1,584 Data Centers Studied: The Real Job Numbers Behind the Boom

Data centers opened across the United States from 2000 to 2023 created 1,000 to 1,300 nonconstruction jobs in their host counties, on average, according to a new study of 1,584 facilities. The catch: those jobs sit outside the data center walls, arrive three to five years after opening, and cluster in industries that serve the facilities—not in restaurants or shops.

The research, conducted by six academics including Simon Buechler, assistant professor of finance at Miami University, and Lyndsey Rolheiser, professor of urban and real estate economics at York University, Canada, lands amid escalating fights over data center construction. Local opposition contributed to delays or cancellations of $156 billion in proposed data center projects in 2025 alone, from Ohio to Texas.

The question the authors set out to answer sits at the center of those fights: Do data centers actually create jobs where they are built?

What did the researchers measure?

The team assembled records on 1,584 U.S. data centers opened between 2000 and 2023. For each facility, they recorded location, opening date, and power capacity—a standard yardstick for comparing facilities of different sizes. They then matched those records with federal data on employment, wages, and business counts in each county.

Isolating the effect was not straightforward. Operators do not pick locations randomly. They seek reliable electricity, fiber connections, available land, and friendly regulations. Those places may already be growing, so new jobs might have arrived anyway.

To control for this, the researchers tracked employment in counties that received data centers—Loudoun County in Virginia, Berkeley County in South Carolina, and Morrow County in Oregon among them—before and after each facility opened. They then compared those trends against similar counties without data centers, and tested whether larger facilities produced bigger gains.

How quickly do the jobs appear?

Slowly. During the first two years after a county's first data center opened, the researchers found little employment change outside construction. After that, jobs begin to grow. Three to five years out, host counties carried roughly 1,000 to 1,300 additional nonconstruction jobs compared with similar counties that received no data center.

Scale matters. Five years after new capacity is installed, each additional megawatt is associated with about 7.4 more nonconstruction local jobs.

Part of that gain reflects follow-on expansion. Each megawatt installed is associated with another 1.7 megawatts of capacity added in the following years. Counting that later growth, the estimates work out to about 2.7 nonconstruction jobs per megawatt of eventual capacity. A 100-megawatt facility, for instance, correlates with roughly 275 additional local jobs outside the data center itself.

Where are the jobs, and where are they not?

The composition explains the mechanism. Employment gains concentrate in three areas:

• Transportation and warehousing
• Professional and technical services
• Real estate

These industries serve the data centers directly or benefit from the investment and infrastructure the facilities bring.

The researchers found little evidence of job growth in restaurants, stores, and other local services. The spending by data center employees and supplier workers that boosters often predict did not register in the data.

Expansion also follows a pattern. Companies already operating in a county tend to expand their facilities there, because the electricity, fiber, and other infrastructure are already in place. Other operators generally do not move into the same area to build new data centers.

What does the study leave unanswered?

The authors flag clear limits. The dataset covers facilities opened through 2023—before the current wave of massive AI-driven construction. Larger data centers do not necessarily create proportionally more jobs, so the findings may not apply to today's biggest projects.

The study also measures job gains only. It does not weigh them against the costs that worry communities: large consumption of electricity and water, noise, and substantial tax breaks from local and state governments, several of which states have begun scaling back.

The authors' bottom line is measured. Data centers do create jobs beyond the few dozen technicians, engineers, and security guards who work inside their walls. Whether those additional jobs justify the costs remains an open question—one that counties weighing the next $156 billion in proposed projects will have to answer for themselves.