Guest opinion column by Eirini Pajak of Pinal Unlocked, published in Pinal Post’s Opinion section. Views are the author’s own.
LA OSA DATA SERIES
Part 5 of a Series | Case No. PZ-003-26 and PZ-PD-003-26
At the April 16 Planning and Zoning hearing on La Osa, Planning and Zoning Commission Chairman Robert Klob looked at the employment figures presented by the applicant’s attorney and said: “This would be an awesome thing, you know, potentially 30,000 employees. That’s a great thing for our county. But if it, if the reality turns out it’s, it’s, you know, 500 for the entire site, well, we got sold a bill of goods.”
Quick Look
- The employment estimates presented at the hearing were projections with no legal weight.
- The applicant’s per-building estimate does not seem to be supported by industry data. Court Rich, the attorney representing Vermaland at the April 16 hearing, estimated 100 to 500 permanent employees per each of the 59 buildings, implying between 5,900 and 29,500 permanent jobs at full build-out. Chairman Klob cited national data showing 5 to 10 per building. The independent sources reviewed for this article generally align more closely with Klob’s skepticism.
- The 200 to 500 figure cited by the Arizona Commerce Authority applied to hyperscaler campuses specifically, not data centers generally. It described staffing for three to six facilities at a time.
- Vermaland, the developer and applicant, is a land developer, not a data center operator. The La Osa application confirms that no data center operators or tenants are identified anywhere in the public record.
- Industry-sponsored employment estimates presented at public hearings overstate actual impact by a factor of three, according to a May 2026 Brookings Institution study.
- Vermaland’s own Traffic Circulation Analysis implies a small permanent workforce. The trip generation assumptions in Vermaland’s own engineering documents are consistent with a facility employing hundreds of workers, not thousands.
- A discrepancy between the initial figures announced and the actual outcomes is documented across multiple states. Indiana and Illinois both show differences between what communities are told publicly and what developers are legally required to deliver.
The employment projections and the fiscal projections are the two central justifications for this rezoning. Neither is legally binding on anyone. The fiscal figures are examined in Article 1 of this series.
What Is Not Enforceable
Vermaland is a land developer. Its press release describes it as “a Phoenix-based land development company specializing in solar farms, industrial development and data center infrastructure.” The La Osa application and hearing testimony confirm that Vermaland is not itself a data center operator. The county is handing out an industrial rezoning of nearly 2,400 acres based in part on employment projections made by a developer on behalf of future occupants who do not yet exist in the formal record.
Supporters of large-scale data center projects argue that permanent on-site staffing is only one measure of economic impact, pointing to construction employment, long-term tax base growth, and secondary economic activity. The Arizona Commerce Authority made this case at the April 15 Board of Supervisors special session, and the Arizona Technology Council has made similar arguments in county proceedings. That some economic benefit could occur is not the question. Whether the employment expectations discussed publicly will actually pan out is.
Vermaland can project thousands of jobs, but nothing in this approval, as structured, requires them to deliver any.
What Was Said at the Hearings
Arizona Commerce Authority Numbers
According to the Brookings Institution, there are two distinct types of data centers. A hyperscaler is a single major cloud or AI company that builds and operates its own campus for its own workloads. A colocation facility is a building where a landlord leases server space to multiple tenants.
At the April 15 Board of Supervisors session, Ryan Reece of the Arizona Commerce Authority provided general industry context. On permanent employment, he said: “we’re really talking anywhere between 200 and 500 full-time employees for these data center campuses,” but explicitly qualified that figure as applying to hyperscaler and AI-focused campuses with three to six facilities, as opposed to colocation-style facilities. He also cited a 3-to-1 economic multiplier, meaning that for every direct job created at the data center, three additional indirect jobs are created in the surrounding economy. (Brookings found that industry-sponsored multipliers of this kind overstate actual impact by a factor of three.)
Reece also drew the colocation distinction himself, in response to a direct question from a supervisor: “hyperscaler versus like a co-located facility. I’d, I would say like think of it as a shopping center, right? When every kinda cubby is a different company, the permanent employees there are not likely to be many, and the quote-unquote ‘permanent’ employees that are there are likely contracted, which we don’t necessarily count as employees, ’cause they’re contracted through a third party that was likely already here.” The Arizona Commerce Authority’s own representative told the Board that colocation facilities produce few permanent employees and that contracted workers already in the region are not counted as new jobs.
