OPINION:
There has been a lot of churn in the data center conversation in the past six months or so. I get that; the tech bros tend to bring out the worst in everyone — sometimes, even themselves.
There appear to be a lot of people who are opposed to data centers in their neighborhoods, and politicians are looking for cover. It is not clear how Team Data gets back on the right foot.
Now that one side of the ship seems gunwale under, let me offer a few alternative thoughts.
Let us take a look at the survey work. Despite all the noise and fury, data centers are not really that important to voters.
When asked about important issues facing the country, the voters invariably zero in on the economy, gas prices, healthcare, immigration and Iran. In a recent Harvard-Harris survey of 2,100 registered voters, essentially no one answered “data centers” when asked (in an open-ended question) about important issues facing the United States today.
That is not an aberration.
In May, the Rainey Center asked 1,010 registered voters about data centers. Most respondents worry about their effect on electricity prices (of course). That is a legitimate concern. There is only one problem: A lot of data confirms that electricity prices are more closely tied to whether a state is run by Democrats or Republicans. You can guess which ones produce high rates and which ones produce lower rates.
In August, Echelon Insights conducted a survey of 1,002 registered voters. They discovered that when people are exposed to messages focused on what data centers are doing on water, noise, prices, education, local development, etc., the needle winds up on the plus side.
Those who want a moratorium on data centers fall into three camps: those opposed to artificial intelligence for ideological reasons (Bernard Sanders); Republicans who have trouble staying on message (Mike Rogers in Michigan and Jon Husted in Ohio); and opportunists such as Texas Gov. Greg Abbott, who will probably approve all the pending projects after the election (once they have bent the knee).
It is inescapable that the data center team and its allies have performed poorly to date. They have allowed the issue to be framed around electricity costs, water, noise and whatever else their opponents could devise. The political allies of data center development have played right into their opponents’ hands (just think about the pointless ratepayer protection pledge).
Eventually, though, Team Data will get better. It will hire smart people to reframe the conversation, re-pitch the effort, become good landmen and make the investments in communities that lead to wins in siting and construction. It will start centering the conversation on generational prosperity, the future and the absolute requirement that we — not the regime in Beijing — win this race.
Finally, the context of this fight is fairly simple and strongly tilts toward data centers. People love their phones and are unwilling to dispense with them, no matter the cost.
Each day, 25 billion texts are sent, 2 billion hours of YouTube are watched, 20 billion shares of stock are traded. Half of American households and a shade less than 90% of workers are already using artificial intelligence.
None of those numbers will decrease in our lifetimes.
There are more than 4,000 data centers in the United States. That is more than half of all the data center capacity on the planet. About 600 additional American data centers are planned. One-third to one-half of the centers planned for this year have been delayed, mostly because of suboptimal electrical infrastructure.
There is no question that more data centers will be built or that Americans will build them.
Data center owners will maintain a dispositive advantage with respect to cash, political power and (should it come to this) a government that is itself highly dependent on the output of data centers.
About 20 data center projects have been outright canceled this year. The number may grow, but it will not change the fundamentals of the conversation. Growing demand, a salient mismatch in resources, and the essential nature of the product offered mean that more data centers — a lot more data centers — will be built in the next five years.
• Michael McKenna is a contributing editor at The Washington Times.
Follow the author
Please read our comment policy before commenting.