The Data Center Backlash is Not About Data Centers
Brief #4 from the UVA NSI Data Center Opinion Monitor | Source: UVA NSI survey fielded September 3 2026
October 23, 2026
The backlash to data centers has spread across the country, reaching from the major cities to wealthy suburbs to small rural communities. Opposition cuts across party lines, fueling a rare bipartisan movement. The movement has resulted in hundreds of data center bans across the country and an estimated $68 billion in blocked or stalled projects.
Public opposition to infrastructure projects is not new. Residents often resist facilities that bring local costs with no benefits to their communities, even when they acknowledge that there are national-level benefits. This pattern is referred to as “NIMBYism” (“not in my backyard”), and for decades it has shaped debates over projects ranging from affordable housing to nuclear power plants. Data centers are the latest facilities to trigger this kind of resistance. But is opposition to data centers just NIMBYism, or is something else driving it?
We ran a randomized survey experiment comparing data centers to three other facilities to identify the why behind the opposition: distribution warehouses, nuclear power plants, and landfills. Each of these facilities have been known to be the subject of NIMBYism, and they share some features with data centers but differ on others. This allows us to isolate the features that are driving the opposition to data centers. For example, distribution warehouses and data centers share an association with Big Tech (Amazon for warehouses, AI for data centers), but differ on the amount of resources they consume and the amount of permanent jobs they bring to the area. Accordingly, this allows us to isolate the effect of a Big Tech association from the effect of local economic benefits. Landfills take up space but are an obvious necessity. Nuclear power plants bring along environmental concerns and obvious hazards, but aren’t associated with big business.
Table 1
Facility Features: Developer and Technology
| Facility | Big Tech | Tied to AI | Newer / Unfamiliar Technology | Privately Owned | Often Run by a Public Utility or City |
|---|---|---|---|---|---|
| Nuclear Power Plant | ✔ | ||||
| Warehouse | ✔ | ✔ | |||
| Landfill | ✔ | ||||
| Data Center | ✔ | ✔ | ✔ | ✔ | ✔ |
Table 2
Facility Features: Costs and Concerns
| Facility | Requires a lot of Land | Environmental Concerns | Safety Concerns | Property Value Concerns | High Water Use | Strains Grid / Raises Utility Bills |
Health Concerns | Heavy Truck Traffic | Noise | Odor | Eyesore |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Nuclear Power Plant | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | |||||
| Warehouse | ✔ | ✔ | ✔ | ✔ | |||||||
| Landfill | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | |||||
| Data Center | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
Table 3
Facility Features: Benefits
| Facility | Many Permanent Local Jobs | Directly Serves Local Residents | Main Benefits Go to People Outside the Community | Essential Public Service | Large Local Tax Revenue | Produces Something Tangible |
|---|---|---|---|---|---|---|
| Nuclear Power Plant | ✔ | ✔ | ✔ | ✔ | ✔ | |
| Warehouse | ✔ | ✔ | ✔ | |||
| Landfill | ✔ | ✔ | ✔ | |||
| Data Center | ✔ | ✔ |
We recruited a politically representative sample of 809 American adults and randomly assigned them to evaluate one of four facilities. We introduced the hypothetical facility using the prompts listed in the table below.
Table 4
Four Conditions and Prompts
| Condition | Prompt |
|---|---|
| Nuclear Power Plant | Imagine a company is interested in building a nuclear power plant within 10 miles of your home to generate electricity for the regional power grid |
| Warehouse | Imagine a company is interested in building a distribution warehouse within 10 miles of your home to store and ship products to consumers and businesses |
| Landfill | Imagine a company is interested in building a landfill within 10 miles of your home to manage household and commercial waste from the surrounding region |
| Data Center | Imagine a company is interested in building a data center within 10 miles of your home to support AI technology |
Americans would rather live by a nuclear power plant than a data center
We asked all respondents how they would feel about their respective facility being built within 10 miles of their home. Each person was randomly assigned to evaluate just one of four facilities: a data center, a landfill, a nuclear power plant, or a distribution warehouse. Because no one saw more than one, respondents judged each facility on its own terms rather than ranking them against each other. This allows us to rigorously isolate the psychological drivers behind NIMBYism for each facility and the associations with the particular parameters of each type of facility.
Of the four facility types we tested, people liked distribution warehouses the most and landfills the least. Nuclear power plants have had a decades-long reputation for being notoriously difficult to site because resistance to them is so widespread and intense. But opposition to data centers now outpaces opposition to nuclear power plants. Americans would rather live by a nuclear power plant than a data center.
