Politics
AI data center boom raises questions about housing, power costs and economic risk
A projected $10 trillion U.S. data center buildout is reshaping investment, while experts warn that electricity costs, public subsidies and a narrow reliance on AI could leave communities exposed. Construction spending on data centers has surged as residential construction value has fallen.
The U.S. is entering a data center construction boom on a scale that could rival the great infrastructure expansions of its past, but the rapid shift in investment is raising concerns about housing, electricity bills and the economy’s dependence on artificial intelligence.
A Columbia Business School analysis estimates that data center investment could reach $10 trillion and account for nearly 4% of U.S. gross domestic product. By comparison, railroad construction from 1870 to 1890 cost about $173 billion in today’s dollars and represented more than 2% of annual economic output over two decades, according to figures provided by the report’s author, economist Stijn Van Nieuwerburgh.
The boom is already reshaping construction spending. The value of new data center construction more than doubled over two years, reaching $85 billion annually, while the value of other construction fell 8%, according to an analysis of U.S. Census Bureau data through August. Since 2024, data center construction value rose by $48 billion as the value of new homes and apartments declined by $29 billion.
Adam Shapiro, a vice president at the Federal Reserve Bank of San Francisco, described the trend in a September post as a shift in investment from residential construction toward computers.
The largest projected data center expenditures over the next three years are in Texas, at $638 billion, followed by Virginia at $249 billion, Ohio at $169 billion and Georgia at $160 billion, according to Industrial Info Resources. Projects are also planned across a wide range of states, including Arizona, California, Florida, Iowa, North Carolina, Oregon, Pennsylvania and Washington.
States that once offered tax breaks to attract facilities are reassessing whether the deals cover the public costs. Data centers require large amounts of electricity and water, while their tax exemptions can reduce local and state revenue. Nevada, for example, has scaled back some tax abatements and is seeking commitments that companies cover the energy and water demands of their projects.
Power use is a growing point of tension. Van Nieuwerburgh estimates the buildout could double electricity consumption. A study by researchers at Carnegie Mellon University and North Carolina State University found that data centers and cryptocurrency mining together could increase electricity bills by 8% nationally and as much as 25% in parts of Virginia.
Oregon officials have continued a freeze on approvals for data center projects following a state review that identified utility affordability as the most frequent concern in public comments. Virginia, meanwhile, began applying a surcharge on data center electricity use in July as part of a budget deal that kept major tax incentives in place. Some Virginia localities are weighing pauses on new projects.
The economic case for such rapid expansion remains uncertain. David Pickering, senior vice president of research at Industrial Info Resources, said the industry’s profitability has not yet caught up with its costs, though he expects major companies to find ways to make investments pay off as the market and regulations develop.
Economists have also warned that the boom could prove fragile if expected profits from AI do not materialize. Josh Bivens, chief economist at the Economic Policy Institute, said reliance on a single sector creates a vulnerable base for growth. The International Monetary Fund has similarly warned that weaker-than-expected AI investment or profitability could trigger a market correction.
Paul Krugman has argued that the long-term employment gains may be limited because data centers require relatively few workers once they are built, even as construction projects create jobs in the near term. Mark Higgins, a historian of U.S. financial markets, sees a parallel with the railroad era, when heavy investment was followed by financial panics and distressed assets.
The scale of the buildout puts local governments in a difficult position: attracting investment and construction jobs while ensuring that residents are not left to shoulder higher utility costs or lose revenue needed for public services. Whether AI-related profits eventually justify the spending will help determine how durable the boom—and its benefits to communities—will be.