The promise of new jobs often accompanies significant industrial projects, yet the reality frequently diverges, with the anticipated employment opportunities failing to materialize.
The town of Jay, Maine, a rural community situated approximately 67 miles northwest of Portland, once thrived around the Androscoggin paper mill. At its zenith, the mill supported around 1,500 employees until a pulp digester explosion in 2020 led to its permanent closure.
In 2023, the vast 1.4 million-square-foot facility was acquired through a joint venture involving JGT2 Redevelopment and other holding and capital companies, spearheaded by developer Tony McDonald. Over the subsequent three years, McDonald and his team systematically dismantled the mill's machinery, shipping it to Pakistan, and undertook extensive cleanup of the industrial site in preparation for resale. This resale agreement was finalized earlier this year, according to McDonald, positioning Jay at the forefront of a growing national debate surrounding the proliferation of large-scale data centers.
Maine presents an attractive proposition for data center developers due to its consistently cool year-round temperatures, accommodating land-use regulations, and an impressive 54 percent renewable energy mix, ranking it eighth highest nationally. The state currently has several planned data centers, prompting the legislature to pass a bill for an 18-month moratorium on permits and construction for any proposed data center consuming over 20 megawatts of power. Lawmakers aimed to pause development to thoroughly assess the impact of data centers on local economies, the power grid, and the environment.
However, this legislative effort, which would have been the first of its kind in the country, was vetoed last month by Maine Gov. Janet Mills. Her primary justification for the veto was the potential for job creation. She argued that a proposed $550 million data center at the former Jay paper mill site would generate 125 to 150 permanent, high-paying positions for a town that had recently lost its largest employer.
Across the nation, from former mill towns in Maine to agricultural counties in Indiana and desert expanses near Abilene, Texas, data center developers are pitching local governments with a consistent message: provide the necessary infrastructure, grant tax incentives, and jobs will inevitably follow. Over 35 states have responded by offering a range of incentives to attract this burgeoning industry.
Despite these promises, there is limited research confirming that such massive industrial sites consistently deliver the long-term economic benefits they tout, with early reports often suggesting otherwise. Experts highlight that rural communities frequently lack the governmental expertise required to adequately evaluate the potential impact of data centers. Recent Pew Research Center data indicates that 67 percent of planned data centers in the U.S. are slated for rural areas, with 39 percent heading to counties that currently host none. As data center development accelerates, it's becoming apparent that what rural communities are often receiving is not a significant boost in permanent jobs, but rather a power- and water-intensive industrial facility that temporarily employs a workforce comparable in size to a midsize restaurant.
Tony McDonald's initial vision for the mill involved selling it to Godfrey Forest Products, an oriented strand board company that would have employed approximately 150 individuals. However, federal tariffs disrupted the project's financial backing, leading McDonald to explore alternative proposals he had been receiving.
“Most of the people that were contacting us, you know, they were all hat and no cattle,” McDonald recalled, describing numerous calls from individuals he termed “data center cowboys.” These contacts often claimed to represent major tech companies and sought sites for new data centers, but deeper investigation frequently revealed a lack of credible backing.
Following several such encounters, McDonald began discussions for a data center partnership with Sentinel Data Centers, a New York-based firm specializing in facilities for the healthcare, financial, and hyperscale industries. Sentinel Data Centers did not respond to multiple requests for comment for this story.
McDonald explained that, as he understands it, the Jay project will be a neocloud data center, a specialized facility designed for high-performance GPU computing to support AI and machine learning workloads. Such a facility, if installed in the old mill, would demand over 100kW of energy per rack, adhering to industry standards, and would necessitate advanced cooling methods like direct-to-chip or immersion cooling, both of which require substantial space and water resources.
“Most of the people that were contacting us, you know, they were all hat and no cattle.”
