A view from DC: What can history teach us about data center resistance?

When one tries to describe the full transformative potential of the technological revolution of the moment, it can be easy to lapse philosophical.

It "is an enormous, an incalculable force practically let loose suddenly upon mankind; exercising all sorts of influences, social, moral, and political; precipitating upon us novel problems which demand immediate solution; banishing the old before the new is half matured to replace it; bringing the nations into close contact before yet the antipathies of race have begun to be eradicated; giving us a history full of changing fortunes and rich in dramatic episodes."

Such was the promise and potential of the railroad, at least as viewed through the mid-19th century eyes of Charles Francis Adams, grandson of a U.S. president and later himself president of the Union Pacific. Though by no means a critic, Adams was a relative reformer among the railroad's zealots, who had become captive to their own narrative of inevitable progress.

The rise of railroads in North America, especially as achieved at great cost through what became the so-called transcontinental project, provides fascinating historical echoes of today's generative AI rush. The lessons of the era are probably best captured — whether we decide to heed them or not — in historian Richard White's 2011 book "Railroaded: The Transcontinentals and the Making of Modern America."

'No secret deals'

Even as AI investment is forecast to exceed USD1 trillion in 2026, public sentiment about data center development in the U.S. has soured at a remarkable rate. A new Annenberg Public Policy Center survey found that 61% of U.S. adults opposed data centers in their area, up from 49% in its spring survey. Opposition is bipartisan, with the survey recording 69% of Democrats, 54% of Republicans, and 53% of independents as opposed.

Hundreds of U.S. jurisdictions have adopted or initiated some form of data center pause, ban or restrictive action. Last month, New York became the first state to place a one-year moratorium on data centers over 20 megawatts. State-level legislators in at least 14 other states tried to pass similar legislation this year, according to a tracker managed by the National Conference of State Legislatures. 

Responding to the rapid politicization of this issue in the midterm elections, governors are taking matters into their own hands. Gov. Greg Abbott, R-Texas, announced a moratorium at the beginning of August, and last week Gov. Josh Shapiro, D-Penn., signed an executive order to place some limits on Pennsylvania data center development, a complete about-face from his prior position on the issue. 

Election wonks are buzzing and campaigns are rapidly shifting tactics amid what Punchbowl called the "hot summer of data center politics." Most recently, the Senate GOP campaign arm sent letters to AI companies warning that Sen. Jon Husted, R-Ohio, seems poised to lose his seat as Democrats make data centers a "centerpiece" of their campaign against him.

From railroads to powerlines to server farms

Westward expansion, as delivered by railroad development, was not so unpopular among U.S. citizens, caught up as they were in the excitement of progress. But there was no shortage of local resistance at the time, as railroads facilitated the final phase in the extermination, subjugation and relocation of Native Americans. New infrastructure enabled U.S. military power to exert control over vast swathes of territory that had previously only been colonized on paper. Local resistance to this development cycle is a tale told in blood, sweat and trails of tears.

Today's development projects do not bring nearly this level of risk or harm to local communities. Nevertheless, the major efforts at resisting the development of hyperscale data centers deserve closer inspection. For a more compatible historical precedent, we can look to other major infrastructure projects, such as the electrification of the U.S. through the spread of high-voltage overhead power lines. 

Another excellent book tells this story, with a focus on how rural farming communities in Minnesota responded to the perceived threats posed by long-distance electrification — as their farms were overtaken by development over which they felt little control and realized no benefits. In "Powerline: The First Battle of America's Energy War," Barry M. Casper and Paul David Wellstone show how local resistance, if not taken seriously, can evolve from polite town hall meetings, street signs and administrative law complaints to coordinated efforts at sabotage and asymmetric warfare.

The sacrificial lamb problem

"Powerlines" highlights one of the clear preconditions for concerted community resistance to development. Distributed technological systems localize their social costs among populations that derive no distinctive benefit from accepting them. This is termed the "sacrificial lamb problem." If not thoughtfully structured, a national project of progress and innovation can become a highly localized reallocation of land, authority, risk and public wealth.

Data centers concentrate the physical requirements of an ostensibly immaterial digital economy in particular communities. According to local activists, the lived experience of this development cycle feels like the appropriation of land, electrical capacity, water, tax benefits, atmospheric quiet and local planning authority. 

Communities feel left behind and taken advantage of. This has already led to a repeatable pattern of resistance in disparate communities across the U.S., from Saline Township, Michigan to Peculiar, Missouri to Hampton, Georgia. 

The playbook starts with discovering a development project too late. Locals create a Facebook group, hold meetings in informal civic spaces, pack zoning hearings, investigate and demand a moratorium. If local leaders refuse to budge, activists turn to a legal playbook, combing through zoning rules, land records and utility filings, turning to litigation, recalls or ballot initiatives. Successful or not, they spread what they learn to other communities. National networks are growing rapidly, with the Stop Data Centers Coalition reporting more than 500 participating organizations.

Thus, we see a recurring American drama playing out: infrastructure is justified at a national scale, planned in a competitive commercial environment in abstract technical terms, and then contested at the scale of the farm, household, town or ancestral landscape.

Avoiding the bolt weevils

For now, data center activists do not seem to be turning to self-help to stop development projects. This was not the case for the farmers in Minnesota, who became disenchanted with the democratic process when they concluded their county and utility commissions were already captured by special interests. 

They escalated to asymmetric techniques, physically blocking development projects with farm machinery and massive piles of manure and eventually organizing into gangs of vigilantes. Calling themselves "bolt weevils," the farmers shot out thousands of glass insulators or even toppled entire transmission towers. Eventually this resulted in the deployment of state troopers, but it delayed transmission line developments for years.

The railroad story has still other warnings for the current moment. Professor White's 12-year project of contextualizing the most important infrastructure achievements of the 19th century led him not to conceive of railroad tycoons as robber barons or even genius-level monopolists, as others have done, but rather as shockingly incompetent grifters who used political capture and investment mania to their personal advantage while building unsustainable corporate giants that eventually failed. 

He is careful to acknowledge that this was never a choice between having railroads and not having them. Much of the crippling economic by-products of the era — which, incidentally, led to the rise of antimonopoly sentiment and the eventual creation of the U.S. Federal Trade Commission — were related to infrastructure development simply outpacing demand. Railroads were built to reflect national pride in anticipation of a new world order that did not materialize quite as predicted. Transcontinentals were not built "in response to a popular demand for development of these lands. Instead, they created the demand through vast promotions unlike anything seen until that time."

White's book is a warning about the dangers of developing too fast, without a clear understanding of how new technology will shape the future. Writing in the middle of the last recession, he could not help but see the parallels: "As in the nineteenth century, highly leveraged corporations, marketing dubious securities that were more inventive than comprehensible even to their creators, precipitated massive losses, receivership, government rescues, and severe economic downturns. The present seems so nineteenth century."

Political and economic forces are rapidly adapting to the reality of AI. This election season will teach us a lot about how the current technological revolution compares to those of the past.

Please send feedback, updates and railroad stories to cobun@iapp.org.