Does Texas Have the Energy to Reindustrialize?

My last post raised the prospect of an American manufacturing boom after decades of relative decline. The question of whether Texas could attract substantial new industrial investment depends on having the infrastructure and workforce to support it. Electric power generation is one such essential requirement. Factories today require fewer workers–they have lower job density per square foot. Automation allowed this to happen. The more automation a plant uses, the more electricity it needs. 

The fact that Texas is growing and that there are more competing uses for electricity may put a bind on manufacturing growth. Since 2020, Texas’ population growth rate has accelerated. More power will be needed for new homes, shops, and schools. Electrification of the car fleet continues, and this will require more power. Finally, the AI boom, and the ever-increasing volume of digital content will only be sustained by building many more data centers. U.S. Census Bureau construction data suggests that data center development has grown rapidly and will exceed the total volume of office construction by early 2026. Are the state’s power generators up to the task? 

In the year ending May 2025, Texas plants generated a monthly average of more than 48,000 gigawatt hours (GWh) of power. Peak generation in this interval was in August 2024 at 58,500 GWh. Ten years prior, the monthly average was 36,700 GWh. The peak month then, also August, was 45,200 GWh—less than today’s monthly average (see figure). That means the annual average increased by 31 percent in a decade. Electricity growth has been faster than the state’s physical and economic growth. Over the last decade total population has gone up 15 percent, and total real economic output is up 26 percent.  

Given rapid growth in electricity demand overall, what additional power might be needed to support more manufacturing? 

In a national assessment of electricity demand, consulting firm Deloitte called attention to the role of manufacturing. Their study cited data on recent manufacturing plant announcements. Many industry segments are represented among the 150 factories identified, including metal manufacturing, semiconductors, motor vehicles, and plastics. This diversity is a good proxy for the types of projects we might expect Texas to attract. These facilities are estimated to require 13,000 GWh annually. If we divide this annual demand by 150 plants, we arrive at 86 GWh per plant per year, or a little more than 7 GWh per month.  

Using the data at hand, we can estimate how much more manufacturing Texas can support given power requirements. The U.S. Energy Information Administration (EIA) tracks proposed power plants. Their data show that many new plants are projected to come online in the U.S. through 2030. The total output of these new plants is estimated at 30,800 GWh per month. This is over and above today’s 48,000 GWh per month. If realized, this would be a two-third increase!  

This astounding growth is even more impressive considering that EIA projects that one third of all new U.S. capacity is planned in Texas. If all this new capacity were dedicated to running factories like the ones described above, the state could power over 4,000 new manufacturing facilities. Of course, the many competing uses for electricity mean that far fewer factories will find available power. For instance, although estimates vary, a data center can use 10 times the amount of electricity as a manufacturing plant, per square foot. So, each new data center saps enough power to run scores of factories. 

This assessment shows that Texas power generators are working feverishly to meet equally rapid growth in demand. Ultimately, electricity will be allocated like all market goods. Those willing to pay more will gain access to the power they need. Whether Texas manufacturers can outbid the data centers and other users will depend on how much economic value they can create with that power. That in turn will depend on how much society as a whole values steel, computer chips, and tortillas over streaming movies and AI art.  

Future installments in this series will consider Texas’ manufacturing prospects in light of transportation infrastructure, water, workforce, and other essentials. 

Views expressed on The 338 are those of the authors and do not imply endorsement by the Texas Real Estate Research Center, Division of Research, or Texas A&M University.