Data Centers Need to Adopt a Baseline on Electricity Costs

The most dangerous issue in the data center debate is not land use or even water. It is electricity prices. When residential bills rise, projects do not get delayed. They get stopped.

This is where the industry faces its clearest test. Not whether it can secure power, but whether it can do so without shifting costs onto the communities it enters.

For too long, the conversation has been framed around load growth. More demand, more infrastructure, more investment. But that framing misses the point. The real question is who pays for that growth, and when.

If a data center increases residential electricity bills, it is a failed project structure.

That is not an activist position. It is a baseline for legitimacy. Without it, no amount of economic development messaging will hold once utility bills begin to rise.

The problem is not demand itself. The grid has always evolved to meet new forms of industrial load. The problem is how that demand is integrated into the system.

In markets such as Virginia, Texas and parts of the Midwest, utilities and regulators are already facing growing scrutiny over infrastructure investments tied in part to large-load growth from data centers and AI facilities. Those investments are often recovered through rate structures that spread costs across all customers. 

When that happens, households effectively subsidize infrastructure built to serve private, large-scale users.

That is the point where opposition becomes political.

Residents may tolerate construction, noise and land use changes. They will not tolerate higher monthly bills tied to projects they do not directly benefit from.

This is why the next phase of data center development requires a different standard. Not just access to power, but accountability for its cost.

The principle is simple: No cost shifting.

Large energy users should operate under dedicated rate structures that isolate their impact on the grid. These structures are already emerging in multiple states and utility territories, but they are not yet universal.

They need to become the default.

That includes minimum take provisions that require data center operators to pay for a substantial portion of their contracted capacity regardless of actual usage. If infrastructure is built to serve them, they should carry the financial responsibility for that commitment.

It also includes upfront contributions to grid infrastructure, ensuring that new substations, transmission upgrades and related investments are not financed through general ratepayer pools.

These mechanisms are not punitive. They are protective.

They protect existing customers from absorbing costs they did not create. They also protect developers by reducing political risk and increasing the likelihood that projects move forward without prolonged opposition.

This is where the industry has an opportunity to reset the narrative.

Instead of defending load growth as an abstract economic benefit, it can define what acceptable load growth looks like in practice. Growth that is fully priced, fully disclosed and fully aligned with those who create it.

That is a stronger position.

It acknowledges the reality of rising demand while setting clear boundaries around who bears its costs. It shifts the conversation from whether data centers should consume large amounts of power to how that consumption is structured.

Right now, there is no consistent framework for evaluating this at the project level. Rate structures vary by jurisdiction. Assumptions about cost recovery are often opaque. The connection between a specific project and its impact on local rates is rarely presented in a way that communities can understand.

That lack of clarity creates space for fear, and fear turns into resistance.

It also creates risk for capital.

Projects that trigger ratepayer backlash face delays, renegotiations, and in some cases cancellation. In a market where speed to power is critical, those delays carry real financial consequences.

The industry has the tools to address this.

Dedicated rate classes, minimum take agreements and upfront infrastructure payments are not theoretical constructs. They are being implemented in parts of the country today. What is missing is their adoption as a baseline expectation rather than a negotiated exception.

The next phase of development will require that shift.

The industry does not need broader promises around growth. It needs a consistent way to structure and evaluate power demand at the project level.

This is not about limiting growth. It is about structuring it in a way that communities can accept.

Because once electricity bills become the story, the project is already at risk.

And, in this cycle, the projects that survive will not be the ones with the most demand. They will be the ones that prove they can grow without making someone else pay for it.

Suhail Y. Tayeb is clinical assistant professor at New York University’s Schack Institute of Real Estate and director of the Center for the Sustainable Built Environment.