The question facing Hermantown has already been tested elsewhere: What happens after a data center opens, the construction crews leave and the community begins collecting taxes while living with the facility?
In places with years of operating experience, the record includes substantial public revenue, permanent employment and infrastructure investments. It also includes persistent neighborhood complaints, pressure on water systems and decisions to tighten rules governing the next development. Those outcomes sometimes occur in the same community.
Are those communities pleased? Some officials point to measurable financial gains. Some neighbors describe a diminished quality of life. Others support existing operations while questioning how much additional development their communities can accommodate. The records reviewed for this column do not establish a representative, communitywide satisfaction rating. They do show why a city’s financial results and its residents’ experiences can produce different answers.
Loudoun County, Virginia, provides one of the clearest examples of the financial benefits. The county reports that data centers generated $1.2 billion in real estate and personal property tax revenue in fiscal 2026, representing 39% of its overall budget. Its account of the industry’s development describes how data centers helped diversify a tax base that had depended heavily on residential property before the housing market collapse. This is an operating industry producing public revenue, rather than a developer’s projection of what might eventually arrive.
A December 2024 study by Virginia’s Joint Legislative Audit and Review Commission identified specific uses for data center revenue. Local government staff reported that it helped Loudoun and Prince William counties lower real estate tax rates, supported an affordable housing trust fund in Henrico County and helped Mecklenburg County construct schools. The review also found that local tax rates and incentives substantially affected what communities collected. A large investment did not produce an identical public return everywhere.
Loudoun’s financial success has nevertheless been accompanied by a change in development policy. The county eliminated “by right” data center development in 2025, requiring special-exception approval for future proposals. Its public explanation acknowledges growing concerns about noise, scenery, water, energy infrastructure and other neighborhood effects. It also says the county is working to avoid excessive dependence on data center revenue. Loudoun’s experience demonstrates that collecting substantial taxes and imposing tighter controls on further growth are compatible positions.
Employment requires a separate accounting. Virginia’s legislative review found that much of the industry’s economic benefit occurred during construction. Industry representatives told researchers a typical 250,000-square-foot facility might employ about 50 full-time workers, roughly half of them contractors. Those operating jobs tended to pay well, but their number was substantially smaller than the construction workforce. A community can receive a significant permanent tax contribution from a facility with a comparatively modest permanent staff.
Council Bluffs, Iowa, offers a more recent example of disagreement within a community that already hosts Google data centers. In June, Mayor Jill Shudak sought a pause of up to a year on new projects so the city could examine infrastructure and development standards. The council unanimously rejected the request. Council members argued that the city could review its policies while continuing to consider developments. Shudak acknowledged substantial economic benefits and said she was not seeking to prohibit data centers permanently.
The infrastructure concern there was specific. Public Works Director Matt Cox told local reporters that requests for water could reach 1 million or 2 million gallons daily, and that only a few such projects could consume available capacity. He also described the administrative burden of negotiating each development separately. That dispute concerned the pace and terms of further expansion in an established data center community. It did not establish that existing facilities had failed financially, or that residents uniformly supported additional ones.
In Chandler, Arizona, the experience of some neighbors centered on sound. ABC15 reported complaints from residents near a CyrusOne facility about a continuous hum. One resident described a formerly quiet neighborhood and said eliminating the noise would make it livable again. Those accounts concern a particular facility and neighborhood; they are not evidence that every data center produces the same disturbance. They do illustrate an operating impact that can persist after construction traffic and building activity have ended.
Chandler subsequently adopted detailed requirements for data center development. Its ordinance, effective in January 2023, called for a preconstruction sound study, disclosure of the findings to residents, mitigation to prevent noise from exceeding the measured baseline and annual sound studies during peak operations for five years after construction. It also established neighborhood communication requirements, an on-site liaison and restrictions governing routine backup-generator testing. The city explained that evolving facility designs had increased potential noise impacts and prompted stronger oversight.
Virginia’s statewide review adds perspective: Most data centers did not generate noise complaints, which researchers attributed to location or design. At some facilities near homes, however, continuous low-frequency sound affected residents even when it rarely exceeded existing noise limits. Legal compliance and neighborhood satisfaction were therefore not always equivalent.
The Dalles, Oregon, presents another combination of benefits and resource demands. OPB reported that Google paid $7.3 million in property taxes and a community service fee in 2025, while receiving substantial tax discounts under its agreements with local governments. Separately, city records supplied to OPB showed Google-related water use of 434.4 million gallons in 2024, up from 383.9 million gallons the previous year. The operation was producing public revenue while also functioning as a major municipal water customer.
Google has also financed water infrastructure there. In October 2025, the company announced completion of an aquifer storage and recovery system designed to capture water during wetter months for use during drier periods. Google said it transferred the system and associated groundwater rights to the city, providing more than 100 million gallons of additional annual water availability. That is a documented infrastructure contribution, although the capacity figure is the company’s account of the system’s benefit.
The relationship has nevertheless generated disputes over transparency and how development agreements are described. OPB reported a legal fight over disclosure of Google’s water use, with records eventually released. City officials also challenged aspects of OPB’s reporting. A former public works director said infrastructure improvements were required as a condition of service to protect existing customers, while tax incentives were provided through separate agreements. The distinction matters when determining what a company paid for, why it paid and which obligations belonged to the public.
The Oregon water figures cannot simply be applied to Hermantown. Hermantown’s published description says its proposed campus would not use water for cooling and would use municipal water for domestic purposes, with consumption of no more than 50,000 gallons daily at full buildout. Differences in cooling systems, climate, project scale and utility arrangements limit the usefulness of direct comparisons. The Dalles provides evidence about that community’s experience, not a forecast of Hermantown’s water demand.
Nor do the Virginia tax totals establish what Hermantown would collect. Virginia’s results depend in part on taxes on computer equipment, local rates, depreciation schedules and incentives. The relevant comparison is the revenue a community actually retains under its own rules, alongside the services and infrastructure it must provide. Employment comparisons similarly require separating construction workers, permanent operating employees, contractors and jobs attributed indirectly to spending.
The evidence from established communities supports a qualified answer to whether they are pleased. Public officials can identify schools, tax relief and infrastructure supported by the industry. Residents near some facilities can identify continuing disturbances. Communities that accepted early developments have subsequently reconsidered location standards, monitoring requirements and the pace of expansion.
For Hermantown, those experiences provide a record of outcomes to examine: what was collected, what was built, what neighbors experienced and what local governments changed after the facilities began operating.
