For years, Texas offered many of the things technology companies wanted: relatively inexpensive energy, abundant land, low taxes and a government eager to attract investment. Data centers followed.
Now state officials are trying to figure out what all that success is going to cost.
On Monday, Gov. Greg Abbott ordered the Public Utility Commission of Texas and the Electric Reliability Council of Texas, or ERCOT, to audit data centers seeking to connect to the state’s electric grid. New approvals are effectively on hold while the review proceeds.
The numbers help explain the concern.
Paul Cobler of The Texas Tribune reports that ERCOT is tracking more than 1,800 proposed projects representing 474 gigawatts of potential electricity demand. About 90% of those requests come from data centers, according to Abbott. ERCOT’s record peak electricity demand is about 91 gigawatts.
No one expects all 474 gigawatts to materialize. Many projects will never be built, and others may be speculative requests from developers trying to reserve a place in the queue.
But that presents its own problem. Texas has to plan its electric grid years in advance. Regulators cannot intelligently decide where new transmission lines, substations and generation will be needed if they cannot distinguish serious projects from placeholders.
Abbott wants considerably more information. The audits will examine expected electricity demand, on-site generation, water consumption, cooling systems, ownership, neighborhood impacts and tax incentives. ERCOT has paused its existing process for evaluating large new electricity users while the review proceeds.
Texas’ experience also raises a familiar economic-development question: What happens when government succeeds in attracting the investment it has been subsidizing?
Texas has been generous.
In April, Cobler and Apurva Mahajan of The Texas Tribune reported that the state’s sales-tax exemption for data centers is costing more than $1 billion annually. The state comptroller estimated Texas would forgo $3.2 billion in sales-tax revenue over the following two years.
Supporters can reasonably argue that Texas receives something in return. Data centers represent billions of dollars in private investment and are increasingly important infrastructure for cloud computing, artificial intelligence and other digital services. Texas may genuinely be one of the best places in the country to locate them.
But the tax breaks are only part of the public cost.
As viable projects move forward, someone also has to pay for the transmission lines, substations and other infrastructure needed to serve them.
Abbott has directed state regulators to make the data centers themselves bear those costs.
In June, he directed state regulators to require data centers to fully fund the electric infrastructure needed to serve them rather than shifting those costs to residential customers. He also called for phasing out what he described as outdated tax incentives and requiring better reporting of electricity and water use.
Texas regulators had already been considering the same issue. Shelby Webb of E&E News by POLITICO reported in May that regulators were considering changes to transmission charges for large electricity users, primarily data centers, to keep residential customers and small businesses from bearing a disproportionate share of billions of dollars in grid improvements.
That is an important distinction in economic-development policy. A project can bring substantial private investment to a community and still impose substantial public costs. Counting the former while ignoring the latter makes almost any development look like a bargain.
Texas is also discovering how difficult those costs can be to calculate.
According to the Tribune, regulators surveyed data center operators about water consumption, cooling systems, electricity demand and power sources. Only 28 companies, representing 92 facilities at various stages of development, responded. The PUC could not say how many surveys it had sent.
That is not much of a basis for planning billions of dollars of infrastructure.
Still, Texas should be careful not to overcorrect. A lengthy or unpredictable approval process could discourage viable projects along with speculative ones.
Webb reported Tuesday that Abbott’s order has disrupted ERCOT’s plans for evaluating new large-load projects. ERCOT had expected as much as 200 gigawatts of proposed projects to qualify for its first review group, although officials believe roughly 65 gigawatts are more likely to be built through 2032.
That gap illustrates both the problem and a possible solution.
Texas does not need to choose between subsidizing data centers and stopping them. It can require developers to provide credible information, establish predictable rules and make projects pay for the infrastructure necessary to serve them. Companies can then decide whether the economics still work.
For years, states and cities have competed to attract large developments by offering tax breaks and other incentives. The announcement usually focuses on how much money a company promises to invest.
Texas’ data center boom is a reminder that investment is only one side of the ledger.
One must count the costs, too.
