Republicans Don't Have a Data Center Problem. Our Platform and Principles are Clear.

Twenty years ago a coalition told Texans that hydraulic fracturing would poison our water, ruin our land, and make our state unlivable. They said tap water near the Barnett Shale would catch fire. The Permian would never recover. The aquifers would empty inside a decade. None of it happened.

Thankfully, instead of giving in to the fear and propaganda, Texas instead pioneered the techniques the rest of the country and most of the developed world now uses, regulated them appropriately, and withstood every campaign of imported foundation-funded litigation and federal pressure designed to stop us. They then ran the same playbook on coastal natural gas, on pipelines, on drilling in the Gulf, and on every refinery expansion any Texas operator has developed since 2010. Same Washington money. Same NGO infrastructure. Same misguided arguments in slightly different costumes.

Because Texas held the line we led the shale revolution, broke OPEC's grip on global pricing, and built the eighth-largest economy on earth, while emissions per unit of energy produced fell.

The skilled-trades paychecks that funded a generation of Texas families, the energy security that protected American foreign policy for two decades, and the manufacturing base that quietly rebuilt parts of the Gulf Coast all trace back to Texas Republicans standing firm against the attempted spread of anti-energy panic.

The principles that won that fight haven't changed, and the opposition's methods haven't changed either. Now they've just moved on to a new target, and it's just as important that conservative principles win the fight again this time.

The new target is the physical infrastructure of American AI. As the factory floor for the next industrial revolution, data centers are where the most consequential technology contest of the next half-century gets won or lost.

A rancher in Milam County signed a lease with Stargate. He is not Microsoft. He is not in San Francisco. He is the kind of Texan a carve-out would tell to go ask permission of agitators who don't live on his road and don't want him doing anything with his property. A landowner in Nueces County signed for Hut 8's Beacon Point campus. The $9.8 billion fifteen-year lease announced last week is the first phase of a $17 billion buildout. Real Texans, real land, real contracts, real permits.

Opponents are agitating for arbitrary mistreatment of a legal business. Separate utility rate classes, retroactive permit hurdles, county-specific bans, water-use restrictions written to bind only one buyer. Each does the same thing.

Each singles out one lawful industry for treatment we would never accept for any other. In some places the carve-out is winning. Maine's Democratic legislature passed a moratorium bill the governor had to veto. Delaware House Democrats just passed a party-line bill saddling data centers with a separate utility rate class. Baltimore. Harford County. Bulloch. Sedgwick. Streetsboro. New Jersey. Madison. The map of carve-out proposals isn't Texan, isn't conservative, and isn't Republican.

And it isn't organic. Last week Power the Future asked Rep. James Comer and Sen. Rand Paul to open a formal investigation into what their letter called “a coordinated, billionaire-funded, and potentially foreign-backed political campaign” against data center and AI infrastructure.

Power the Future identified 188 local opposition groups across 24 states. It tracked more than $13 million flowing through three of the largest left-aligned dark-money conduits: the New Venture Fund, the Sierra Club Foundation, and the Sixteen Thirty Fund. Earthjustice. Food and Water Watch. Good Jobs First. The Piedmont Environmental Council. The Southern Environmental Law Center. MediaJustice. The Athena Coalition. Familiar anti-conservative and anti-Texan names. Donors who in many cases don't live in Texas, don't pay property taxes here, and don't bear the consequences of a carve-out they fund.

The script their money has bought was carefully chosen. The campaign has worked partly because it has aimed at AI. AI is new, AI is scary, AI is the easy hook. But the data center buildout in front of us isn't an AI project alone.

Every call any of us places, every text we send, every order we put through Amazon or Walmart or our local florist, every visit to a county website, every prescription our pharmacy refills, every payment our small business runs, every photo grandparents share with their grandkids moves through multiple data centers. AI has accelerated the demand curve, but it's not the only driver.

The campus rising in Milam County and the one at Beacon Point are going to serve a Texas neighbor's grocery delivery, a Texas grandmother's video call, and a Texas small business's payroll. The agitators know that. They present AI as the only target because no Republican primary voter is going to vote against the device in his or her pocket.

We know that large data center companies and customers like AWS and Microsoft have not been friends of the conservative movement. They've funded the wrong politicians, deplatformed the wrong voices, and structured their corporate posture for the other side.

