Make Data Centers Good Grid Citizens

June 2026

PROBLEM

Thirty-eight states offer dedicated tax incentives for data centers. The electricity these facilities consume is on track to triple between 2024 and 2028. At the same time, public backlash to data centers is accelerating: at least 11 states have moved to restrict or ban data center development since 2025, and Maine’s legislature became the first to ban construction outright.

Several jurisdictions that granted tax abatements years ago are now moving to revoke them, responding to constituent anger over electricity prices and a broad sense that the deal struck with industry has not delivered what was promised. Governor Shapiro in Pennsylvania announced in May 2026 that the state would condition data center tax incentives on meeting grid and energy standards, requiring facilities to pay for their own transmission infrastructure, use clean energy, and reduce demand during peak hours. Since then, Democratic governors in Illinois and Arizona have paused incentives, Ohio’s governor froze new exemptions, Utah’s governor issued an executive order requiring data centers to demonstrate benefits, New York’s legislature passed a moratorium bill with broad conditionality requirements, and Texas’s Governor Abbott is pledging to repeal data center tax incentives.

The problem is not that states created these programs; it is that states have handed out subsidies without requiring anything about how these facilities interact with the grid. Without tax incentive conditionality, the data center buildout will strain infrastructure, raise residential rates, lock in fossil generation, and generate political backlash that forecloses the investment the country needs.

IDEA

States should pass legislation that conditions the award or renewal of data center tax abatements on the recipient demonstrating “grid citizenship”: a defined set of behaviors that ensure large load customers contribute to, rather than extract from, the grid infrastructure they depend on. Specifically, a qualifying data center would need to meet the following requirements:

Take-or-pay load commitments. Certify that projected electricity demand forecasts are accurate within a defined tolerance, and accept financial liability for infrastructure built to serve demand that never materializes. Only 13% of capacity that entered interconnection queues between 2000 and 2019 reached commercial operation by 2024, leaving ratepayers to absorb the difference.

Full cost causation. Agree that transmission and distribution infrastructure built specifically or primarily to serve the facility will be cost-allocated to the facility, not shifted to the general ratepayer base. This must be legislated or otherwise mandated because utilities earn returns on rate base investment and face structural incentives to prefer cost recovery from ratepayers over cost causation.

Verified demand flexibility. Demonstrate the technical capability to curtail or shift a defined percentage of load during grid stress events, and commit to participating in utility or grid operator demand response programs, with third-party metering, real-time telemetry, and actual dispatch tests before certification.

Real clean energy procurement. Rather than purchasing clean energy certificates that do not change what is actually being delivered to the grid or the data center facility, they should contract for new, additional clean and firm generation before the facility comes online. This means the facility is causing clean capacity to exist that would not have otherwise.

Transparency and reporting. File annual verified load forecasts with the state utility regulator, including actual versus projected consumption data. Grid operators are trying to plan a trillion-dollar infrastructure buildout on unverified hyperscaler projections that may or may not be accurate.

Banning data centers is a protest, not a policy. The buildout will happen; the only question is whether it happens with ratepayer protections, grid commitments, and clean energy requirements attached, or whether it happens somewhere else without them. States that move now can still answer that question on their own terms.

EXPECTED OUTCOME

If successfully implemented, abatement conditionality would produce four durable outcomes:

Ratepayer protection. Cost causation requirements prevent transmission and distribution costs from being shifted to residential bills. Take-or-pay commitments eliminate stranded infrastructure costs when facilities underdeliver.

Better grid planning. Verified annual load forecasts can give utilities and Regional Transmission Organizations (RTOs) the data they need to plan proactively rather than reactively, directly supporting implementation of the Federal Energy Regulatory Commission (FERC) Order 1920’s proactive regional planning requirements.

Clean energy acceleration. Mandatory procurement of new, additional clean firm energy can drive cost declines and deployment in technologies like long-duration storage, nuclear, and geothermal. Long-term offtake commitments from hyperscalers are among the few demand signals large enough to pull these technologies down the cost curve.

A public record for federal and regional reform. State conditionality creates defined terms that hyperscalers either accept or refuse, on the record and in public. That record is what federal and RTO reform requires, and it makes hyperscalers active stakeholders in fixing the grid.

