Could Abbott’s Data Center Directive Wall Off Competition?

Estimated Time to Read: 11 minutes

Some of the biggest companies involved in artificial intelligence (AI) and data center development are lining up behind Gov. Greg Abbott's (R) new approach to Texas data centers.

Meta, OpenAI, Digital Realty, MARA, Skybox Datacenters and others have publicly embraced standards aimed at electricity infrastructure, water use, taxpayer subsidies, transparency and impacts on surrounding communities. Their endorsements follow Abbott's direction to the Public Utility Commission of Texas (PUC) and the Electric Reliability Council of Texas (ERCOT) to conduct a comprehensive verification and audit of data centers advancing through ERCOT's interconnection process before additional projects are allowed to move forward.

There is nothing inherently troubling about companies agreeing that they should pay their own costs, protect neighboring property owners, or operate transparently. In fact, many of those principles are consistent with what Texas Policy Research (TPR) has argued throughout the growing debate over data centers. Still, the enthusiastic response from some of the industry's largest companies raises a different question that Texas policymakers should not ignore. Could Texas be building a regulatory wall that is relatively easy for the industry's largest companies to scale but increasingly difficult for smaller competitors and new entrants to overcome?

That does not require assuming that Abbott's standards were written at the request of large data center companies or that their endorsements are evidence of some coordinated effort to suppress competition. There is no evidence to justify that accusation.

The concern is simpler. Regulations can apply equally on paper while imposing very different economic burdens in practice. When the companies being regulated already have the lawyers, engineers, consultants, and regulatory infrastructure necessary to comply, new government requirements can become another barrier separating established players from potential competitors.

Major Data Center Companies Back Abbott's Texas Rules

OpenAI's response to Abbott illustrates why the corporate endorsements deserve closer examination. In an August 7 letter to the Governor, OpenAI said it supports protecting grid reliability, conserving water, preventing infrastructure costs from being shifted onto residential and small-business customers, and increasing transparency surrounding its projects.

More noteworthy for the competition question, however, was the company's description of Abbott's standards. OpenAI said the standards “largely reflect practices already embedded” in how the company develops and operates projects.

The company committed to fully funding project-driven infrastructure costs rather than shifting them onto residential or small-business customers. It also pledged to work with utilities and ERCOT on generation and infrastructure needs, minimize water consumption, address site-specific community impacts and provide information about electricity use, water use, infrastructure investments, public incentives and community protections.

Again, many of those commitments are reasonable.

The interesting policy question is what happens when government transforms practices already routinely followed by some of the world's largest technology companies into a regulatory process that every affected competitor must navigate. For OpenAI, Meta or another hyperscaler, extensive engineering analysis, regulatory filings and interactions with state agencies and utilities are already part of doing business. These companies operate projects measured in billions of dollars and can employ specialists devoted specifically to regulatory compliance.

The same requirements may look very different to a smaller developer.

Are Texas Data Center Rules Building a Regulatory Wall?

Regulatory costs do not necessarily scale evenly with the size of a business. A multibillion-dollar company can spread the costs of attorneys, engineers, consultants and compliance personnel across enormous investments. A smaller company may need many of the same professional services despite having substantially less capital over which to distribute the expense.

That matters when regulatory approval itself becomes a condition for entering the market. Abbott's approach requires affected data center projects to provide extensive information concerning their proposed operations as part of the verification process. The PUC and ERCOT are then expected to review that information before projects advance through the interconnection process.

For the largest companies, that may amount to another compliance function within an organization already built to handle it. For smaller developers, the combination of professional costs, project delays, and uncertainty about regulatory approval can affect financing and ultimately determine whether a project proceeds at all.

This is how a regulatory wall can emerge even when the law never explicitly favors a particular company. Government does not have to exempt large companies or intentionally discriminate against smaller ones to protect incumbents. Every additional compliance requirement adds another brick to the wall confronting anyone trying to enter the market. Individually, those requirements may appear reasonable. Collectively, their fixed costs, delays and uncertainty can become a significant barrier to entry.

For a hyperscaler investing billions of dollars, that wall may be relatively easy to climb. The company already has attorneys, engineers, consultants, regulatory specialists and access to capital. For a smaller developer or new entrant, the same wall can determine whether a project is financeable at all.

That is the competition problem Texas policymakers should be examining.

The effect can be particularly significant when the government process is uncertain. A company does not merely need to know what information it must provide. Investors and lenders also need some confidence about how long approval will take, what standards regulators will apply, and whether satisfying today's requirements will actually allow the project to proceed tomorrow.

That uncertainty has a price. The larger the company and the deeper its access to capital, the easier that price is to absorb.

This does not prove that major data center companies support Abbott's standards because they expect to benefit competitively. They may sincerely believe the standards are good policy. But policymakers should evaluate the consequences of regulations independently of the motivations of those supporting them.

When some of the companies most capable of absorbing a regulatory burden enthusiastically embrace it, Texas should at least ask how that burden will affect competitors that are not similarly situated.

Texas Data Center Accountability Is Still Necessary

Raising concerns about regulatory barriers does not mean data centers should operate without rules or accountability.

Texas is facing legitimate questions about the industry's rapid expansion. Large data centers can require substantial amounts of electricity. Some projects require significant water resources. New facilities can require transmission, generation, roads and other infrastructure. Nearby property owners may experience concerns involving noise, light, traffic, drainage and other impacts.

