Data Center Energy Demand and the $23 Billion Electricity Cost Debate

The $23 Billion Figure: Revenue Increase vs. Public Cost

The claim that data centers have increased electricity prices for the general public by $23 billion is a simplification of PJM (Pennsylvania-New Jersey-Maryland Interconnection) market data. The $23.1 billion figure actually represents the total increase in capacity market revenue for the 2025/2026, 2026/2027, and 2027/2028 Base Residual Auctions (BRA).

This total includes both the payments made by the data centers themselves as new customers and the increased costs passed on to other customers due to higher market prices. Some analysts suggest the actual cost borne by non-data center customers is closer to $16 billion, rather than the full $23 billion.

The Role of Infrastructure and Policy

Whether the public pays for data center grid upgrades is a matter of policy and regulatory structure rather than an inevitability of energy physics.

Shared Infrastructure Costs

In many current utility models, investments in substations and additional electricity sources required to handle data center loads are treated as general grid improvements. These costs are often shared across all ratepayers, meaning residential and small business customers may subsidize the infrastructure that enables large-scale AI compute.

Alternative Regulatory Models

Some regions are implementing different approaches to prevent this cost-shifting. For example, Oregon has utilized the POWER Act to approve significant rate hikes specifically for data centers (such as a 29.7% hike for PGE) to ensure that the industrial users driving the demand bear the brunt of the associated infrastructure costs.

Arguments for Data Centers as Economic Drivers

Some perspectives suggest that data centers act as "anchor tenants" for the power grid, providing the financial justification for critical infrastructure upgrades that benefit everyone.

• Financing Improvements: Large-scale data center investments can accelerate the modernization of aging electrical grids that would otherwise lack the funding for upgrades.
• Economic Revitalization: Data centers are frequently built in regions with declining populations and limited economic prospects, providing a localized economic boon through construction and tax revenue.
• Market Dynamics: Proponents argue that increased demand typically attracts new suppliers, which eventually restores price stability and increases overall capacity.

Systemic Bottlenecks and Grid Inefficiency

Industry observers point out that the primary driver of price increases is not the demand itself, but the inability of the grid to scale quickly enough to meet that demand.

Regulatory Barriers

Significant regulatory hurdles often delay the connection of new power plants and interconnects. This prevents the supply side from expanding at the same rate as the demand side, leading to artificial price spikes.

Underutilization of Assets

Grid operators often maintain conservative buffers, leaving many assets (such as long-distance transmission cables) underutilized. Experts suggest that integrating batteries as energy buffers could improve infrastructure utilization, potentially lowering costs by allowing more customers to share the same fixed infrastructure costs.

Broader Externalities of AI Infrastructure

Beyond electricity costs, the rapid expansion of AI data centers introduces several systemic externalities:

• Water Consumption: Thermoelectric power plants (natural gas, nuclear, and coal), which provide the majority of US electricity, require massive amounts of water for cooling. In 2021, the US electric power sector withdrew approximately 47.7 trillion gallons of water.
• Supply Chain Pressure: The surge in demand for GPUs, CPUs, and high-capacity RAM for AI servers can drive up prices for general-purpose computing hardware for all consumers.
• Energy Sourcing: There is a growing call for data centers to build their own dedicated renewable energy generation to avoid placing additional strain on the public grid and to incentivize clean energy investment.