Data Centers are Infrastructure
My fourth in an initial four-part series on data centers in Ohio.
In September 2025, Jerome Township in Union County, Ohio, became one of the first jurisdictions in Ohio to place a temporary moratorium on new data center applications. In April 2026, St. Louis, Missouri, took a different path. After months of public hearings, the city’s Board of Public Service approved a conditional use permit for a data center, but attached detailed operating conditions and required the project not to seek local tax abatements.
These are two ends of the local policy spectrum: pause and study, or approve with teeth. This final newsletter is about what falls in between.
Before getting to the menu of local responses, I want to start with the frame I have found most useful for thinking about all of this:
Data centers are infrastructure, not traditional economic development.
The distinction matters.
First, benefits accrue diffusely. People all over the region and country benefit from highways, electric grids, pipelines, and digital infrastructure. The local host community captures only a portion of the value created. Data centers follow the same pattern. The value of the facility is consumed mostly by users somewhere else.
Second, costs are concentrated locally. Land is consumed. Resources are used. Local roads, water systems, emergency services, and electric infrastructure may need to be sized around the facility. Neighbors may experience noise, traffic, light, air quality effects, or peak water withdrawal concerns. These are not abstract costs. They show up in specific places.
Third, local control is limited, but not nothing. For some types of infrastructure, state or federal agencies can preempt local zoning. In Ohio, the Ohio Power Siting Board has jurisdiction over major utility facilities. That matters because data center campuses may include large backup generators, behind-the-meter generation, fuel cells, or other energy infrastructure that can move parts of the project outside ordinary local land-use control.
That does not mean communities should reject data centers. It means they should negotiate like host communities, not like desperate bidders for a new economic lifeline.
A Menu of Responses
Communities facing data center development have more tools than many realize, though fewer than some would like. The available tools depend heavily on local zoning authority, utility control, state preemption, and bargaining leverage.
Here is a simplified menu and tradeoffs of each policy option.
Moratorium. A temporary pause can give a community time to study the issue, update zoning, and decide what conditions it wants in place before accepting applications. The strength is that it buys time. The limitation is that it does not solve the underlying problem. A moratorium only matters if the community uses the time to write rules or negotiate terms.
Zoning Ordinance. Local zoning can address setbacks, design standards, screening, noise, fire department coordination, and mitigation plans. The strength is that zoning applies generally, not just to one applicant. The limitation is that zoning only works where zoning exists, where rules are already in place, and where state preemption does not override local authority.
Permit-Conditioned Approval. A conditional use permit can allow a project to move forward while attaching binding requirements. These might include closed-loop cooling, noise limits, decommissioning plans, emergency response coordination, or reporting obligations. The strength is that the community can approve with conditions rather than simply approve or deny. The limitation is that this requires strong local permitting authority and the political will to enforce the conditions later.
Community Benefits Agreement. A CBA is a negotiated contract between the developer and community parties. It can cover water use, noise, e-waste, local hiring, emergency services, bonding, reporting, or other project-specific issues. The strength is flexibility. The limitation is leverage. The company has to agree, and the community needs the information and organization to negotiate effectively.
PILOT with Clawbacks. Payments in lieu of taxes, service payments, and clawback provisions can help communities capture revenue and protect themselves if promised jobs or investment do not materialize. The strength is that the agreement can be tailored. The limitation is that clawbacks are often weak, rarely triggered, or difficult to enforce unless the terms are specific and measurable.
The key point to all of these policy options is timing. Communities have the most leverage before zoning approvals, water agreements, tax abatements, and service commitments are finalized. Once a facility is entitled, constructed, or routed through state siting processes, the local bargaining position weakens considerably.
Understanding What You Can and Can’t Control
Ohio’s recent energy legislation, including HB 15, matters less because it directly regulates data center buildings and more because it changes the landscape around energy generation and siting. For large electric generation facilities, OPSB jurisdiction can sharply limit the practical role of local zoning.
