On January 6, 2026, 44 people testified before the City Council of Wilmington, Ohio, urging elected officials to withhold approval of the proposed site plan for an Amazon data center in their community. A video of that meeting is available here. The council’s Planning Commission voted unanimously to withhold approval of the Amazon data center Site Plan. With partners at Good Jobs First, Policy Matters supported the community’s effort with an open letter, reprinted below.
State legislators have been extraordinarily generous to companies that operate data centers, though Amazon and other corporations have said they would have built in Ohio even without the incentives. The data centers that already exist are driving up electricity costs throughout our state. The people of Wilmington provide an example of how Ohioans can ensure our elected officials understand the serious questions data centers raise for our communities.
An Open Letter to the City Council and residents of Wilmington, Ohio:
We are writing to strongly recommend that Wilmington defer any further action and not enter hastily into an agreement regarding a proposed Amazon Data Services, Inc. data center.
Good Jobs First is a 27-year old, nonprofit, nonpartisan resource center for anyone in the U.S. seeking to improve economic development deals or programs. We published our first study on data center incentives in 2016. Policy Matters is a 25-year-old, nonprofit, nonpartisan research institute supporting a vibrant, equitable, sustainable, and inclusive Ohio. We have published extensively on tax and budget issues, including incentives, throughout our history.
We have reviewed the proposed Community Reinvestment Area project agreement and associated materials. Based on those specifics, as well as our own cumulative experience with incentives generally and data centers specifically, we believe that Wilmington can strike a far better deal than has been proposed. Specifically:
The Amazon data center project is already extremely tax-advantaged
Amazon Data Services, Inc. is asking for a 30-year, 100% property tax abatement offset only by a 30% Payment in Lieu of Taxes (PILOT) for an effective 70% tax abatement on its “real property” (i.e., land and buildings). We consider that an unnecessary concession for these reasons:
Ohio repealed its corporate income tax years ago. So Amazon will not be paying any state tax on its Ohio profits.
The state also no longer taxes corporate personal property (i.e., machinery and equipment), so the vast majority of the installed value of the data center will be property-tax free. By far the costliest parts of a data center are the servers, AI-capable microchips, cooling and ventilation systems, power and fiber-optic cabling, stand-by power generators, software, data storage, batteries, switches, firewalls, and operations centers.
As well, Ohio’s data center incentive law exempts all of the data center-specific building materials and equipment from sales and use taxes, so Amazon won’t pay either the state (5.75%) or the local Clinton County share (1.5%) of the sales tax. This is an enormous subsidy and would constitute a large sum of foregone local revenue, given that equipment and software can account for 80% or more of the capital expenditures for a hyperscale data center.
Amazon will also likely benefit from the additional tax breaks available to so-called “Megaproject Operators” under Ohio law. This law doubles the duration, from 15 years to 30 years, of Job Creation Tax Credits (which are a diversion of workers’ state personal income taxes away from the state to the employer), Community Reinvestment Areas, and other tax-based subsidies.
So the only tax left standing is the local tax on real property, and this agreement proposes effectively surrendering 70% of that. Given all of the other tax exemptions and abatements this project would enjoy, we believe it is unlikely that paying 100% of its real property tax would be a material site location issue to Amazon.
Every state that has estimated its fiscal return on data center tax abatements has found sharply negative returns in practice. Given Ohio’s income and property tax codes, and its sales and use tax exemption for data centers, we have no doubt that Ohio is also shifting the tax burden away from data center operators and onto working families and small businesses. There is no reason Wilmington should amplify that shift.
Finally, regarding cost-benefit analysis: we see no evidence that the City has attempted to account for the public health costs of a data center’s air pollution, given that diesel emissions (from required standby power generators) are associated with higher rates of asthma, other respiratory diseases, cancer, and heart disease.
For all of these reasons, we recommend Wilmington give no local tax abatement at all.
The proposed agreement fails to provide electricity ratepayer safeguards
The Clinton County Port Authority’s Data Center FAQ states that AES Ohio has confirmed that there is sufficient generation capacity to meet the project’s electricity demand. However, AES Ohio customers have already seen a 13.1% increase in residential energy costs per kW over the summer months in 2025, compared to prices in June-August of 2024 – the second-largest increase across all of Ohio’s utility providers.
Soaring electricity costs for Wilmington residents are heavily driven by data centers’ projected load growth, widening the imbalance between electricity supply and demand in the 13-state transmission grid that serves Ohio. This growing imbalance supports the need for local agreements that incentivize the addition of new generation supply to the grid through measured, responsible investment, while expanding protections for residential ratepayers.
The Union of Concerned Scientists, in a close analysis of utilities’ rate filings, has reported that Ohio ratepayers paid $1.3 billion in 2024 for approved data center transmission costs.
The Wilmington agreement should require Amazon to enter into a contractual agreement (a “will serve” letter) with AES Ohio confirming adequate existing capacity to meet the project’s energy needs. The agreement should stipulate that the approval of the project hinges on this “will serve” letter, detailing whether additional investments in new generation resources or infrastructure are necessary to connect the data center to the grid.