Extrapolation of the Employment Figures
The following day at the formal Planning and Zoning hearing, Court Rich of Rose Law Group, representing the applicant, offered an estimate of 100 to 500 permanent employees per building, applied to each of the 59, in response to a question from Vice-Chairman Mooney. Commissioner Tom Scott followed up, confirming the figure and doing the math on the record: “on the low side you’re gonna have 5,900 employees…” Rich’s response was: “And we’re excited about that, I think that’s a good opportunity.”
Chairman Klob ran his own calculation. At 100 to 500 per building across 59 buildings, he noted, the campus implies between 5,900 and 29,500 permanent employees.
In response to Klob’s challenge, Rich reached back to Reece’s campus-level figure from the day before to support his numbers: “I’m not here telling you that you should do this for the jobs, although I think that’s a great piece of it. Yesterday at the Board of Supervisors, when they talked about jobs, the gentleman from the Arizona Commerce Authority said that for a campus of six data centers, they would expect between 250 and 500. So if we use that math here and every six buildings has something like that, we’re certainly talking thousands of jobs based on all the different metrics that I’ve seen.” He acknowledged he was not promising the top number: “I don’t stand here telling you this is gonna be 30,000 jobs. Those are the ranges that I’ve heard.” However, Reece from the ACA’s figure described staffing for a hyperscaler campus of three to six buildings total, not a per-building multiplier that scales across 59 buildings of unknown operator type.
Vermaland’s own Traffic Circulation Analysis, prepared by EPS Group, projects 1,396 AM peak hour trips for the entire 26.55 million square foot campus. Those assumptions are difficult to reconcile with a facility employing thousands of workers. If the actual permanent headcount is in the hundreds as independent research suggests, the economic development case presented at the hearing does not hold.
Who Actually Gets the Jobs
Klob also raised a labor market concern during his questioning of Court Rich, citing a conversation he had with a Wyoming developer about a month before the hearing: that data center construction pulls electricians, plumbers, and concrete workers away from local builders at wages they cannot match, leaving existing projects understaffed. He asked what commitments were being made on workforce training with community colleges. Rich said the Arizona Commerce Authority was in discussions on that front.
That concern is documented nationally. According to The Birmingham Group, a construction recruitment firm that tracks the data center labor market, when multiple large campuses break ground simultaneously in a region like Phoenix, local talent pools are exhausted within months and firms must import workers from other states. CRE Daily reported that senior project managers are now being hired remotely and flown to job sites, and that workers are relocating from other states to take data center construction roles in the Phoenix market.
Good Jobs First found that permanent data center jobs are not uniformly high-skill or high-wage: in Wyoming, 35 percent of data center workers were in unskilled roles, and a developer application in North Carolina showed that out of 69 permanent positions, 25 were security and janitorial staff. Inclusion Cloud, a staffing firm that places remote technical workers and has a commercial interest in that model, found that many specialized operational roles can support multiple facilities remotely without requiring physical presence at the site. The job count announced at a public hearing rarely specifies how many positions will be filled locally, how many will be remote, and how many will be low-wage support roles.
What Independent Research Shows
The independent sources reviewed for this article generally align with Klob’s skepticism. These facilities are not identical to La Osa in scale or configuration; the data nonetheless points consistently in one direction.
