- A USC researcher found utilities like Rocky Mountain Power absorb data center demand costs without passing them onto the consumer, but that may not last.
- University of Utah research shows data centers initially lower local property tax rates, but equipment depreciation leads to volatility.
- State officials project 106 full-time jobs averaging $160,000 for one Millard County project.
Brenna Williams, a resident of Box Elder County, is one of many advocates in the Box Elder Accountability Referendum, or bear, who challenge the proposed Stratos data center which will be built in their backyard.
Aside from environmental concerns, Williams — a co-founder of bear — says the promises of economic advances from data centers sound appealing, but she believes “it’s all smoke and mirrors.”
Across Utah, residents are asking the same questions: Will data centers will raise their electricity bills? Will property taxes truly go down — and stay down? Will data centers developers deliver on their promises of jobs for rural communities?
Researchers say the answers depend on who provides your power, how rural your county is, and whether policymakers can build “the plane while you’re flying it.”
Electricity prices depend on the power provider
Shon Hiatt, an associate professor at the University of Southern California, told the Economic Development and Workforce Services Interim Committee on Aug. 19 that electricity prices nationwide have risen over the past six years.
In a study, he found that investor-owned utilities, like Rocky Mountain Power, which serves most of Utah, experienced no effect on power prices from data center entry from 2020 to 2025. Hiatt said these companies have the tools necessary to handle a big, new customer without passing the cost onto other consumers.
But that isn’t always the case, he said.
Non-investor-owned utilities, he explained, “have no idea how to deal with” new data centers coming in and using so many resources, “so they just raise the prices.”
Even for investor-owned utilities like Rocky Mountain Power, Hiatt said this trend in the past may not hold: “I would expect data centers coming in now, especially these larger ones, to have an impact on power prices wholesale ... .”
The boomerang effect: Volatile property tax rates
Researchers are calling it the “boomerang effect”:
Utah’s Truth in Taxation system splits property into two buckets: real property — the land and warehouse — and personal property — the computer chips and servers.
The building, which counts as a “new construction,” when first built, adds new tax money to the county’s existing property tax base. The more the tax base increases, the less county residents have to pay in property taxes that year.
At first, the data centers can lead to lower property tax rates for other residents.
However, because personal property depreciates, some of what they pay will decrease over time.
Maddy Oritt, director of public finance research at the University of Utah, and author of a study released earlier this month, said volatility follows the initial construction.
As the value of the data center’s personal property — computer chips and servers — depreciates, “the certified tax rate has to creep back up to generate the same amount of revenue,” she said.
The depreciation is fast. The study found the valuation would drop from $1 billion to $620 million after the first year, and down to $70 million by the fifth year. This means volatile rates for citizens.
But across the states, it has become an industry norm to require data centers to refresh their equipment every five years, another study from the Tax Foundation found — meaning property tax collections could be going up or down depending on when equipment was purchased.
Utah has yet to enact legislation mandating the replacement of hardware.
However, Oritt explains that as the rate begins to creep back upwards, based on the study’s modeling, “we don’t think that rate is going to ever exceed what property taxpayers were paying before.”
The volatility of the property tax weighs heavily on how rural the county is: The larger the existing property tax base, the fewer residents will be affected.
“If a two billion dollar data center went into San Juan County, that would more than double their existing property tax base, which means that the effect on existing property taxpayers would be quite a bit larger as opposed to a county like Salt Lake County where a $2 billion data center is only 1% of our existing property tax base,” said Oritt. “So, that’s going to mitigate the effect that property taxpayers would feel.”
106 jobs per data center — state officials call it a win
Lance Soffe, senior vice president of economic growth of the Governor’s Office of Economic Development, said there are many economic benefits to data centers —especially in rural areas — at the Aug. 19 meeting.
For the future Creekstone Energy Project data center, Soffe projected $344 million in state revenue, and $333 million in wages for Millard County, with 106 full-time jobs created, averaging $160,000 per job.
Soffe explained that hiring locally — and providing citizens with the training they need — is priority No. 1.
Jordan Hill, the newly hired workforce director for the Governor’s Office of Economic Development, works “hand-in-hand with Talent Ready Utah” and is “assigned on every single one of (the state’s) projects,” to build a direct pipeline between rural residents and employees, Soffe said.
Utah doesn’t write checks, but it does give tax incentives
Soffe claims the state does “not write checks to companies.”
Instead, Utah offers a post-performance tax incentive, rebating up to 30 cents per dollar of tax revenue in urban areas and 50 cents in rural areas. “There’s no reason to give back any of our tax revenue to jobs or companies that aren’t building exactly what we want to be and what runs our economy,” he said.
As with property taxes, Oritt claims that legislation around tax incentives will vary depending on the rural or urban nature of the county.
“There are certain areas of the state where I think it’s almost certain that we would have economic development anyway (without tax incentives) because they’re population centers, they’re employment centers, and if a data center weren’t to locate there, it’s almost certain that other development would happen there,” she said.
However, the math becomes very different when incentivizing a data center development in a rural area — one that doesn’t have a large population or employment center. There, her study shows, it might be beneficial.
Overall, Oritt explained that creating economic legislation on emerging technology “is a little bit like building the plane while you’re flying it.”
“We should be looking at what policies we want to make to make sure that we are getting the intended effects in our communities,” she said.
“I don’t know that we know what the future is going to look like in a year or five years or 10 years — policymakers know that they’re going to have to adapt on the fly.”