Infrastructural development in the U.S. has been politicized differently in different industries. For example, construction on oil pipelines and gas plants is often framed as a jobs vs. environment issue, gaining support from labor unions but facing opposition from environmental protection advocates. But the current scale-up of infrastructure necessary for American competitiveness in the global arms AI arms race, including the rapid construction of data centers, power grid upgrades, and the expansion of energy production necessary for AI training, from natural gas to batteries to nuclear, currently lacks a salient framing in policy and media rhetoric.

Instead, data center development is viewed as a jobless form of growth, and a huge energy sinkhole whose costs are being subsidized by residential ratepayers. Rather than viewing data centers as a public utility, whose costs should be absorbed by all users, most media reporting has framed data center development as a conflict between narrow Big Tech interests and the public.

As a result, in some regions this growing industry is facing popular opposition from residents who decry its environmental externalities and suspect policymakers of ushering through sweetheart deals between data center developers and utilities monopolies, resulting in pricing schemes that push the costs of AI energy use onto the public. In these regions, this popular opposition is leading to slowdowns in data center development, despite federal government support for the industry and an increasing realization among local governments that data center development can provide a substantial and long-term tax base.

To investigate regional variation in the controversial politics of data center development, I am interviewing energy professionals, data center professionals, local political leaders and members of lobbying groups and civil society organizations concerned about data center development.

All research participants are granted full anonymity, and I am not interested in industry-proprietary knowledge. See here for consent forms detailing privacy agreements for research participants.