But none of that changes the decision. The rule of law and the rest of our principles apply to every buyer or they aren't principles at all. A property right we apply only to buyers we approve of stops being a property right. A free-enterprise commitment we honor only when the industry suits us isn't a commitment. The Texas conservative test is whether the buyer is following the laws: every permit obtained and obeyed, every appraisal complete, every limit observed.

Even when buyers comply, the opponents move the goalposts, and only for this one industry. Until they recycle the playbook against the next opportunity they want to kill.

When a policy proposal lands in front of a Texas conservative, the platform tells us where to start. Free enterprise. Limited government. Property. Sovereignty. Personal accountability. The platform's ten Principles have been ratified and refined by convention delegates every two years for decades.

The 2024 delegates introduced them with a deliberate commitment: “We, the 2024 Republican Party of Texas, believe in this platform and expect our elected leaders to uphold these truths through acknowledgment and action.” Principles are not advisory. They are designed to function exactly as they're functioning here: as the criteria against which proposals get measured. Opposition to data center development loses by every principle that applies.

Government's proper role

The Republican Party of Texas platform commits us in Principle 9 to “a free enterprise society unencumbered by government interference or subsidies.” It commits us in Principle 4 to “limiting government power to those items enumerated in the United States and Texas Constitutions.”

An arbitrary carve-out is government inserting itself between a willing buyer and a willing seller on private land, after every permit has been obtained and obeyed, because a third party doesn't like the buyer's industry, even though it's completely legal.

Texas Republicans spent two decades fighting that exact instinct when it came from Washington and the anti-energy opposition. We must not do it here just because the buyer is a tech company instead of an oil company. The position that government can prevent a willing buyer and a willing seller from completing a legal transaction in full compliance with all regulations on private land is the opposite of a conservative position.

The Texas tradition

Platform Principle 8 commits us to “the inalienable right of all people to defend themselves and their property.” That principle didn't first appear in 2024. It is older than the state itself. Texas was settled by people who signed land grants, walked onto the property, and built something. The position that opposes every data center and every alternative would have told Stephen F. Austin and his Old Three Hundred families never to show up with their Land Grant. That was not the Texas way.

The landowners in Milam and Nueces aren't asking the legislature for a favor. They're asking the state of Texas to honor its own most basic promise: that a Texan who follows the law on his own land can build on it. A carve-out says no to that promise. It says no for reasons our own platform rejects. And it says no on behalf of donors who don't live here, don't pay property taxes here, and don't bear the consequences.

Sovereignty and the next factory floor

The Preamble of the platform names the loss of sovereignty as the loss of everything else: “If we fail to maintain our sovereignty, we risk losing the freedom to live these ideals.” Principle 3 reinforces it. Our broader platform demands we “eliminate dependency on adversarial nations, such as China, for critical medical, technological, energy, and other vital resources.” China is building over 60 gigawatts of data center capacity by 2030. They're field-testing truck-mounted ten-megawatt nuclear reactors to power them. They've lined up $8.4 billion in financing for orbital data centers. The Trump administration's AI Action Plan prioritized exactly what this contest requires: faster permitting, regulatory clarity, infrastructure investment. Get those three right and we lead. Get them wrong and the lane goes to Beijing. Every Texas data center we don't build is capacity Beijing is happy to add.

Our country made the wrong choice last time we faced this kind of challenge, in manufacturing. We spent forty years telling ourselves the factories didn't matter, because the engineering, design, and management jobs would stay. They didn't. We woke up dependent on the Communist Party of China for our antibiotics, our magnets, and our chips. Data centers are that factory floor. We can't afford to outsource and offshore them again this time.

This is also where the energy abundance fight lives. Our delegates ranked Securing the Electric Grid in their top eight legislative priorities for the 2025 session, calling for hardening the grid and for Texas to “build and operate traditional and next-generation nuclear energy plants.” We've put $350 million into a Texas Advanced Nuclear Development Fund. GE Vernova and Blue Energy are building a gas-plus-nuclear plant on Texas soil. The PUCT's proposed rule under Project 58481 puts real pricing on speculative interconnection requests so the queue gets cleaner, not slower.

Data center load is the customer that finally pays for the grid we've needed to reliably power our population and economic growth since long before Winter Storm Uri. We are closer to delivering it today than we have ever been, because the customers who can finance it are finally at the table. More than a thousand people move to Texas every day. They need power, they need water, and they need an industrial base big enough to share the property-tax load so the residential homeowner isn't carrying it alone.