| Source | Finding |
|---|---|
| Brookings Institution (May 2026) | Counties receiving a first large data center see county-level employment gains of roughly 2,000 to 4,000 additional jobs after six years (in a typical county with roughly 98,000 workers), with the range depending on facility type. Hyperscale counties see significant information sector job growth. Colocation counties do not see those gains. These are jobs at other employers across the county: construction firms, IT contractors, support businesses, not jobs inside the data center itself. |
| Hamm Institute for American Energy (Nov. 2025) | Hyperscale campuses above 100 MW operate with as few as 18 to 30 permanent staff per 100 MW across low to high automation scenarios. Applied to a 3,000 MW campus: roughly 540 to 900 permanent staff at complete build-out. |
| Virginia JLARC (Dec. 2024) | At peak construction, approximately 1,500 workers are on site. A typical 250,000-square-foot data center employs about 50 full-time workers once operational, roughly half of them contractors. Economic benefits are concentrated in the construction phase, not operations. |
| Latitude Media (May 2026) | A Vantage Data Centers facility outside Reno: 73 permanent positions for a 1.1-million-square-foot site. Meta’s $10-billion Lebanon, Indiana campus: roughly 300 people at completion. Amazon’s $35-billion Virginia announcement: 1,000 permanent positions phased across 17 years. |
| Data Center Knowledge | Total on-site staff at major facilities typically numbers only several dozen. A Microsoft UK data center employed 101 people. An Amazon Mississippi AI campus projected roughly 200 direct roles. |
| Food and Water Watch (Jan. 2026) | Virginia holds 33% of national data center capacity but only 4% of sector employment. As few as 23,000 people worked across approximately 4,000 data centers nationally in 2024, averaging about 6 per facility. |
What Brookings Found About Facility Type
A May 2026 Brookings Institution paper by economists Dany Bahar and Greg Wright, drawing on approximately 770 U.S. data center facilities linked to county-level employment records from 2003 to 2024, distinguishes between hyperscale facilities, built and operated by companies like Amazon, Google, or Microsoft to run their own workloads, and colocation facilities, built by landlords who lease server space to remote tenants. Hyperscale counties see significant information sector job growth. Colocation counties do not.
La Osa does not map cleanly onto either Brookings category. Vermaland plans to sell or lease parcels to future operators whose identity is unknown. If those operators are hyperscalers, some larger employment spillovers may materialize for that portion of the site. But even under that scenario, the 200-500 employee figure Reece cited would apply to a campus of several facilities, not to each of the 59.
Hyperscale facilities attract investment regardless of subsidies; colocation facilities depend on them. Brookings found that state incentives represent 2 percent of construction investment in hyperscale counties and 62 percent in colocation counties, meaning public subsidy flows to the facility type that generates the least employment return. Industry-sponsored employment estimates overstate actual impact by a factor of three. Whether the construction activity translates into a permanent tax base is examined in Article 1 of this series.
The Gap Between the Projection and the Outcome
The gap between what gets announced publicly and what gets written into binding agreements is documented. In Indiana, the governor announced a Google data center by citing 200 permanent jobs; according to Good Jobs First, which tracks subsidy accountability nationally, the binding local property tax abatement agreement required 30. In Illinois, the state’s data center sales tax exemption requires participating companies to create just 20 positions, and Good Jobs First found that recipient disclosure data shows many facilities employing fewer than 50 workers. La Osa has no comparable agreement.
The Opportunity Zone Structure
The La Osa site carries a Federal Qualified Opportunity Zone designation. The investor tax benefit is captured by holding the investment for the qualifying period. It does not require completing all 59 buildings, finding tenants for all parcels, or creating a specific number of jobs. An investor who begins construction and holds the investment long enough captures the full capital gains benefit regardless of how much of the campus actually gets built or how many people end up working there.
Unanswered Questions
- Definition of permanent job: Direct employees or contractors? Full-time or part-time? On-site or remote? High-skill technical roles or security and janitorial staff? The figure presented at the April 16 hearing was 100 to 500 permanent employees per each of the 59 buildings.
- Workforce training: Chairman Klob asked about community college partnerships at the April 16 hearing. Court Rich said discussions were underway. Nothing was added to the record after the P&Z hearing, and no workforce training commitment has been made part of the public record.
- Local hiring: What share of construction and permanent operational jobs is realistically expected to be filled by Pinal County residents? The Phoenix data center market is already pulling skilled workers from out of state, and permanent operational roles typically require specialized skills.
The Pinal County Board of Supervisors public hearing on this case is scheduled for May 27, 2026 at 9:30 a.m., 135 N. Pinal Street, Florence. Agenda packet.
Disclosure
This article is based on publicly available documents, hearing transcripts, and third-party research cited throughout. It is not legal, financial, or economic advice. Figures should be independently verified before being relied upon for any official purpose.
About the Author
Eirini Pajak is a licensed real estate agent and Pinal County resident. She covers local land use and development decisions through her Pinal Unlocked page on Facebook and runs the Pinal Code Watchers community group. Her dog Peso joins her on county rounds.
