Data Centers Shift Grid Upgrade Costs to Low-Income Residents
Cornell researchers analyzing power grids from 2020 to 2025 across four major U.S. regions found that data centers paid only 38% of infrastructure costs required for their growth, with residents absorbing the remaining 62% ($5.13 billion of an $8.25 billion 2025 bill). Lowest-income households faced energy cost increases up to 28% of their incomes. This disparity reflects how current cost-allocation rules—based on historical consumption patterns—don’t account for who actually drives new grid demand.
Data center power consumption is accelerating sharply: transmission infrastructure investment rose from $2 billion in 2020 to $34 billion in 2025. These facilities are projected to consume 8-17% of U.S. electricity by 2030, with 70-80% of that attributed to artificial intelligence. While each gigawatt of capacity does create approximately 15,000 jobs across the economy, economic benefits flow unevenly—California saw $180 in worker income returns per dollar of state incentive spending, while Virginia saw only $2.
The cost burden falls hardest on rural and lower-income communities. Rural households’ share of high energy burdens rose 76% more than urban households. Costs spill across state lines, with New Jersey, Pennsylvania, and Ohio residents helping fund Virginia’s data center infrastructure. Researchers recommend allowing data centers to reduce consumption during peak demand hours, charging them directly for grid expansion they trigger, establishing state programs to offset costs for low-income households, and tying tax incentives to job creation and wage growth.
Read the full article
Continue reading at Hacker News →This is an AI-generated summary. Read the original for the full story.