.png)
Watts, water, and land: Who benefits from data center development?
Weighing data center benefits against community costs is critical for the Pacific Northwest and Great Basin.
The AI boom is driving the largest U.S. infrastructure buildout since the interstate highway system. More than 700 data centers are under construction across 38 states. Proponents argue the investment brings tax revenue and construction jobs and strengthens national competitiveness. They also cite efficiency gains and voluntary sustainability commitments.
Opponents raise concerns about rising electric bills, strained power grids, depletion of local water supplies, industrial-scale noise, destruction of habitat, draining of public funds through large tax breaks, and a lack of transparency. Communities nationwide have enacted or are considering moratoriums, restrictions, and bans on data centers.
The full story is more nuanced and complex. Decision-makers must weigh potential economic benefits against environmental, social, and fiscal costs.
In the Pacific Northwest and Great Basin, the conflict extends beyond electricity bills to the long-term allocation of finite shared resources: water, power, and land. Columbia Basin water rights are already oversubscribed. Decisions must also account for salmon habitat, clean-energy mandates, and tribal treaty obligations.
As an economist and mediator who has worked with communities and industry on competing resource needs, I've learned that a flat-out no isn't the best response. Data centers done well can provide benefits. It shouldn't be a yes-or-no question.
Too often, development processes bypass public participation. Communities need good information and enough time to use it. Early participation helps developers and decision-makers address concerns before costly disputes arise.
What does data center development bring to communities?
Data centers can offer genuine economic benefits to host communities. Microsoft points to property taxes from its facilities in Quincy, Washington, that helped fund a new high school and police station. But the conditions that produced those benefits have changed. The surplus hydropower that helped attract early development is largely spoken for.
Construction brings jobs and spending. Once facilities open, permanent employment is more limited. Tax revenue can support public services, but decision-makers must account for tax abatements, infrastructure costs, and long-term obligations. A large investment can still cost a community more than it returns.
Industry sustainability and efficiency commitments deserve the same scrutiny. Company-wide averages say little about what one facility will draw from a local watershed or utility. What improvements are proposed, and which commitments can be enforced?
Commitments carry more weight when they're written into local agreements. In Hermiston, Oregon, a 25-year water service agreement requires Amazon Web Services to pay the full cost of an aquifer storage project and return at least 100 million gallons of drinking water to the city each year.
What do data centers cost communities, and who pays?
Water demand includes direct consumption for cooling and indirect use through electricity generation, which can be many times larger. Cooling choices can involve trade-offs between water and energy use. Annual averages can obscure local concentration and seasonal pressure on supplies.
The documented impacts extend beyond utilities. Diesel backup generators can affect air quality. Persistent noise and night lighting affect nearby residents. Large campuses can convert farmland, wetlands, and rural land. Environmental justice requires examining how these impacts compound burdens in communities already facing pollution and underinvestment.
Timing compounds these costs. Siting, permitting, and contracting decisions often happen before the their impacts are fully assessed, so communities discover costs after commitments are made.
Communities need that information before agreements are signed. In The Dalles, Oregon, the city fought to keep Google's water consumption confidential before settling a public-records lawsuit and agreeing to disclose it. Communities shouldn't need a lawsuit to learn how much water a facility uses.
What makes the West different?
In Virginia and Texas, the core data center conflict is electricity costs. In the West, it centers on water, hydropower, and the obligations tied to them. In Nevada's Great Basin, around Reno, data centers draw on scarce water in one of the driest regions in the country. In the Pacific Northwest, the pressure falls on the Columbia River, where reliance on hydropower creates distinct choices. Relatively inexpensive, low-carbon power helped draw data centers to the region. The river also supports irrigation and salmon habitat, and decisions affecting it must respect tribal treaty rights.
Hydropower is low-carbon, but it carries ecological costs. The dams profoundly affected salmon and steelhead populations and the Columbia Plateau tribes whose food systems, cultures, and economies depend on them. That makes hydropower less clean than aggregate carbon accounting suggests.
New demand also tests the region's clean-energy mandates. Washington State law requires utilities to provide 100 percent clean electricity by 2045, so serving data center growth means new clean generation, new transmission, or both. Where will that power come from, and who will pay?
These decisions also reach treaty-protected resources. Yet tribal representatives were not initially included in Washington's Data Center Workgroup; the Yakama Nation advocated for a seat. Tribal nations must be engaged as sovereign governments in decisions affecting treaty-protected resources, not brought in after plans are settled.
What should policymakers require?
How data centers are governed will determine whether communities benefit or bear the risks, as the World Resources Institute argues. Legislators need a thoughtful conversation about responsible AI and the rules that should govern data center development. That starts with public meetings and transparency: local officials shouldn't be bound by nondisclosure agreements that keep the public from learning what's proposed. Data centers should meet 100 percent of their power needs with renewable energy. And decision-makers should use new development to improve lives, economies, and water resources, or at least ensure it results in no net negative.
Meaningful participation gives the public a real chance to influence the outcome, whether that means revised terms, another design or location, or a decision not to proceed.
What should communities ask before approving a data center?
Before a deal is signed, communities and decision-makers should ask:
Where will the water come from? Is the source disclosed, what water rights authorize its use, and how will demand be met during drought or a low-snowpack year?
Where will the electricity come from, and who pays? Will the developer cover the costs of serving its demand without shifting them to existing ratepayers?
Do the fiscal terms add up? How does projected tax revenue, after abatements, compare with public-service costs and infrastructure obligations over the project's lifetime? Who pays for decommissioning and any remaining obligations when the facility closes?
What are the cumulative impacts? How will this facility, together with existing and proposed projects, affect water, power, land, and nearby residents? How will those combined impacts be addressed?
Is the process transparent and inclusive? Are tribes at the table from the start? Are project information and deal terms available for public review? Can the public influence decisions before commitments are made?
None of these questions is anti-development. The goal is to help communities, developers, and decision-makers determine whether a proposal's benefits justify its full costs, and how finite resources should be shared.
Bea Covington will discuss these questions as a panelist on "Watts, Water, & Land: The Data Center Debate" at the American Water Resources Association Washington Section's "The Future is Now" conference on October 8, 2026.
Bea is a principal economist and the sustainability director at Greene Economics. A certified mediator, she helps communities, governments, and industry work through complex decisions about land, water, and natural resources.