That creates an important risk. A community might negotiate a detailed tax agreement for a data center building, only to later discover that associated energy infrastructure is handled through a state-level siting process with much less local control.
The practical implication is simple: communities need to address power generation early. If the project may include behind-the-meter generation, large backup generators, fuel cells, or other on-site energy infrastructure, those issues should be part of the original negotiation.
The same is true for water, roads, emergency response, and decommissioning. These are not details to be worked out later. They are core terms.
Zoning Realities
Several practical points about zoning are worth making clear.
First, areas without local zoning generally have very limited ability to restrict where a data center locates. If a township or county is unzoned, there may be no simple legal pathway to prohibit placement after the fact.
Second, even where zoning exists, local officials may not be able to block a data center if the use fits within the existing zoning district. The rules need to be in place before the proposal arrives.
Third, it is hard to write a zoning code that says, in effect, “this use is fine when we like it, but not when we don’t.” Rules that restrict data centers may also restrict other large industrial or infrastructure-like uses. That may be appropriate, but communities should understand the broader implications.
This is why planning ahead matters. Once a proposal is on the table, the debate often shifts from “What rules should we have?” to “Can we legally apply new rules to this applicant?”
What Local Governments Can Still Do
Even with state preemption and zoning constraints, local governments still have meaningful authority.
They can negotiate tax terms. CRA agreements, TIF arrangements, PILOTs, service payments, and other local fiscal terms are not automatic. Their structure, duration, and conditions are negotiated. County commissioners, township trustees, city councils, school districts, and other affected public bodies should understand what revenue is being forgone and what the community receives in return.
They can negotiate water access. Local utilities and water providers can attach conditions to service agreements. These can include peak capacity limits, reporting obligations, efficiency requirements, drought contingency plans, and cost-sharing for system upgrades.
They can negotiate land-use approvals before state siting takes over. The building, site design, access roads, screening, drainage, and related local approvals may remain within local authority, depending on the jurisdiction and the specific project.
They can negotiate community benefits agreements. CBAs are contracts, not zoning ordinances. That makes them a potentially important tool where zoning authority is limited or preempted.
Within the tax abatement conversation specifically, communities should pin down a few basic questions:
What is the current tax revenue from the site?
What revenue will the site generate during the abatement?
What revenue will it generate after the abatement expires?
Are there other local tax streams, such as municipal income taxes or sales taxes?
What new costs will the facility impose on fire, EMS, roads, law enforcement, water systems, or schools? And most importantly: are those costs reflected in the agreement?
Resources for Negotiations
Two resources are especially useful for local officials thinking through these issues. The first is the Georgia Data Center Ordinance Hub at Georgia Tech’s Energy Policy and Innovation Center. It catalogs ordinances from more than 180 Georgia jurisdictions across topics including setbacks, water and sewer, noise, design standards, and decommissioning. It is Georgia-focused, but the categories and model language are useful for communities anywhere.
The second is the Alliance for the Great Lakes’ regional playbook on data centers and water. For communities in the Great Lakes region, it offers a useful way to think through water supply, withdrawals, reporting, and long-term resource concerns.
No model ordinance or playbook can substitute for local legal advice. But these resources help communities ask better questions.
Questions Communities Should Ask
Water Resources and Efficiency: What are the facility’s projected daily and peak water needs? What is the source of that water? Will the local utility need to expand capacity? Who pays for that expansion? Does the agreement include specific efficiency metrics, reporting requirements, and consequences for non-compliance?
Energy and Air Quality: Will the facility generate power on site? Will it use backup generators, fuel cells, or other behind-the-meter generation? What emissions controls are required? What approval process applies? What monitoring and enforcement requirements are in place?
Noise and Light: What are the noise and light limits at the property boundary? Are they written into the agreement? Who monitors compliance? What happens if the operator violates the limits?
Land Use: Where is the facility locating? Does the land-use change align with the community’s comprehensive plan or long-term development goals?