If Texans throw up arbitrary obstacles to data center development we will forfeit a generational chance to harden our grid, expand our water supplies, and ease our homeowners' property-tax burden.

A new industrial use isn't automatically good for Texas. The case for this one is that it is uniquely good at three things Texas conservatives have been demanding and not getting.

The agitators want the conversation framed as “what do we lose if we say yes.” The correct framing is “what do we forfeit if we say no.”

Some of you read all of that and still feel uneasy because someone told you data centers get a free ride on tax breaks. An abatement is a lower tax, and lower taxes are a conservative principle.

When Texas cuts taxes for an oil major or a manufacturer to land a plant, we don't suddenly call it a subsidy. The vocabulary doesn't change because the buyer's industry is new.

What matters is net contribution: what the project pays in versus what it costs the community in services. Data centers don't enroll kids in schools. They don't generate police calls. They don't fill highways at shift change. Modern hyperscale designs are moving to closed-loop and air-cooled architectures that consume a fraction of the water older facilities used. Whatever water any project uses flows through the same county and state permitting any landowner faces. The full water comparison comes in a later piece.

The same opponents never want to answer the property-tax question either. Even during the abatement period, the property-tax revenue from some data center projects we've seen has exceeded 50% of a school district's entire current tax base.

There is unlikely to be any better opportunity to significantly lower residents' property taxes. And when the abatement window closes the resulting tax revenue increase is another meaningful lift, again with no additional corresponding effort or demand on local resources. The choice in front of a county isn't “data center or nothing.” It is “industrial tax base or residential tax base.” Singling out one lawful industry is a tax-and-spend argument wearing a fake Republican costume. That is not the Texas way.

The last time American conservatives let the left write the policy on dispatchable, scalable, clean energy, we lost forty years of nuclear power. We could have built two hundred reactors. We built about ninety. We exported the technology, paid Russia to enrich our fuel, and watched France build the grid we could and should have built ourselves. That was the cost the last time the conservative movement deferred its own principles to the moral framing of its opponents.

The civilian nuclear industry we should have run instead got run by French utilities and Russian enrichment plants. The skilled trades who would have built our plants built somebody else's. The hundred-plus reactors that never got built were a hundred-plus baseload customers our turbine manufacturers, fuel suppliers, and grid integrators never got to serve. The American emissions reductions that those reactors would have delivered we ended up importing from European utilities and global atmospheric averages instead of producing ourselves. A generation of carbon policy was written around a hole the opposition punched in our energy system, and we still pay the bill in higher prices and worse grid reliability every winter when the wind doesn't blow and the gas pipelines freeze. The conservatives of 1979 who declined to defend nuclear didn't know they were forfeiting a forty-year industrial advantage. The conservatives of 2026 do not have that excuse.

Texas has the chance to write the better ending. The $350 million Advanced Nuclear Development Fund. The Vernova partnership. The legislative work to stand up our own state nuclear regulatory authority. The trades training that BlackRock just dropped $30 million into so a thirty-eight-year-old electrician with a quote on his first house can take the job. Microsoft can absorb a carve-out. He cannot. All of it sits on the table because the data center load sits on the table. Walk away from the load, walk away from the comeback.

Three of the eight legislative priorities our delegates set in San Antonio point in one direction: Secure the Electric Grid. Texas Is Not For Sale. End Federal Overreach. Build at scale, on Texas soil, with Texas labor, for American purposes. The platform is unambiguous. The donors and agitators pushing the carve-out narrative are not Texan and they are not Republican. They're the same coalition that was wrong and thankfully lost in Texas on fracking, was wrong on nuclear inside the conservative movement, and is now trying to sway good Texas Republicans on data centers by laundering left-wing arguments through populist lies. We cannot let them.

The carve-out violates Platform Principle 4 (limited government), Principle 8 (defense of property), Principle 9 (free enterprise unencumbered by interference or subsidies), and the Preamble's commitment to sovereignty. Which conservative principle is it actually defending? None.

The platform isn't ambiguous. The platform expects action. The fight isn't new. The donors writing the script aren't ours.

This is not the Texas way.

Acknowledge. Act. Hold the line.