While McDonald has consistently stated his disinterest in immediate tax breaks for the project, he has left open the possibility for future tax benefits should the town deem them appropriate. Coincidentally, just before Governor Mills vetoed the moratorium, the Maine state legislature passed a law that generally excludes data centers from some of Maine’s standard business tax breaks. However, it preserves the authority of local municipalities to negotiate and offer their own tax break agreements and other municipal incentives. This provision could pose challenges for smaller communities like Jay, which might anticipate significant tax revenue to invest in schools, public infrastructure, and community facilities.
Maine State Representative Melanie Sachs, sponsor of the state’s moratorium bill, asserted that McDonald did not inform the Jay Select Board of the new data center plan until late February 2026, merely days before her bill was scheduled for a floor vote, despite its initial introduction to committee on January 30th.
The Select Board reviewed McDonald’s presentation in March and voted 4-0 in favor, as reported by The Maine Monitor. The moratorium bill successfully passed both the state House of Representatives and Senate on April 14th, only to be vetoed by Governor Mills on April 24th, who cited job creation in Jay as her primary justification.
Sachs, who chairs Maine’s House Energy, Utilities and Technology Committee, clarified that her legislation was not intended to ban data centers. “This bill was about creating the playbook,” she stated. “And we were told, ‘Don’t worry your pretty little heads about it, data centers are not coming to Maine anytime soon.’ They came anyway, and without a framework in place, towns have no mechanism to evaluate the claims developers are making.”
“Even if it’s 30 jobs, that means a lot to Jay, then, okay, but you’ve swept away protections for 1.4 million Mainers for 30 jobs,” Sachs emphasized, highlighting the perceived imbalance.
Michael Hicks, director of the Center for Business and Economic Research at Ball State University, is an economist and professor who last November published one of the first causal analyses of data center employment effects in the United States. His research involved studying data center openings across 254 Texas counties to measure their actual impact on long-term local employment.
His findings revealed that net job creation was effectively negligible, with any long-term jobs within data centers being offset by losses elsewhere in the same sector.
“As you drive by a data center, you see people working on it,” Hicks noted. “You see construction workers. The hotels locally are packed. But there’s no net pulse of that. A lot of these workers are there for three weeks to do their part of it, and then they’re gone … The real question is whether there are permanent jobs associated with data centers, and in Texas, the answer is no.”
“A lot of these workers are there for three weeks to do their part of it, and then they’re gone.”
Texas serves as an ideal case study due to its distinct power grid and a diverse landscape encompassing large, rapidly growing metropolitan areas, tech hubs, and remote rural towns, mirroring the broader national context. As Hicks explained, it functions as a “mini-United States” in its regional variety, making the economic lessons learned there broadly applicable across the country.
Anthony Elmo, a public education funding defender at Good Jobs First, a nonprofit research center focused on corporate and government accountability in economic development, observed that rural towns are frequently “outgunned” when attempting to negotiate agreements with major data center developers.
“They don’t have the resources to negotiate,” Elmo stated. “They don’t know what to ask for. They don’t have the legal expertise, and they don’t feel like they have the leverage, which I think is part of the issue.”
The national picture is equally challenging. Business Insider recently reported that the national subsidy for data centers surpassed $2 million in costs per permanent job. In some extreme instances, such as one in New York, companies received nearly $77 million in tax breaks for a facility that generated only one permanent position.
Microsoft’s facility in Quincy, Washington, which is roughly comparable to the initial plans for Jay, employed up to 500 workers during its construction phase but now operates with merely 50 full-time employees. The specific type of data center being developed significantly influences the long-term employment figures for a community. Neocloud data centers, like the one proposed for Jay, typically require between 30 and 50 full-time staff, depending on their scale.
“Of those, say, 50 jobs, a chunk of them are maintenance, a chunk of them are technicians in charge of backup generators. The high-tech jobs make up maybe 10 percent of the facility,” Elmo highlighted. Furthermore, data center companies often include remote workers located in other states as employees of the state where the data center is situated. “We may get a little bit of an economic effect from that, but it isn’t nearly as much as if it were a physical person in Maine buying goods and adding to the local economy,” he added.
