A new cost-benefit analysis from the Columbus firm Scioto Analysis suggests that Ohio should either require new data centers to operate on 90% renewable energy or implement an outright ban on their new construction. The analysis, which examined future data center development in the state, determined that either of these approaches would offer a better economic outcome for Ohio than the current status quo.
Ohio currently ranks fourth nationwide in the number of data centers. Taxpayers have already provided $2.3 billion in subsidies through tax breaks to support these facilities. However, questions have arisen regarding their overall benefit, as data centers have more than doubled the state's electricity consumption, contribute to noise and air pollution, and generate artificial intelligence (AI) that leaders in major AI companies warn could lead to significant social harms and job displacement for millions of Americans.
Against this backdrop, Scioto Analysis conducted a detailed financial assessment of potential data center growth in Ohio. The Buckeye State follows only Virginia, Texas, and California in its number of data centers, and these "gargantuan facilities" are expected to expand significantly into rural areas across Ohio and other Midwestern states. According to the Scioto Analysis report, "About 87% of current data centers in the United States are in urban areas, while 67% of planned data centers across the country are in rural areas," driven by the substantial land and power needs of modern hyperscale facilities. The report projects that data centers in the Midwest will experience the highest percentage growth nationally, with a 64% increase compared to current figures. By 2026, the report states, 38% of Americans are expected to live within five miles of at least one operational data center.
While data center development is credited by the Scioto Analysis study with creating 96,000 new jobs and generating $62 billion in social benefits for Ohio since 2013, a criticism of their economic impact is that once constructed, they require only a skeleton staff to operate. The study specifically questions what new benefits these centers will create moving forward, focusing on the substantial electricity drain and the social cost of resulting pollution.
The cost-benefit analysis warns that if all currently planned data center projects come online by 2028, the number of such facilities in Ohio will increase by half, their annual energy consumption will more than triple, and their annual carbon emissions will also more than triple. The report highlights the scale of this energy demand, noting that in 2026, Ohio data centers use enough electricity to power 5.7 million homes, surpassing the 5.4 million homes in Ohio. By 2028, the analysis predicts, they will consume enough electricity to power approximately 3.5 times as many homes as there are in Ohio.
Rob Moore, the principal of Scioto Analysis, emphasized the importance of considering the negative impacts a business can have on third parties, such as the public. Moore stated that while data centers represent a new category of high-intensity energy use, they pose similar challenges to other industries like transportation, manufacturing, or power generation. He explained via text message that emissions, if not priced, could lead to local air-quality issues, lung and heart conditions, and higher carbon emissions.
Moore and his team evaluated five distinct scenarios, running 10,000 simulations to project outcomes between 2027 and 2034 with varying conditions. The analysis expressed its findings in "benefit-cost ratio" terms. Maintaining the current approach, or status quo, yielded a ratio of 1.04, indicating that its benefits only marginally exceeded its costs.
The scenarios that offered the greatest public benefit, in descending order, were identified as: a behind-the-meter requirement for 90% renewable energy, which produced both the highest net present value and benefit-cost ratio; followed by an outright ban on new construction, which resulted in a higher net present value and benefit-cost ratio than the status quo. A behind-the-meter requirement with 90% natural gas produced a similar net present value to a ban but with a lower benefit-cost ratio.
The report clarified that a ban on new construction would generate a better outcome than the status quo primarily because the employment benefits from existing data centers would largely remain, while the cost burden of higher electricity prices on ratepayers would decrease.
However, the analysis included an important caveat: its findings are highly dependent on assumptions about job creation. The report noted, "Our results depend heavily on our employment assumptions." If the study overestimates the number of jobs and earnings generated by data centers, even a ban or a 90% renewable energy requirement could potentially have a negative net present value. Conversely, if these employment effects are underestimated, all alternatives would appear much more favorable. The analysis concluded that, across its various tests, the status quo was the alternative "most likely to have costs that exceed its benefits."