Local Service Burden: Will the project increase demand on fire, EMS, roads, law enforcement, water systems, or schools? Has that burden been costed out and reflected in the agreement?
Decommissioning, Bonding, and E-waste: Is the operator required to post a financial bond or other guarantee for decommissioning and remediation? What happens if the company restructures, sells the facility, or leaves the market? Is there a vacancy trigger? How will e-waste be handled? What recycling, disposal, and remediation protocols are required?
Long-term Accountability: Who monitors compliance? Are reports public? Are tenants disclosed? How can the community enforce the agreement? Can the agreement be terminated or revised if the operator fails to meet its obligations?
I want to emphasize the decommissioning and bonding question because it is one I rarely see addressed in Ohio’s current data center agreements.
Ohio requires oil and gas well operators to post a plugging bond before drilling. The bond is a financial guarantee. If the well is abandoned, there is at least some money available to plug it and remediate the site. The requirement exists because Ohio has tens of thousands of orphaned wells whose owners could not be held accountable after the fact.
A hyperscale data center is not an oil well. But the risk profile has parallels: a large, specialized structure; a rapidly expanding industry; complex ownership and tenant arrangements; and the possibility that firms restructure, sell, or exit markets faster than local governments expect.
Decommissioning costs are real, and they should be planned for upfront.
The State Policy Picture
This series has focused mostly on what communities can do locally. But many of the most important decisions are being made at the state level, and they are moving quickly.
As of this writing, the Ohio House is signaling a possible override of Governor DeWine’s veto of the data center sales tax repeal. A Data Center Study Commission is also advancing through the legislature, tasked with studying data centers' economic, environmental, energy, and community impacts and recommending policy within six months of enactment.
One aside on the sales tax issue is worth flagging. Justin Ross has written a thoughtful piece arguing that Ohio's data center sales tax exemption sits on different conceptual ground than the property tax abatements I focused on in Newsletter 2. His core point is that data center equipment is an intermediate business input, and exempting it prevents the same value from being taxed at multiple stages. It is an argument worth reading before the override vote. The property tax abatement side of the local fiscal ledger is a separate question, and one Justin and I both think deserves harder scrutiny.
The broader state policy question is whether Ohio will establish clearer standards for data center development or leave communities to negotiate one project at a time. State policy could address water reporting, bonding, on-site generation, local service payments, utility tariffs, transparency, and minimum agreement terms. Or it could entrench the status quo.
Local officials, extension educators, landowners, and community members should pay attention to that process. The rules being written now will shape the bargaining environment for years.
A Final Note on Leverage
I want to close this series where I started it: with the question of leverage.
If data centers are infrastructure, then communities being asked to host them are not really competing for jobs the way they might compete for a manufacturing plant. They are being asked to absorb the local costs of facilities whose benefits flow primarily elsewhere.
That changes what “good economic development” looks like. It is not about beating the next county on incentives. It is about making sure the price of hosting includes compensation for the costs the host community bears. It is about making sure long-term risks are someone’s contractual responsibility from day one.
Communities have often negotiated as if they had no leverage, when in fact they hold assets these companies need: land, water, infrastructure access, local approvals, and community cooperation. The question is whether the communities most affected have enough information, enough time, and enough organization to use that leverage well.
Special thanks to Justin Ross, a public finance economist at Indiana University’s O’Neill School, who helped me sharpen my thinking on data centers. Justin writes thoughtfully on state and local public finance at his Substack, Capital & Capitol, which is worth subscribing to if these topics interest you. Thanks also to Dale Arnold and Leah Hetrick of the Ohio Farm Bureau for sharing legal provisions relevant to data centers, particularly for Ohio’s townships and counties.
This newsletter is supported by the C. William Swank Program in Rural-Urban Policy, housed in the Department of Agricultural, Environmental, and Development Economics at The Ohio State University. Its contents are, and always will be, free.
Please reach out to me at lade.10@osu.edu if you have questions or suggestions for future newsletters.