From cotton to cloud: In the data center debate, is rural Georgia once again being asked to carry the cost?

For generations, the story of the American South was written in cotton.

The crop transformed the region into an agricultural powerhouse, supplying textile mills across the United States and Europe. By 1860, the South produced the overwhelming majority of the world’s cotton, yet only a small fraction was processed locally. Most left the region as a raw commodity, bound for Northern factories and overseas mills, where the greatest share of the manufacturing, financing and commercial profits were realized.

The South supplied the land, the labor and the natural resources. Others captured much of the long-term value.

The consequences reached far beyond ledgers. As production intensified through the 20th century, Southern agriculture grew dependent on chemical pesticides. Toxaphene — a mixture of some 700 compounds registered for use against 167 insect pests, and for a time the most heavily applied insecticide in the country — was sprayed primarily on cotton; more than 409,000 metric tons were used worldwide between 1946 and 1974 before the EPA cancelled it in 1982. DDT, aldrin, dieldrin and parathion were laid down across millions of acres alongside the mixture.

Farmworkers handled chemicals that would never be approved today. Families lived beside treated fields. And the residues did not leave when the crops did. Because toxaphene breaks down slowly in soil, it persisted in Southern farmland for decades; sampling of former cotton fields in Georgia and South Carolina has detected DDT and its breakdown products in nearly every soil sample tested, generations after application stopped.

The human cost is harder to see, but is increasingly well-documented. These organochlorine compounds are fat-soluble, so they accumulate in the body and can be measured in exposed people many years later. Decades of epidemiological research have linked long-term pesticide exposure to elevated rates of Parkinson’s disease, certain cancers and neurological harm. One large study of farmers and their spouses tracked such outcomes across more than 60,000 people over 20 years. The burden falls first on the people who worked the fields and the families who lived beside them, long after the crop that justified the spraying was gone.

Decades of continuous cultivation also depleted soils, accelerated erosion and reduced productivity across large portions of the South. And communities dependent on a single crop proved vulnerable when markets shifted, mechanization cut labor demands and textile manufacturing moved elsewhere. The population declined. Downtowns emptied. Schools consolidated. Young people left in search of opportunity.

Across much of rural Georgia, the abandoned cotton warehouses and shrinking tax bases remain, enduring reminders that industries built to serve distant markets do not necessarily stay rooted in the communities that first sustained them.

Today’s frontier is digital rather than agricultural, but the debate feels familiar.

Georgia now hosts more than 200 data center facilities and has become one of the fastest-growing markets in the country for new development, according to Georgia Tech’s Energy Policy and Innovation Center. The largest of these — hyperscale facilities serving cloud computing and artificial intelligence — consume enormous amounts of electricity, require large tracts of land and, in many cases, substantial quantities of water for cooling.

Local governments are responding at a remarkable pace. As of early June, 34 counties and 23 cities across Georgia had adopted moratoriums, passed data center ordinances or begun drafting them — roughly one in five counties statewide, a threefold increase since last October.

The wave runs from metro Atlanta to Southwest Georgia. Bulloch County extended its moratorium through the end of 2026 and has openly discussed a permanent ban. Augusta imposed a pause in June after more than two hours of public comment. Camden County adopted a six-month moratorium in May, with Terrell County following suit in early July. And in Lee County, commissioners are weighing whether to extend their own moratorium into 2027 while they write rules from scratch. As County Manager Scott Addison put it: “We currently do not have any provisions in our ordinances to identify, define, or tell us how to permit a data center.”

At the state level, House Bill 1012 would have barred counties and cities from permitting any new data center until March 2027. It stalled — as did similar efforts in at least a dozen other states — but the fact that such bills are being filed at all signals how quickly public skepticism has grown.

Nowhere is the conversation more immediate than in Albany and Dougherty County, where officials say it is only a matter of time before a data center developer — depending on one’s point of view — blesses or targets the community.

No concrete proposal has been filed in the city or county. But there have been inquiries, and the industry is already circling the region. In June 2025, the Crisp County Commission approved rezoning for a 2.1 million-square-foot data center expected to cost more than $6 billion, to be built in phases between 2029 and 2033. In Early County, a developer is eyeing Blakely for a proposed 12 million-square-foot campus, and commissioners there approved a data center ordinance this week over residents’ concerns about water contamination and pollution.

Albany is trying to get ahead of the question. In July, the city launched a series of four community town hall sessions, the first held at Dougherty Comprehensive High School.

City Manager Terrell Jacobs told residents that Albany’s city-owned electric utility holds one of the key cards in attracting the industry: 20 to 30 megawatts of excess power capacity and, according to Jacobs, “… with that aspect comes the opportunity to sell off that excess.” That is not enough to support a hyperscale campus, but it is comparable to the amount of electricity consumed by dozens of smaller data centers operating throughout metro Atlanta.

