Microsoft president Brad Smith recently called electrical talent shortages the number one problem slowing U.S. data center expansion. He is not exaggerating. According to CNBC, the United States needs approximately 300,000 new electricians over the next decade just to keep up with data center demand, and that is before accounting for the 200,000 current electricians expected to retire. Meanwhile, data center construction spending is projected to surpass $52 billion in 2026 alone, and according to McKinsey, cumulative global data center investment could reach $6.7 trillion by 2030.
The Salary Numbers That Are Reshaping Career Decisions
The economics of skilled electrical work have changed dramatically. According to Fortune, young electricians under 30 are earning between $240,000 and $280,000 annually on data center projects in places like Plano, Texas. Journeyman electricians in Northern Virginia - the epicenter of U.S. data center construction - are earning approximately $59.50 per hour, which translates to roughly $120,000 base salary and up to $200,000 with overtime, according to CNBC.
According to a Randstad workforce report cited by Fortune, construction workers on data center projects average $81,800 annually - 32% more than those working on non-data center construction. Specialist electricians who understand the unique power distribution, redundancy, and cooling requirements of hyperscale data centers command even more.
These numbers are driving a measurable shift. Electrical trade program enrollment surged 400% over four years at Midwest Technical Institute, according to Fortune. Commercial apprenticeship applications nationally jumped 70% from roughly 70,000 to 120,000 between 2022 and 2024.
Why Electricians Are the Real AI Bottleneck
Electrical work represents 45% to 70% of total data center construction costs, according to industry estimates cited by CNBC. A single hyperscale data center requires hundreds of electricians working for 18 to 24 months. The power requirements are staggering - a modern AI training facility can consume as much electricity as a small city.
The constraint is not just headcount. Data center electrical work is specialized. Standard commercial electricians need additional training in high-voltage power distribution, uninterruptible power supplies, generator paralleling, and the specific safety protocols for facilities that cannot tolerate even milliseconds of downtime. The gap between a residential electrician and a data center electrician is wider than the gap between a general practitioner and a surgeon.
Oracle recently pushed a major project completion date from 2027 to 2028 partly due to labor shortages, according to industry reports. When trillion-dollar companies cannot build fast enough because they cannot find enough electricians, the constraint is structural, not cyclical.
How Big Tech Is Responding
The largest technology companies are now directly investing in electrical workforce development. According to CNBC and Fortune, Google committed $15 million to the Electrical Training Alliance. Amazon pledged $700 million in skills training aimed at certifying 15,000 workers by 2026. Microsoft is partnering with community colleges to certify 8,500 technicians per year. Google separately invested $50 million in apprenticeship programs across 12 metropolitan areas.
These investments acknowledge a fundamental reality: the AI revolution depends on physical infrastructure, and physical infrastructure depends on skilled human labor that cannot be automated or outsourced. Nearly 30% of union electricians are between 50 and 70 years old, according to industry data. The retirement wave is coming regardless of how much money tech companies throw at the problem.
What To Do About It
1. If you are in construction, pivot toward data center specialization. Electrical contractors who build expertise in hyperscale power systems will have pricing power for the next decade. Invest in training your crews now.
2. If you are planning a data center build, lock in electrical contractors early. Lead times for qualified electrical crews are stretching to 12-18 months in major markets. Do not assume labor will be available when your permits clear.
3. If you are in workforce development, target this gap. Electrical trade programs focused on data center skills are the fastest path to high-income employment without a four-year degree. The demand is real and growing.
4. If you are evaluating AI infrastructure, factor in construction timelines. The compute you need in 2027 requires construction decisions in 2025. Labor availability is now a first-order constraint on AI infrastructure capacity.
HRIM's Take
The irony is rich. The most advanced technology in human history - artificial intelligence - is bottlenecked by one of the oldest trades. No amount of software can replace the electricians who wire the facilities where AI runs. The companies that figure out the labor pipeline will control the pace of AI infrastructure buildout. Everyone else will be waiting in line. We think this is the most underreported constraint in the entire AI ecosystem, and the salary data proves the market agrees.