This framework does not by itself build the grid. FERC must finalize interconnection reform and implement Order 1920. RTOs need to incorporate verified large load forecasts into forward-looking planning. Federal investment in clean firm power must close the gap between what markets will finance today and what the grid needs. But state conditionality is a place for that work to start — and it can begin immediately.

CONSIDERATIONS

Benefits:

Direct ratepayer protection through cost causation and take-or-pay liability, preventing large load infrastructure costs from being spread across residential bills.

Strong public support: concern about electricity costs is nearly universal, and voters support specific accountability measures for large energy users.

Generates the public evidentiary record — verified forecasts, defined commitments, documented compliance — that federal and RTO reform requires.

Converts hyperscalers from passive grid customers into active stakeholders in fixing the infrastructure they depend on.

Costs and Risks

Race-to-the-bottom risk: developers may shift investment to states with laxer rules. This risk is real but mitigated by hyperscaler location constraints (fiber corridors, labor markets, existing infrastructure) and diminishes as more states adopt similar frameworks.

Full compliance is not immediately achievable everywhere. Most facilities today cannot meet the clean firm procurement requirement because the product does not exist at scale. The framework must be tiered, with achievable immediate requirements and a graduation pathway as infrastructure and policy catch up.

Several utilities are negotiating special electricity service agreements (“large load tariffs”) with data center customers that impose cost allocation, timing, and–occasionally–clean energy requirements for conditions of service. While useful, these are insufficient on their own: they are negotiated by utilities that profit from rate base growth and kick in after states have already committed their subsidy. Conditionality operates upstream of both problems.

POLLING*

Polling from The States Forum shows that–across party lines–opposition to data centers flips to 56% support when states require data centers to pay for their own grid infrastructure, use clean energy, and reduce peak demand.

47% of voters say the problem with data centers is that they do not pay their fair share (versus only 21% who say the problem is that they were built at all). Again, this is consistent across party lines.

65% of voters would allow data center construction if they pay for their own grid infrastructure. This decreases if it is framed as anti-business or anti-growth.

FAQ:

Who is likely to support this idea?

Ratepayer advocates and consumer groups focused on preventing cost being passed to ratepayers
Clean energy developers who stand to benefit from mandatory hyperscaler procurement
State policymakers facing constituent anger over electricity bills and data center proliferation
Grid operators and transmission developers who need verified load data for planning
Hyperscalers with genuine clean energy commitments who want a level playing field against competitors making only nominal claims

Who is likely to oppose this idea?

Data center developers and hyperscalers facing new compliance costs
Utilities with structural incentives to prefer socialized cost recovery
State economic development agencies worried about losing investment to competitor jurisdictions

Won’t data centers just move. to states without these requirements?

Hyperscalers are constrained by fiber corridors, labor markets, existing infrastructure, and proximity to power transmission. The race-to-the-bottom risk diminishes significantly as more states adopt similar frameworks. The goal is interstate coordination, not unilateral action — and the more states that move, the stronger the baseline becomes.

Doesn’t this just ban data centers by another name?

No. Immediate requirements are achievable everywhere today (verified forecasting, take-or-pay liability, transparency reporting). The clean firm procurement condition includes a graduation pathway that rewards facilities for advancing toward full grid citizenship as infrastructure and policy catch up. The goal is to steer the buildout in societally desirable directions, not to stop it.

What about the jurisdiction problem in restructured markets?

This is a real constraint. In RTO markets, cost allocation is governed by FERC-approved tariffs, limiting what states can require directly. However, state conditionality operates through the abatement. a state tool, rather than through rate regulation, which preserves state authority. Also, the framework generates the public record and political proof of concept that federal and RTO reform requires, without depending on that reform to take effect.

*Polling data cited from a Data for Progress survey of 1,159 U.S. likely voters conducted May 1-4, 2026. The sample was weighted to be representative of likely voters by age, gender, education, race, geography, and recalled presidential vote. The margin of error is ±3 percentage points. For more information on methodology, visit dataforprogress.org/our-methodology.