MUNCIE — Data centers have almost no local economic effects, with one possible exception.

Michael J. Hicks

That’s the result of my recent study examining the effects of data centers on Texas employment and incomes. Texas provides an ideal case study because it offers both property tax abatements and sales tax exemptions for data centers, it posts transparent data on tax incentives, and it experienced the data center rush before other states.

To understand the economic impact, I measured employment in counties with new data centers against those without, focusing on total employment and the sectors most likely affected: information, construction, and professional and technical services.

The results were pretty clear.

No data center had a measurable effect on employment. To double check that finding, we performed a second type of analysis designed to detect whether jobs come early, as in a construction phase, or later, once the data center is operating.

There was no evidence that data centers led to more employment or incomes. There’s just no effect at all, which is not terribly surprising. After all, these are not complex buildings. Much of the value comes from purchased equipment, not the structure. Also, there’s very little labor used to operate them after they’ve been built.

This Texas study offers data center proponents little evidence to justify attracting them in the name of economic development. But there is one potential exception.

Large data centers — those over 100,000 square feet — did appear to influence professional and technical employment. The average effect was reasonably strong, but many places experienced no effect. This finding suggests some future benefits, but it is far from definitive. It is difficult to know precisely how these very large data centers would differ from smaller ones.

The likeliest explanation is that smaller data centers support cloud computing, while larger ones are used for artificial intelligence. The AI facilities might bring with them technicians and engineering or computing staff in larger numbers. That would be a hopeful sign for the local effect of data centers, but it is a very tentative hope.

One finding was very clear. Statewide tax incentives for data centers are unjustifiable. It shouldn’t take an economic study to determine this. Large fiscal incentives for businesses without an employment effect have always been an especially dubious venture.

There’s no real support for local incentives either. But, in the rare case when there are economic development benefits, it will be very localized. Again, there is low probability of these incentives ever paying off, but that should be a local choice, not something taxpayers across Texas (or Indiana) should bear.

It’s also worth noting that, as part of this study, I tested the effect of data centers on electricity prices and price changes for residential consumers from 2010-2023. Again, there was no evidence that data centers led to higher residential electricity prices.

Indeed, the share of the family budget Americans spend on electricity has never been smaller than it is today. Any claims that data centers are driving higher electricity prices likely fail on a number of facts, not least of which is that inflation-adjusted electricity prices are low by historical standards.

Of course, this study only evaluated jobs and electricity prices through 2023 or 2024. There could be changes in the future, so it is best to keep an open mind.

Despite these results, data centers are important to our economy today and will grow in importance as data storage and the computing power of AI becomes more critical to household, government and business applications.

The big takeaway from this research isn’t so much that data centers aren’t a local economic benefit. I think even a modicum of judgment would make that clear. The lesson is that the rush to attract the newest fad blinded legislators, economic developers and county and city councils across the country.

Take Indiana, for example. In the coming weeks, the Indiana Economic Development Corp. is set to release a data center study. It’s worth reiterating: Any study that doesn’t attempt to assess the causal effect of data centers on local employment, wages and other economic measures is probably useless.

The IEDC’s work is unlikely to use rigorous experimental design — we don’t yet have enough variation in the location and timing of Indiana’s data centers to make such analysis possible.

Yet, Indiana’s first attempt at analysis comes more than five years after the state effectively gave the farm away to any new data center that wanted to move here. It’s part of the collapse of judgment that accompanied economic development policy in 2019 and 2020.

That failure cost Hoosier taxpayers billions of dollars and has saddled the Braun administration with many difficult choices. It is a case study on poor oversight, mission creep and a culture of accommodation to any business request for tax relief. In short, a very costly mistake.

The data center lesson should be simple. Do your homework on economic effects, be reluctant to give special tax deals and don’t wait until there’s a taxpayer rebellion to perform due diligence. Those are evergreen policy recommendations.