AI Data Center Boom Raises Electrician Pay
On July 29, The New York Times reported that AI data center construction is drawing electricians and other skilled trades workers to large projects, including OpenAI's Michigan facility, where hundreds of electricians are working long shifts. PYMNTS, citing the Times and Indeed data, reported that data center installation and maintenance roles pay 42% more per hour than comparable jobs in other fields.
AI data center construction is increasing competition for electricians and other skilled trades workers, with projects offering unusually high pay, bonuses and per diem rates in markets where large facilities are being built. The New York Times reported July 29 that OpenAI's data center project in Saline Township, Michigan, requires hundreds of electricians working 10-hour days, seven days a week.
At the Detroit Electrical Industry Training Center in Warren, Michigan, the Times reported that apprentices are training amid growing demand from nearby data center projects. Tyler Shelton, an electrician in training, told the publication that his employer had sent many workers to a data center that recently broke ground about an hour away: "Everybody wants that money."
Pay competition reaches construction labor markets
PYMNTS, summarizing the Times' reporting, cited an Indeed analysis finding that hourly installation and maintenance positions at data centers pay 42% more than comparable roles in other fields. Its report described bonuses and higher per diem rates as mechanisms employers are using to recruit workers, particularly in rural locations where many large facilities are under construction.
The Times reported that the demand has helped counter weaker hiring in some other construction segments, including housing, while also tightening labor availability for other projects. PYMNTS quoted Aerotek's director of data center market development describing a "really delicate" labor tension as workers weigh opportunities in the data center buildout.
A separate May report from the New York Post described high compensation in certain Texas data center markets. It quoted television host Mike Rowe as saying that electricians he met at a Plano data center were making $260,000 annually. That account is anecdotal rather than a sector-wide wage measure, but it illustrates the premium that can emerge in constrained local labor markets.
Infrastructure constraint beyond GPUs
Data center capacity also depends on construction labor as well as accelerators, power generation, grid interconnection and networking equipment. A facility cannot bring compute online until electrical systems, cooling infrastructure and related physical construction are complete.
Comparable infrastructure booms commonly shift labor toward projects with the strongest compensation, potentially creating bottlenecks for unrelated commercial and residential construction. The reported wage premium is therefore relevant to AI infrastructure timelines and capital costs, even though the sources do not quantify the resulting effect on data center delivery schedules or model-compute pricing.
Key Points
- 1The Times reports OpenAI's Michigan project needs hundreds of electricians, illustrating that AI capacity expansion depends on physical construction labor.
- 2PYMNTS cites Indeed data showing data center installation and maintenance roles pay 42% more per hour than comparable work.
- 3Comparable infrastructure booms can redirect scarce skilled labor, adding construction-cost and delivery-timeline pressure beyond GPU and power constraints.
Scoring Rationale
The reporting identifies skilled electrical labor as a material, often overlooked constraint in AI data center deployment. It does not introduce a new AI product or technical capability, but it is relevant to practitioners tracking the cost and timing of compute-capacity expansion.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