The proposed agreement fails to ensure specific levels of job creation or capital investment
For several decades in the United States, economic development incentive agreements, or the legislation that enables them, have required companies to create jobs, invest capital, and/or achieve other tangible goals with the clear understanding that failure to deliver will result in a reduction or cancellation of the tax incentive. This is usually achieved by a “clawback” (or rescission or recalibration) provision that specifics goals and penalties, or a “performance-based” structure in which subsidies are effectively backloaded and only provided after the company proves it has delivered.
The proposed Wilmington-Amazon agreement would turn the clock back about 50 years to a “no strings attached” model. This is very unusual.
The modest 100 permanent jobs that Amazon estimates will be created are not guaranteed. In fact, the agreement specifies that the job estimates “will not limit the amount or term of the tax exemptions provided under this Agreement or allow the City to compel the Company to create positions.” The agreement has the same “will not limit the amount or term of the tax exemptions provided under this Agreement or allow the City to compel the Company to make investments” language covering its estimated $4 billion capital investment.
The proposed agreement does allow the City to reset the real property tax rate to 100% under certain circumstances involving cessation of operations. But that is, of course, the most drastic scenario, and does not discernibly cover partial job-creation or capital-investment shortfalls. Other provisions allowing the City to terminate the abatement if Amazon fails to fulfill its obligations are notably weak, with no provision for recovery of previous lost revenue.
The proposed agreement fails to maximize local Community Benefits
Twenty-four years ago, Good Jobs First published the first “how to” manual for community groups to negotiate Community Benefits Agreements (CBAs) with developers in order to maximize the local benefits of major development projects. This framework has since taken root in many local, state, and federal projects and programs, and is considered an economic development best practice.
In this case, a CBA could address every issue local residents have had with hyperscale data centers across the U.S., such as: minimizing air pollution with state of the art emissions controls; ongoing air quality monitoring in downwind areas; negotiating the routing of power-supply lines to minimize disruption to existing land uses; maximizing distance from residential property, schools, and parks; requiring closed-loop water cooling to reduce noise pollution and water supply problems; use of state-certified construction apprenticeship programs to maximize local construction employment; and a local hiring agreement to give incumbent residents first crack at obtaining the permanent jobs. A CBA could also fund energy-efficiency measures such as investment in a low-income home weatherization and conversion to heat pumps fund.
A CBA can first exist as a legally enforceable contract between two private parties: the community coalition and the developer. A CBA may also be appended to a development agreement between a local government and the developer, making it legally enforceable that way as well.
CBAs also typically include a joint, long-term, community-developer monitoring process to oversee implementation and ensure that commitments are met.
We would caution against considering any corporate “strategic philanthropy,” such as making special one-time contributions, as “community benefits,” in lieu of true benefits like those we enumerate here, or in lieu of paying 100% of a real property tax obligation. We note that the proposed agreement calls a payment of $1.85 million to the Clinton County Port Authority a “One-Time Community Benefit Payment.” One-time payments cannot take the place of ongoing tax payments.
Wilmington has handled the Amazon data center project undemocratically
At $4 billion in estimated capital expenditures, this project would be, we are told, the biggest in Wilmington’s history. It would rank among Ohio’s largest as well. Given its scale and potentially disruptive impact, we would normally expect the City to have provided extensive public disclosure of the project details far in advance of public hearings at which any tax abatements or other project concessions would be considered.
The city’s inadequate public disclosure was evident at the July 1, 2025, Planning Commission and August 7, 2025, City Council rezoning hearings. No residents testified at either hearing, apparently unaware that 480 acres adjoining their neighborhoods were being rezoned from Rural Residential to Light Industrial.
Even if the City sought to provide a bare-minimum level of notice to adjoining property owners, we believe that when such a large, potentially disruptive project is under consideration, elected and appointed public officials have an affirmative obligation to actively notify all residents who could be affected and intentionally seek their input.
Had Wilmington residents known that an industrial use that would emit air and noise pollution was proposed so close to their neighborhoods, many would likely have attended the early hearings. Especially given heightened public awareness today in the United States about problems associated with data centers, if Wilmington residents had known of that proposed use, much more civic engagement would certainly have occurred. This is confirmed by the fact that after the nature of the project was known, 40 people spoke to the Wilmington City Council about the project on November 25, and chambers “overflowed” on this date as well as December 4.
Instead of robust engagement, the City signed a non-disclosure agreement (NDA) on January 30, 2024, and withheld many project details for more than a year and a half while conducting secret negotiations with the company. The resulting proposed agreement reflects this lack of community involvement.
We hope that our suggestions can help Wilmington achieve a better deal.
Sincerely,
Greg LeRoy
Executive Director
Good Jobs First
Zach Schiller
Research Director
Policy Matters Ohio
Lame duck is an opportunity for legislators to act on data centers (and more)