Data centers are popping up across North Texas and many communities are pushing back, expressing concerns about water and energy consumption as well as pollution.
The sales pitch for data centers though, which house the hardware required for A.I. and other technologies, is that they're good for the local economy.
But how true is that?
Daniel Yue, an assistant professor in the Scheller School of Business at the Georgia Institute of Technology, studies the economic impact of data centers and told NTX Now's Miranda Suarez that the answer is complicated — and depends on where you live.
These interview highlights have been edited for length and clarity. To hear the full conversation, click the 'listen' button above.
'Modest' economic impact
We researched the entry of data centers into different counties around the U.S. and wanted to know what effect they had on the local economies where they were built. We were primarily interested in economic outcomes and what we found is that when data centers enter, they spur economic development in a modest way.
We were looking at outcomes like new business establishments, the creation of jobs and wages. What we find over the first three years after the entry data centers is about a 1% increase in jobs and corresponding increase in establishments.
These gains are not just constrained to the construction industry. They're not purely short-term, but actually the effect seems to amplify over time, which we interpret as evidence of modest economic spillover, the development of economies that goes beyond just the effort that went into building these data centers.
Although data centers are extremely large investments, a lot of that money is actually going to the computers themselves and not necessarily to the labor that's constructing it or operating these centers. So, seeing some of this spillover is a promising sign that some of the economic promises associated with these investments are actually going to be realized.
Higher electric prices
The main way that data centers can affect electricity prices is if they are connected to the broader grid and they start drawing power from it.
Data centers are running very power-hungry computers that can often be consuming on the order of tens of thousands of households worth of electricity at a given point in time. So, if a grid is not able to expand its capacity, then greater demand will lead to an increase in prices for electricity and this can affect local households.
In the counties where they have their own electrical grids, we do find that there is a modest increase in electricity prices — about a 5% increase over the time window that we're looking at.
That is a significant increase statistically, but it's nowhere near the level of some of the narratives that have come out about these A.I. data centers, doubling or tripling the price of electricity. That's not what we find on average, even if it may be true in certain areas.
Rural vs. metro areas
This is a classic pattern that this is not unique to data centers: when there are large, localized investments, a lot of times the spillovers manifest most in more densely populated areas. The reason is that if the investments are made in a very rural area, they may not have the other infrastructure or facilities in place that allow them to capture the value of the investment.
So for example, if someone is traveling into a rural county to a data center, they may not choose to live there and stay there, in which case the corresponding businesses that they'd be frequenting would not be incentivized to go in at the same time.
In more dense areas, we tend to find that co-located investments will have a greater spillover effect. A lot of this is attributed to the fact that there is more infrastructure in place, and this means that the local investments can be amplified more.
Future impact
We certainly have a ways to go to understand the local economic impact in terms of breaking out precisely what other industries are benefiting from the entry of data centers and what is the nature of these spillovers.
I think the bigger question right now is trying to document some of the environmental and other downsides that people have been raising. A lot of the political debate over data centers is because people are afraid of the unknown, right? So, understanding those downsides and understanding what policies we can implement in order to mitigate those downsides or trade them off against each other would be really important, I think, to help communities be able to navigate these decisions as they figure out what will work best for them.
Miranda Suarez is a cohost of KERA's NTX Now. Got a tip? Email Miranda at msuarez@kera.org.
KERA News is made possible through the generosity of our members. If you find this reporting valuable, consider making a tax-deductible gift today. Thank you.
