use these links to subscribe: LCSSP Mailing List, LCSSP BioPolicy Initiative, and LCSSP Democracy Mailing List.
For help or questions, please reach out to [email protected]!