Pima County, Arizona

Pima County has proven it will approve major data center investment, but the political and regulatory environment has hardened noticeably since its first major project was approved in 2025. A new entrant today is entering a market that is actively rewriting its rules, has paused its primary state tax incentive, and has already produced a cautionary precedent in an adjacent jurisdiction involving the same class of project.

None of this makes the market unviable. It does mean that success here depends more on public engagement strategy and timeline flexibility than on regulatory certainty or financial incentives. The sections below walk through the evidence, module by module, with the underlying sources named throughout.

The Board of Supervisors currently has no stable working majority on data center questions. It approved the county's first major project 3-2, then voted 3-2 for a different outcome, moving toward restriction, roughly a year later. District 5 is the seat to watch: it is currently held by appointment, with a 2026 special election that will likely determine the board's near-term posture more than any other single factor.

Both the City of Tucson and Pima County have moved to tighten rules since 2025. Tucson's new zoning ordinance, effective September 2026, requires discretionary rezoning for any large facility, with no by-right approval path anywhere in the city. The county has authorized drafting a moratorium on new approvals and is rewriting its zoning code, a process expected to conclude with a permanent ordinance vote this fall.

Data centers have become their own standing local media beat in this market, not a story tied to any single project. AZ Luminaria, Tucson Sentinel and Arizona Public Media (AZPM) are all producing sustained, recurring coverage of data center policy, public meetings and community organizing, at a level of process detail unusual for routine zoning matters. City officials cited 980 resident comments on a single proposed code amendment.

Coverage regularly names the No Desert Data Center Coalition as an active regional organizer, described in reporting as having traveled nationally to share its organizing playbook and expanded into community teach-ins. Its activity spans both Pima County and the City of Tucson; a new entrant anywhere in the region should expect to encounter it, not treat it as specific to one project.

Of the outlets reviewed, the two most-engaged on this topic (AZ Luminaria and Tucson Sentinel) lean critical in framing, while AZPM's process-heavy coverage is the closest to neutral. No outlet reviewed has produced sustained, independent pro-development coverage, a structural gap in the market's information environment rather than one project's public-relations problem.

Initial power capacity is achievable under Tucson Electric Power's established "growth pays for growth" model: full commercial rate, no subsidy, with the customer funding any capacity expansion beyond an initial allocation. Arizona's regulatory framework for large-load customers is still being built in real time at the state and federal level.

Reflects TEP-attributed owned generating assets in EIA's plant-level filings, not TEP's full contracted portfolio. TEP has signed substantial solar-plus-storage power purchase agreements in recent years that do not appear as owned nameplate capacity in this dataset; a full capacity picture for a specific project should confirm TEP's current contracted position directly.

On water, the City of Tucson's own supply position is resilient even under worst-case Colorado River scenarios, but that security does not automatically extend to unincorporated county land, where groundwater dynamics differ and site-specific verification is essential.

Local building trades are a genuine ally. IBEW Local 570 and SMART Local 359 have supported data center development both locally and nationally, and a 100% union project labor agreement precedent already exists in this market from the county's first major project. Peak construction demand for a large project can still exceed local union membership capacity, so labor import or accelerated training investment should be part of any serious plan.

Highway/street/bridge figure is drawn from a thin base (17 reporting establishments countywide) and should be treated as directional, not a reliable trade benchmark, until corroborated by a larger sample or occupation-level OEWS data.

Read together, the trade-specific data confirms the qualitative finding from the initial pilot: Pima County's construction labor cost sits meaningfully below the Arizona state average across the categories with reliable sample sizes, a real cost advantage for a large build. The contractor base itself (specific firms with the bonding capacity and headcount for a hyperscale build) has not yet been mapped and remains a dedicated research item.

Two points matter most here. First, Arizona's primary data center sales tax exemption is currently paused: a three-year moratorium on new applications took effect in July 2026 as part of a June 2026 state budget deal, running through June 2029. A new project applying today would not qualify. Second, property tax revenue from any project is distributed automatically by formula across county, school, fire and library jurisdictions, not directed to one controllable public benefit pool, which matters for how economic impact gets communicated to a community. The county itself only controls roughly 38% of the property tax it collects.

GPLET, the local incentive tool, is designed around urban redevelopment and does not fit industrial or logistics-zoned parcels particularly well. Given how negatively incentives are perceived in this market's public discourse, not seeking one at all could function as a genuine differentiator rather than a missed opportunity.

This market has an established precedent of expecting data center projects to actively demonstrate alignment with local climate goals, both the city's "Tucson Resilient Together" plan and the county's CAPCO framework, rather than treat them as background policy. The county's first major project tied its energy sourcing to these plans in official materials and shifted its cooling design from water-cooled to air-cooled in direct response to public pressure.

Extreme heat and drought dominate local hazard concerns. Flood risk is real but localized and requires parcel-specific floodplain verification; wildfire risk is more relevant to foothills and wildland-urban interface areas than industrial-zoned land.

The most instructive comparable sits in an adjacent Arizona jurisdiction, involving the same developer active in Pima County: a council-approved project in Marana faced a well-organized referendum challenge backed by a national labor and advocacy organization, ultimately resolved only through litigation, including the developer's own subsidiary suing the town. Elsewhere in Arizona, Chandler's city council rejected a data center proposal outright, 7-0, and Cochise County's own moratorium attempt failed at the board level on the same evidentiary standard (A.R.S. §11-833) that Pima County's moratorium effort will need to clear.

Pima County actively master-plans and markets large industrial sites suited to this use, including the roughly 1,950-acre Southeast Employment Corridor (SELC) and the Aerospace Research Campus, which can shorten entitlement timelines when a parcel fits an already-adopted specific plan. These sites are also marketed to other industrial users (biomedical, manufacturing, logistics), so competition is not limited to other data center developers.

Arizona State Land Department trust land bordering SELC offers a distinct acquisition path, through public auction rather than negotiated purchase, with different timing and competitive dynamics. The private industrial market is separately active, with roughly 18 current listings ranging from $325K to $10M. The county's first major project set a rough benchmark at approximately $20.9M for a 290-acre parcel.

The fundamentals are real: workforce, land and utility capacity all support a serious project here. The friction is also real: political volatility, an evolving regulatory code, a paused tax incentive, and a documented precedent of organized, well-resourced opposition in this exact region.