Much of the lobbying surrounding data centers emphasizes job creation and “upskilling,” or training workers for new, better-paying roles. However, researchers suggest that even the retraining argument lacks substantial support.
Recent census data for Jay indicates that just under 30 percent of its population holds a bachelor’s degree or higher, while 90 percent possess a high school diploma. As Ball State University’s Hicks noted, education plays a crucial role in the success of upskilling initiatives.
Historically, industrial revolutions, such as the entry of farm laborers into Midwest factories in the 1800s, or the post-WWII workforce re-entering civilian production, or increased workforce participation in the post-Civil Rights South, succeeded due to a foundational level of education or acquired skills that made workers trainable for new roles.
Hicks pointed out that these past industrial transformations were successful because they coincided with a significant influx of educated individuals ready to fill new positions. In contrast, the U.S. currently faces negative net immigration, low birth rates, and consistently underfunded education. He argued that there is no comparable "human-capital wave" to sustain a similar jobs boom driven by data centers and AI.
“The waves of industrialization accompanied waves of human capital into the United States,” Hicks explained. “So, where do we think there’s this surge of employment surrounding data centers that can mimic those three events?”
“The big tech companies are investing in upskilling the construction trades,” Elmo observed, citing OpenAI’s recent agreement with NABTU. He noted that in established data center markets, electricians and HVAC maintenance workers typically move from project to project as contractors. “For states that don’t have developed data center infrastructure, like Maine, it’s not a permanent job. It’s an 18-month job. That’s it.”
Hicks contended that the promise of jobs often serves as a distraction from the single reliable benefit a data center can offer a rural community: substantial tax revenue.
Using the Jay data center as an illustration, Hicks calculated that a $550 million data center in a town of 4,620 residents, where the median home value is approximately $215,000, would possess an assessed value exceeding the combined worth of every home and business in the community. The former mill, despite a tax abatement, generated roughly $1.8 million in tax revenue for Jay in its final year of operation, according to the Livermore Falls Advertiser. If taxed at the same rate as other commercial properties, such revenue could fund schools, rebuild infrastructure, and attract new residents for generations.
“You could make that town into a Hallmark Channel town with those sorts of tax dollars, and then jobs would follow,” Hicks suggested, but cautioned that this outcome hinges on whether the town opts to grant the future project special tax breaks.
Rural towns frequently find themselves at a disadvantage when negotiating agreements with large data center developers.
As of this reporting, the precise amount of tax revenue the data center could generate remains unclear, owing to numerous unresolved factors. These include the identity of Sentinel’s potential clients, the exact type of data center to be constructed, the nature of any tax incentives the town’s Select Board might offer, and even the definitive number of jobs the data center could bring to Jay.
McDonald’s job estimates have fluctuated throughout the project’s progression, ranging from 100 to 150 positions. He stated that the figures provided to the Select Board and the state legislature are based on information from Sentinel, but he also acknowledged, “I’m not a data center guy.”
Ultimately, this situation presents a technological ouroboros. The very data center infrastructure, often subsidized under the guise of job creation, is inherently designed to minimize human labor, and the AI it powers is explicitly developed to automate work. Communities are, in essence, being asked to trade vital tax revenue and grid capacity for jobs in an industry whose core output is labor replacement.
“It’s the biggest capital expenditure since the Manhattan Project, and it isn’t going to create tens of thousands of jobs in the long term,” Elmo asserted. “It’s not some economic boom. Meta, Amazon, OpenAI, Oracle, they’re shedding jobs in real time while spending billions on data centers so that other organizations can shed additional jobs through AI. At some point, people need to ask more critical questions about this.”
As Hicks highlighted, rural America is consistently being presented with the same unfulfilled promises it has heard for half a century. He argued that “civic leaders are living 75 to 150 years in the past,” framing
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.