You can accept data centers or you can refuse them, but, if you accept them, major costs are unavoidable and are rarely acknowledged by economic impact studies.
Community and state elected leaders are scrambling to address an array of social, economic, environmental, and quality of life issues that make their constituents resent and fear data centers and, oftentimes, the companies and billionaires that are behind them as well. Elected officials at all levels of government almost uniformly failed to anticipate the breadth and depth of their constituents’ opposition. Dazzled by immense investment figures, grandiose economic development claims, and expectations of greatly increased tax revenues, legislators, mayors, council members, and county commissioners regularly stumbled eyes wide shut into deals with data center developers, often going so far as to sign non-disclosure agreements, which, when discovered by their constituents, added distrust to the anger many felt.
In response, moratoria on new developments are being announced, tax incentives for data centers are being suspended or repealed, and procedural safeguards, which were once relaxed to encourage data center development, are now being enhanced. Meanwhile, electric utilities are being goaded and, in some places, required, to adopt rate structures that ensure data centers will pay, not just for the energy they use, but for the added infrastructure local and regional grids may require to serve them.
But, as determined as elected and appointed officials may be (up to and including President Trump who has demanded that utilities, developers, and power suppliers sign his Ratepayer Protection Pledge) some of the costs data centers impose are not addressed by these efforts and are beyond the reach of any effort to insulate existing customers against fallout from data center development.
First, even though data centers may be made to pay for all the electricity they use, all of the capital costs they induce, and even the environmental damage they do, their mere presence in the marketplace, as incredibly deep-pocketed buyers of goods and services we all use – not just electricity, but skilled labor, electrical equipment, computer chips, and construction materials among others – bids up the prices that we all must pay for those things. Carpenters and electricians become harder to find, computer chips, gas engines and turbines become more scarce as do building materials, and, as a result, the prices for these things and many more rise, making it more expensive for you to put an addition on your house or just pay your electric bill as the inflationary effects seep into other economic sectors.
In some cases, projects and expansions get cancelled altogether because homeowners and business owners decide they just aren’t worth the added cost. This is called the “crowding out” effect and it has already been documented in regions where data center construction represents a major share of market activity.
Data centers and their vast appetite for power also drive up the wholesale price of electricity, which falls not just on them but on all customers. And the impact is compounded if you happen to be served by a utility that finds itself in the position of having to buy or pay for the construction of a new power plant, whether or not the plant will serve a data center. In just the last three years, the cost of building new gas-fired power plants has roughly tripled, from less than $1,000/kw to nearly $3,000/kw. So, even if your utility is merely replacing an old, worn out plant with a new one to serve existing customers, the cost of doing so is suddenly high enough to make your rates rise substantially.
Finally, the induced inflationary effects of data center development now extend to interest rates, not just those we pay when we buy something on credit, but to the rate companies pay when they issue credit to expand and that the federal government must pay for the bonds it issues to finance federal spending.
In these ways, the inflationary impacts of data center development seep into and throughout the entire economy imperceptibly and ineradicably. And that diffusion of cost across the entire economy highlights another problem caused by data centers as they grow at unprecedented rates. The places that benefit most financially from data center development are not the places where most of the costs land.
The financial beneficiaries of data center development are largely limited to those who sell and support data center services and, to a lesser extent, the communities that host data centers, assuming they haven’t foolishly abated taxes. Consequently, a local county or municipality may realize substantial increases in property tax revenue as a result of a data center going into operation. However, the surrounding state and region may experience even greater financial losses as they absorb costs in the form of higher utility bills and crowding out effects.
This asymmetry between where costs and benefits land produces incentives for communities to strike deals which may be financially beneficial locally, but which, regionally, are net losers. This kind of effect was apparently realized in Pennsylvania in 2025, when it appears that, over the course of a year, the aggregate increase in ratepayers’ electric bills was greater than the total amount of new taxes paid by data centers.
In other words, even though data center construction may have produced large revenue increases for some localities, data center development may have resulted in a net reduction in disposable income for Pennsylvanians who faced much higher electric bills.
The costs associated with the induced inflationary impacts of data centers and the asymmetry between where costs and benefits arise are real and substantial . . . and, importantly unaddressed and largely unaddressable by any of the measures currently being developed or deployed to protect communities and ratepayers from the adverse impacts of data center development.
Because many of these costs are baked into our market economy and are rarely acknowledged in the putative economic impact studies that developers foist upon local and state policymakers, there often isn’t anything we can do about them. But they should be recognized, quantified, and considered as public policymakers contemplate whether to support and approve proposed data center projects and the terms on which they do so.