Others are urging caution before any deal is struck. At last week’s Dougherty County Commission meeting, three speakers — including members of the Sowega Aquifer Alliance — asked commissioners to enact a temporary moratorium, echoing the step already taken next door in Lee County. Their concern is the resource beneath everyone’s feet: the aquifer that supplies drinking water and irrigates the farmland of the Dougherty Plain, the same land and water that once carried the region’s cotton economy.

One resident’s assessment, delivered to the commission, distilled the stakes: Data centers can look like a quick fix for the problems facing rural America, but the bargain may be a Faustian one.

The organizing is spreading across county lines, and across political ones. A Mitchell County farmer helping build a regional coalition told Georgia Public Broadcasting that the issue unites people who rarely sit in the same room, because, as she put it, everyone drinks the same water.

The clearest measure of what data centers want from Georgia is written into the state’s own power grid.

In December, the Georgia Public Service Commission unanimously approved Georgia Power’s request for nearly 10,000 megawatts of new energy capacity — roughly a 50% increase, including five new natural gas units — primarily to serve projected data center demand. The company estimates that about 80% of its projected demand growth over the next decade is tied to data centers.

Whether that demand ever fully materializes is an open question. A report commissioned by the Southern Environmental Law Center concluded the utility’s forecast rests on a statistically improbable scenario — a roughly 1-in-500 outcome — and warned that building for speculative demand “risks driving unnecessary investments and infrastructure.” Under the approved agreement, Georgia Power pledged to shield existing customers from related rate increases into 2031 if the boom fails to arrive. Critics note that residential customers, who already pay more than $175 a month on average, are once again being asked to trust that promise.

Then there is the tax question, and here the numbers speak plainly.

Georgia’s sales tax exemption for data center equipment, enacted to attract exactly this kind of investment, cost the state an estimated $474.2 million in forgone revenue in Fiscal 2025, according to a state audit conducted by the University of Georgia’s Carl Vinson Institute of Government. That figure is projected to reach roughly $625 million next fiscal year — up from about $10 million in 2020.

That same audit found that about 70% of Georgia’s data centers would likely have been built even without the exemption, meaning only about 30% of projects were attributed to the tax incentive. Researchers estimated the exemption generated 8,505 construction-related jobs and 1,641 operational jobs attributable to the incentive.

That is not nothing. But set it against the scale of the subsidy, the land, the water and the electrical capacity being committed, and the old question returns with uncomfortable clarity: Who bears the costs, and who ultimately captures the value?

The comparison between cotton and cloud computing is imperfect. Cotton depended on intensive human labor; data centers rely on capital, computing infrastructure and vast amounts of electricity. Cotton’s environmental and economic legacy is well-documented; the long-term impacts of hyperscale development may not be fully understood for decades.

But the underlying arrangement is all too familiar. A hyperscale data center built in rural Georgia will not exist for the benefit of local residents. The information processed inside may serve businesses in New York, hospitals in California, manufacturers in Texas or consumers around the globe. Like cotton before it, the product is destined for markets far beyond the community where it is produced or, in this case, processed.

That does not make data centers inherently good or bad. It does invite a conversation rooted in accountability rather than optimism alone.

What commitments should communities secure before dedicating land, electrical capacity and water to facilities serving national and global demand? How many permanent jobs justify the risk? How should local governments weigh long-term infrastructure costs against projected tax revenue — particularly when the state has already exempted much of it? What safeguards protect water supplies and reliable electric service for the residents and businesses already here?

And perhaps the hardest question: Why have so many rural leaders been willing to accept nearly any major industrial investment?

The answer may lie not in data centers but in decades of economic transition. Much of rural Georgia’s economy was built around industries whose primary markets existed elsewhere: agriculture, timber, manufacturing. When technology reduced labor needs or markets shifted, those decisions were made with little regard for the towns left behind. By then, the leaders who had championed those industries were long gone, leaving future generations to contend with the consequences.

The challenge facing today’s local leaders is not whether to welcome development. Rural Georgia needs jobs and investment to reverse decades of population loss. The challenge is ensuring today’s development does not repeat yesterday’s pattern, where communities provide the resources, shoulder the long-term consequences and discover years later that the greatest benefits flowed elsewhere.

History does not dictate the future. But one in five Georgia counties has already decided the questions are worth asking first. The rest would do well to ask them, too, before another generation stakes its economic future on an industry whose customers, shareholders and business decisions are driven not by the fortunes of rural Georgia, but by markets far beyond its borders.