Today’s conversation is with Mark Muro, senior fellow at the Brookings Institute’s metro policy program. Too often I’m asked, what’s the version of a data center boom that people like? I reached out to Muro because he recently coauthored research into the ways communities and data centers can potentially work together to build more mutually beneficial and popular industry growth. The conversation wound up perfect for The Fight, so I had to include it in full.
The following Q&A was lightly edited for clarity.
What do you identify as the primary driver of the backlash we’re seeing to data center development in the United States?
They are potentially disruptive, large scale developments and also take on a talismanic quality where they stand for something. Both dimensions have really agitated people. On the one hand, often in rural communities there’s a lot of concern about energy use, price impacts, noise in some cases and so on, and for many communities these are a quality of life issue. For others, AI stands in for anxiety about jobs not coming. At a time when people are worried about jobs being displaced by AI, data centers are a convenient Other. They agitate and are focal points for a lot of concerns.
The data is pretty clear: a data center brings to a community an initial surge of construction jobs and then a quite modest level of operational jobs. A community might gain in the near-term several thousand jobs but then the long-term employment is welcome but not as large as had been advertised. Some of them can be decent jobs and we should acknowledge that.
What about tax revenue?
It can be significant but the deals are often worked out quietly. It’s hard to get a systematic take on that. A lot of that also depends on the skillfulness and aggressiveness of local public officials because all of it needs to be worked out in a deal. There are certainly tax benefits in some cases, but those are harder to pin down and seem to range.
Okay, so what is the pathway towards these projects being a more meaningful and positive long-term community investment?
That’s the right question because a data center isn’t inherently a negative for a place.
We think the need is first for communities to use the data center in its own aspirational plans. Places need to know what they want. They should be focusing on high-quality jobs, long-term employment, and in some cases even innovation gains for their local economies. Too rarely have communities taken an aspirational view.The deals are worked out on the fly, without a gameplan for the region.
Communities need to ask for more, require more, and come into these deals with their own priorities.
In some cases there have been communities that for a long period of time built up a number of data centers and felt like they gained benefits. Areas near the Columbia River in the Northwest seem to have worked with Microsoft and other companies to facilitate data center construction while also gaining quality employment and funding for schools. It is possible.
In our report we detail a number of places that have begun to put together these kinds of deals that are beneficial, often in places with a university nearby where there’s interplay on the technology front. I think in those cases, we may be beginning to see a rethinking of how these projects should go down and benefit.
Also, this year the backlash has become such a hurdle for the companies that they’re beginning to rethink how they operate. I think the jig is up for the bad old days and we’re going to see more thoughtful arrangements made in the next few years because everybody agrees, what’s been going down the past few years hasn’t been beneficial for any of the actors.
Do you see industry players picking up on a need to be more mindful of what a community needs? I’m thinking about Meta’s recent announcements around workforce training, for example.
Yes. Both for reasons of seeing what’s needed but also the need to make some concessions to really be a better neighbor. It’s forcing some really beneficial outcomes.
Workforce is one of the key aspects of how Microsoft has been far-sighted in Wisconsin, working with the state university and a community college and so on. I think hyperscalers are beginning to move in a more promising direction.
Do you think we’re still going to be having this same conversation a year from now? Things are moving so fast.
Regions are really up in arms about this. It’s become clear that in many cases they’re going to block development. So to the extent hyperscalers want to continue to build, they’re going to have to pursue a more community friendly way to do that.
I think the conversation is going to change. It’ll have to change if the industry wants to continue building capacity.