Texans are entitled to ask who bears those costs. The limited-government answer is not that private companies should be allowed to externalize their expenses onto everyone else. It is that government should address identifiable problems through clear, predictable, and neutral rules rather than open-ended discretionary power.

Texas Policy Research has previously argued that cost causation should guide infrastructure policy. If a project creates identifiable infrastructure costs, those costs should generally be assigned to the party responsible rather than socialized across unrelated ratepayers. Property rights require a similar balance. The owner who wants to develop property has rights, but so does the neighboring landowner. If industrial noise, drainage, lighting or another impact causes actual and measurable harm, affected property owners should have meaningful remedies. Clear standards, nuisance law and private rights of action where appropriate can protect those rights without creating a political veto over otherwise lawful private development.

Transparency can also be legitimate, particularly when government itself is involved. Texans should be able to determine whether taxpayer subsidies have been promised to a project, whether public infrastructure costs are being shifted onto others, and what commitments public entities have made on behalf of private developers.

The dividing line should be whether regulation addresses an identifiable cost or harm, rather than whether government simply wants more discretion over who is allowed to build.

Texas Should Reject Subsidies and Regulatory Favoritism

The competition question also exposes an important connection between regulation and corporate welfare. Texas has spent years using tax abatements, grants, economic development agreements and other incentives to attract large investments. Data centers have benefited from some of those policies.

Texas Policy Research has consistently opposed that approach. Government should not decide which businesses deserve advantages unavailable to everyone else. If a data center project is economically viable, it should compete based on the merits of the investment rather than its ability to secure a favorable arrangement from government. But ending subsidies while simultaneously building a regulatory wall that established companies are uniquely equipped to scale would merely replace one form of government-created advantage with another.

A limited-government approach should reject both.

Large data center companies should not receive preferential tax treatment because they promise enormous investments, and smaller competitors should not encounter unnecessary government barriers merely because they lack the compliance apparatus of a hyperscaler.

The principle should be equal treatment under predictable law.

That means data centers should pay their own way. It means taxpayers should not subsidize them and unrelated ratepayers should not absorb costs attributable to their projects. It means companies should be responsible for actual harms they cause to neighboring property owners. It also means Texas should be skeptical when regulation becomes a wall around the marketplace, particularly when the companies already inside are far better equipped to navigate government than the competitors trying to get in.

Free enterprise does not mean protecting businesses from responsibility. It means government should neither favor nor punish companies based on their size, political influence, or industry.

Texas Data Center Policy Belongs With the Legislature

This brings the competition question back to the institutional concern surrounding Abbott's directives. If Texas needs new policies governing large electric loads, infrastructure cost allocation, water disclosures, taxpayer incentives or other aspects of data center development, the Legislature is the proper institution to debate and establish those rules. The legislative process forces policymakers to define the scope of government authority. Lawmakers can determine which projects should be covered, what information should be required, what standards companies must satisfy, what due process protections should exist and whether requirements should vary based on the actual size or impact of a project.

Those decisions also occur publicly, with legislators accountable to voters for the policies they enact.

An executive-directed verification process is different. When regulators are given substantial discretion over whether a project can proceed, businesses must determine not only what the law requires but what regulators will ultimately consider sufficient.

That distinction matters for competition. Regulatory uncertainty is easier for sophisticated incumbents to navigate because they have greater access to legal counsel, technical expertise, government affairs personnel and capital. Smaller competitors may be much more vulnerable to delays or shifting expectations. Clear statutory rules do not eliminate compliance costs, nor should they necessarily do so. But they can make those costs more predictable and reduce the importance of having the resources or political sophistication necessary to navigate discretionary government processes. If Texas needs additional data center regulations, lawmakers should identify the specific problems requiring government intervention and write rules directed at those problems.

The objective should not be to eliminate every rule standing between an applicant and approval. Some rules protect legitimate public interests. The objective should be to ensure that every brick in the regulatory wall serves an identifiable purpose, rests on lawful authority, and imposes no greater burden than necessary to address the problem at hand.

Grid infrastructure costs should be addressed as grid infrastructure costs. Taxpayer subsidies should be addressed by eliminating corporate welfare. Water transparency should be addressed through appropriately tailored disclosure requirements where justified. Property harms should be addressed through objective standards and meaningful remedies.

Texas does not need an open-ended permission structure simply because legitimate concerns exist.

Texas Data Center Policy Should Protect Competition

The growing support for Abbott's directives among major technology and data center companies does not establish that the standards were designed to protect incumbents. It would be irresponsible to make that accusation without evidence.

But their support should encourage Texas lawmakers to examine the competitive consequences of the emerging regulatory structure.

A company investing billions of dollars can absorb attorneys, engineers, consultants, regulatory specialists and months of uncertainty much more easily than a smaller competitor. Every new requirement may apply equally on paper while adding another brick to a regulatory wall that becomes progressively harder for new entrants to scale.

Texas should address legitimate concerns surrounding grid reliability, infrastructure costs, water use, transparency and neighboring property rights. The question is not whether government should have any rules. The question is whether those rules are clear, predictable, narrowly directed at identifiable problems and applied equally. As lawmakers consider what comes next, every proposed Texas data center regulation should face a straightforward test: Does this requirement protect Texans from an identifiable cost or harm, or is it simply another brick in a wall that businesses must have enough money, influence and regulatory expertise to climb?

Texas should demand accountability without walling off competition. Data centers are not the problem. Bad policy is.

---