Data Centers in Rural and Suburban Communities

Executive summary

• Electricity: Data centers can trigger costly transmission and distribution upgrades. A Union of Concerned Scientists (UCS) brief focused on PJM states reports that utilities approved $4.356B in 2024 local transmission projects needed only to connect data centers, with costs typically allocated to customers under current rules.^[1]
• Concrete bill example: An Arkansas Public Service Commission docket includes a "Typical Bill Comparison" indicating an estimated increase of $2.02/month (1.51%) for a typical residential customer using 1,000 kWh/month under a proposed recovery rider.^[2]
• Water: Reporting based on a New York Times investigation describes a planned 33% water rate increase over two years in Newton County, Georgia amid rapid growth in data center water demand and infrastructure needs.^[3]
• Jobs: Construction roles can be substantial but temporary; ongoing staffing is usually modest. ITEP describes incentives for a large Louisiana data center that require 500 full-time jobs by 2035 to qualify for maximum benefits.^[4]
• Subsidies: ProPublica reports Washington State has forgone $474M+ since 2018 under a major data center tax break, and the state could not say how many jobs were created by the break (as of the article date).^[5]
• Balanced view: Berkeley Lab work and related coverage note multiple drivers of electricity prices and caution against assuming data centers are always the primary factor in bill increases.^[6]

1) Electricity impacts

1.1 Why electricity bills are the flashpoint

Large modern data centers can demand 100-400 MW, and some projects request 1,000 MW+. That scale can require new high-voltage transmission lines, substations, and other upgrades that utilities recover through rates.^[1]

1.2 PJM case (relevant to Pennsylvania)

UCS reports that from 2022-2024 utilities initiated 150+ local transmission projects needed only to connect data centers, and that $4.356B in such costs were approved in 2024 across seven PJM states, including $491.8M in Pennsylvania.^[1]

1.3 Example of a quantified "per household" increase

In Arkansas PSC Docket 25-054-U (filed Sept. 5, 2025), Entergy Arkansas provides a "Typical Bill Comparison" for a proposed Strategic Investment Recovery Rider (SIR Rider) with an estimated increase to a typical residential bill of $2.02 (1.51%) assuming 1,000 kWh per month usage, moving from $133.90 to $135.92 in the example table.^[2]

1.4 Virginia policy response

Virginia's SCC approved creation of a new rate class for the biggest users, including data centers.^[8] WHRO reporting states Virginia households will see average monthly bills rise by about $13 over the next two years as part of a biennial review, and frames the new large-user class as a cost-protection measure.^[9]

1.5 Balanced evidence: other drivers of electricity prices

Berkeley Lab analysis and related coverage emphasize many drivers of retail electricity prices (poles and wires, extreme weather hardening, inflation, fuel costs) and caution against attributing price rises primarily to data centers nationwide.^[6]

Some utilities argue large-load growth can reduce average bills by spreading fixed costs. PG&E (utility statement) estimated that every 1 GW of new data center load could reduce customers' bills by 1-2% in the long term. This is a utility estimate, not a regulator finding, but it illustrates how results can differ under different rate designs and cost allocation methods.^[10]

1.6 Transparency risk

Harvard's Electricity Law Initiative argues many large-customer power contracts are confidential and can limit public scrutiny, complicating verification of "no cost shift" claims.^[11]

2) Water impacts

2.1 What data center water use can look like

Data centers may use water directly for cooling and indirectly through the water footprint of electricity generation. The Environmental and Energy Study Institute (EESI) summarizes an estimate that U.S. data centers consume about 449 million gallons of water per day (as of 2021).^[12]

2.2 Newton County, Georgia: planned rate increases and infrastructure pressure

Reporting based on a New York Times investigation describes residents near a Meta data center in Newton County, Georgia experiencing worsening well conditions, and local leaders planning major infrastructure upgrades. The reporting states water rates were set to increase 33% over the next two years to fund upgrades and avoid rationing.^[3]

2.3 Great Lakes region: emerging scrutiny

The Guardian reports growing controversy around data center water demand in parts of the Great Lakes region, including concerns about water levels and local impacts as more projects target these areas for cooling advantages.^[13]

2.4 Mitigation options to verify locally

• Cooling design: air cooling, closed-loop, or alternative methods may reduce direct withdrawal but can affect electricity demand.
• Non-potable water: reclaimed wastewater cooling can reduce use of drinking water where capacity exists.
• Reporting: absent mandatory reporting, communities struggle to compare projects and enforce commitments.^[12]

3) Jobs and workforce

3.1 The common pattern

Construction employment can be substantial and short-term. Ongoing operations typically require fewer full-time workers than comparably sized industrial facilities because most activity is automated.

3.2 Louisiana example: incentive requirements and long-term staffing

ITEP notes Meta's Louisiana incentives require 500 full-time jobs by 2035 to receive maximum tax breaks, and emphasizes construction jobs largely disappear after completion.^[4]

Reuters reports the Louisiana project is expected to create 500+ jobs (Reuters language) while supporting AI compute capacity.^[14] Wired reporting highlights debate around who pays for long-lived infrastructure if the customer reduces demand later.^[15]

3.3 Cedar Rapids, Iowa: incentives tied to small headcount

The Gazette (Cedar Rapids) described a proposed $750 million data campus seeking city incentives with a minimum of 30 new full-time jobs (as described in the article).^[16]

4) Subsidies and fiscal impacts

4.1 Why subsidies matter to residents

Even if a data center pays a negotiated power rate, communities can still bear costs via tax exemptions, infrastructure spending, and long-lived grid or water expansions. Evaluating "final cost" to residents typically requires looking at utilities + taxes + infrastructure debt, not only the facility's own utility bill.

4.2 Washington State: large tax break, limited job accounting

ProPublica reports Washington State has forgone $474 million+ since 2018 under a major data center tax break, and the state could not say how many jobs were created by the tax break (as of the article).^[5]

4.3 Louisiana: incentives plus long-lived infrastructure debate

Wired describes Louisiana actions to support the Meta project, including expedited power infrastructure approvals and incentives, while noting concern that some long-lived infrastructure costs could fall on utility customers without sufficient protections.^[15]

4.4 Policy trend: making large users pay marginal costs

Oregon passed a law intended to require data centers to pay for the costs of energy and transmission infrastructure needed to power them; subsequent reporting describes disputes about implementation.^[17]

5) What communities can ask for (practical, documentable items)

Electricity and ratepayer protection

• Who pays for interconnection and upgrades? Require a written statement (tariff or commission order) that assigns customer-specific upgrades to the data center or a new large-load class, rather than spreading by default.^[1]
• Minimum demand charges and long-term commitments: reduce risk of stranded assets if the customer reduces usage later.
• Contract transparency: insist on public summaries and independent review where confidentiality applies.^[11]
• Separate rate class: Virginia provides a current example of a regulatory tool aimed at isolating costs and risk.^[8]

Water system protection

• Water source clarity: potable vs non-potable, groundwater vs surface water, and seasonal constraints.
• Cooling disclosure: cooling method and projected annual and peak-day consumption.
• Infrastructure funding plan: specify which costs are paid by developer vs financed via rates and bonds, and expected rate impacts.^[3]
• Mandatory reporting: annual water use reporting to the public.^[12]

Jobs and economic claims

• Separate construction vs operations headcount and publish enforceable minimums for permanent roles.^[16]
• Local hiring commitments tied to incentives, and training partnerships.
• Clawbacks if job or investment commitments are not met.

Sources (links)

Numbered citations correspond to the sources below. Some sources may be behind paywalls; they are included for completeness.