
New Jersey voters oppose data centers, with concerns about AI and big tech driving opposition, FDU poll finds
NEW JERSEY — Nearly two-thirds of New Jersey voters oppose making it easier to build data centers in the state, but their opposition appears to be driven more by concerns about artificial intelligence and large technology companies than rising electricity costs, according to a new Fairleigh Dickinson University poll released Friday.
The survey, conducted in partnership with the International Union of Operating Engineers Local 825, found that 65% of voters oppose making it easier to build data centers, while just 25% support easing restrictions on their construction.
The findings also show broad support for making it easier to build middle-class housing and new power generation facilities, suggesting voters favor development projects they believe could help address New Jersey’s housing and energy affordability challenges.
“We’ve seen concerns about both electricity prices and data centers increasing at the same time, so it’s been natural to assume that one is driving the other,” said Dan Cassino, a Professor of Government and Politics at Fairleigh Dickinson University, and the Executive Director of the FDU Poll. “But it’s attitudes about big tech companies and AI that are driving views about data centers much more than electricity prices.”
According to the poll, 79% of voters support making it easier to build middle-class housing, while 64% favor easing construction of solar and wind energy facilities and 57% support making it easier to build natural gas power plants.
Voters were more divided over warehouses, with 39% supporting easier construction and 41% opposing it.
“There is a sea change in New Jersey when it comes to increasing our energy supply with an ‘all of the above’ approach and building more homes. Voters in a fully nonpartisan way understand we need more of both to make our state more affordable,” said Greg Lalevee, IUOE 825 Business Manager. “Similarly, voters have a clear view that data centers won’t help them based on what they know.”
Opposition to data centers crosses party lines
The poll found that opposition to easing data center construction extends across political affiliations, although Republicans and self-identified MAGA voters were more supportive than other groups.
Among Republicans and self-identified MAGA voters, 41% favored making it easier to build data centers. However, even within those groups, a majority did not express support for easing construction restrictions.
Researchers found some partisan differences in attitudes toward other development projects, including greater Republican support for warehouse construction. Still, voters across political affiliations generally supported additional middle-class housing and power generation, although preferences differed over the types of energy facilities that should be built.
“Partisanship matters most when it comes to abstract issues,” said Cassino. “When voters are all looking at the same problems, like unaffordable housing and rising electric bills, the obvious solutions tend to be popular no matter what party you’re coming from.”
AI and big tech concerns influence views
To better understand what drives opposition to data centers, researchers conducted an experiment within the survey.
Before answering questions about development, respondents were randomly assigned to consider one of three topics: regulation of large technology companies, regulation of artificial intelligence or whether their electricity bills had increased in recent years.
Researchers then compared support for data center construction among respondents exposed to each topic, including those who expressed negative views about the issue they had been asked to consider.
The results showed little evidence that concerns about electricity prices significantly changed attitudes toward data centers.
Among respondents prompted to think about electricity prices, 25% supported making it easier to build data centers. That figure declined to 21% among those who also reported that their electricity bills had increased, a difference researchers said was not statistically significant.
However, attitudes shifted more substantially when respondents considered regulation of large technology companies or artificial intelligence.
Among respondents prompted to think about big tech regulation, support for easing data center construction was 26%. Among those in that group who also favored greater regulation of technology companies, support dropped to 14%, a 12-percentage-point difference.
A similar pattern emerged among respondents asked about AI regulation. Support for easier data center construction fell from 23% to 14% among those who also favored greater regulation of artificial intelligence, a nine-point difference.
“People in New Jersey don’t like data centers, but putting it in the context of electric bills doesn’t change anything,” said Cassino. “It’s when people are thinking about data centers in terms of AI and big tech that they really start to turn against them.”
The experiment also found that concerns about technology companies and artificial intelligence were associated with stronger support for renewable energy development.
Among respondents considering big tech regulation, support for making it easier to build solar and wind facilities increased from 66% to 77% among those favoring greater regulation of technology companies.
Similarly, among respondents considering AI regulation, support for easier solar and wind development increased from 69% to 77% among those favoring greater AI regulation.
Researchers said the findings suggest that addressing electricity demand alone may not be enough to overcome public resistance to data center projects.
“The ‘why’ of opposition to data centers is important,” said Cassino. “Data centers think that they can mute opposition by building their own power plants, but the data is showing that won’t be enough, and distrust of big tech companies is a much more important barrier.”
Poll surveyed New Jersey voters in October
The FDU Poll was conducted Oct. 16-22, 2026, using a voter list of registered New Jersey voters. Braun Research of Princeton conducted the survey through live telephone interviews and text messages directing respondents to an online questionnaire.
The university reported an overall sample of 805 registered voters, although its methodology also cited 807 voters in the sampling-error calculation.
Respondents were contacted through landline and cellphone interviews, as well as online surveys accessed through text messages. All interviews were conducted in English.
The results were weighted to reflect New Jersey’s likely voter population based on the 2024 and 2025 elections, accounting for demographic factors including age, sex, education, geographic region and race and ethnicity.
The university reported a sampling margin of error of plus or minus 3.4 percentage points at a 95% confidence level, increasing to approximately 5.1 percentage points when accounting for the effects of weighting.
Researchers noted that the margin of error does not account for other potential sources of variation, including nonresponse, question wording and survey context.
The findings come as New Jersey communities continue debating the construction of data centers and their potential effects on energy infrastructure, land use and local development.