Data centers are the only facility where benefits are lowest at home
We also asked participants to report how much they expected each facility to benefit five groups: themselves, their neighborhood, their town or city, their state, and the country. In general, NIMBYism is a story about local costs outweighing local benefits, and that’s what we found. People’s willingness to accept a local facility was shaped in large part by whether they expected the facility to bring benefits to their town or city. Warehouses bring the most local benefits and are the most welcome. Data centers ranked the lowest on their potential to bring local benefits. At the national level, people acknowledge that data centers render benefits to almost the same extent as nuclear power plants. But data centers were the only facility whose expected benefits grew the farther they reached from home: low at the local level, higher at the state level, and highest at the national level. This is a key pattern that drives NIMBYism: people do not want to suffer local costs for a project that is going to render benefits away from home.
People are tolerant of having a warehouse nearby because they expect them to bring benefits locally. In particular, they expect jobs. We asked respondents to write down the thoughts they had as they evaluated their new facility, which allows us to capture the concepts that come to mind naturally. These free responses revealed that people were evaluating data centers and warehouses on a dimension of economic benefits to the town, and were evaluating nuclear power plants on a safety and risk dimension. 1 in 3 people spontaneously brought up “jobs” when evaluating a warehouse, and all of those responses were positive. By contrast, 15% of respondents mentioned “jobs” when evaluating data centers, and 1 in 5 of those responses expressed skepticism.
Table 5
Quotes from respondents' free responses that mentioned 'jobs'
| Warehouse | Data Center |
|---|---|
| “[A warehouse] could create jobs and help the local economy” |
“[Data centers] provide no financial benefit to residents and provide no jobs” |
| “Jobs!! Economic boost to local restaurants and businesses.” |
“AI will ... take away jobs” |
Data center opposition is driven by builder opportunism and a distrust in the federal government
We asked people to consider the type of company that was behind the facility and to rate how opportunistic they are: Would they would break promises and alter facts to get what they want, and are they willing to hurt the community to get ahead? Of the four builders behind the facility types we tested, data center developers seemed the most opportunistic and warehouse builders the least. This goes hand-in-hand with people’s expectations for locally rendered benefits: people view data center builders as opportunistic because they aren’t bringing local benefits; people view warehouse builders as more invested in the community because they expect jobs.
This is surprising because distribution warehouses bring Amazon to mind, and data centers bring AI to mind, both of which fall under Big Tech. It wasn’t long ago that Amazon was the go-to corporate villain. But now, it seems that there is a bigger, badder villain on the scene: AI infrastructure that is taking land and resources without offering jobs, and, in some people’s minds, will eliminate jobs. The perception that data center developers are opportunistic builders is a big driver behind opposition. Among people who saw data center developers as opportunistic, 90% did not support a local build. Among those who didn’t, 41% did not support a local build. Distrust of the developer helps explain why data centers face so much more opposition than warehouses.
Distaste for local data centers is also driven by Americans’ distrust in the federal government, Big Tech, and Big Business. We asked respondents how much trust they have in four American institutions: Big Tech, Big Business, the federal government, and their local government. Trust for the first three is very low across the board. People reported that they “somewhat distrust” those three institutions, an average of -1 on a scale from -2 to +2. That distrust doesn’t factor into their evaluations of landfills, nuclear power plants, or warehouses. For data centers, it matters a lot: people who distrust the federal government are 32% more likely to oppose a local data center.
The data center backlash is not actually about the data centers
This is all pointing to the broader theme: the data center backlash is not actually about the data centers. When people evaluate landfills or warehouses or nuclear power plants, they are evaluating the pros and cons of that facility. When people evaluate data center propositions, they are passing judgment on the AI industry at large. And their distrust for the federal government translates to a sense of economic instability brought on by AI and failed to be regulated by the government. Which means that those that trust the federal government feel less threatened by the AI industry, and by-proxy, data centers.
Indeed, people only mentioned distance (“miles”) 1.5% of the time when they were thinking about a local data center. And the three most common words were water, AI, and environment - notably abstract concepts. In the three other conditions, respondents were talking about the actual facility in concrete terms. The most common words across the three other conditions were: “smell,” “miles,” close,” “safe,” and “traffic.” When people evaluated warehouses or landfills or power plants, they asked themselves: am I close enough to smell it, hear it, or sit in its traffic? Almost nobody did for data centers.
Underneath it all is a crisis of trust. People are uneasy about economic instability and the pace at which things are changing under their feet - and they don’t trust the federal government to steady them. They also don’t trust data center developers to act in their interest. Perceived opportunism and expected benefits go hand in hand: people don’t expect data centers to render local benefits because they think data center developers are out for themselves, and vice versa. The two beliefs reinforce each other, which means either end has a lever to pull. Concrete, visible local benefits can build trust, and more trust makes promised benefits more believable. Offering residents some form of local benefit that promises to support local economic growth and offer some sense of economic stability will go a long way toward restoring trust and cooling opposition. But developers will need to convince people that they are serious when they offer to make an investment.
Data Center Opinion Monitor
A nation’s power is increasingly measured in compute, and compute lives in data centers. This brief is part of our ongoing Data Center Opinion Monitor, which tracks the American public’s support for data centers as they expand their footprint across the country.