Market Update: Texas Tech Researchers Explore Links Between Power Reliability and Economy | January 2026 | Texas Tech Now – Full Analysis

Rawls College professors analyzed how power interruptions affect regional economic
growth and housing prices.

Most Texans who have resided in the state for the past five years will surely remember
the catastrophic winter freeze of February 2021 that lasted nearly nine days, burst
water pipes and left approximately 10 million people without power.

As devastating as those billion-dollar disasters can be, they often occur beyond the
scope of any utility provider’s control.

What is likely to be treated with more annoyance than heartbreak are the power interruptions
that are not the fault of a hurricane or winter storm, but rather are part of a provider’s
routine operations and disrupt countless day-to-day activities, risking billions of
dollars lost in home values and property taxes. 

Seeking to understand the societal impacts of the latter, Texas Tech University researchers led investigations into how power interruptions affect regional economic
growth and housing prices.

Bradley Ewing and Zachary Keeler, professor and assistant professor of practice, respectively, of energy commerce & business economics in the Jerry S. Rawls College of Business, spearheaded the effort.

Keeler highlighted that this work differs from similar studies by separating outages
lasting more than five minutes, known as Major Event Days (MEDs), from those without
MEDs resulting from non-major events.

“We’re one of the first studies that look at minor events, which we see in both studies
are the things that seem to matter the most. Those outages that are relatively more
avoidable,” he said.

They found that counties, especially rural, with shorter and less frequent outages
may experience positive employment growth and reliable power facilitates local business
success, linking reliability with better societal outcomes. Minor interruptions may
reflect an aging or overextended infrastructure, in addition to negatively impacting
business operations and resident well-being.

The study on housing prices identified that normal operating interruptions negatively
impact prices in metropolitan and non-metropolitan counties, as well as both low-
and high-natural-disaster-risk counties, with the market already incorporating the
risks of MEDs. High-risk counties saw the largest effects. 

Additionally, for every increase by one standard deviation from the average number
of minutes customers experience an outage, housing prices decreased by 0.25%, a statistically
significant figure as indicated by the study. Extrapolated to the total value of U.S.
homes at the end of 2022, a one-standard-deviation increase would lead to over $113
billion lost in home values and over $1 billion in property taxes lost each year

Not only were the projects a case of matching available data with their expertise
and research interests in matters such as urban and regional economics, but they also
held the potential to shine a spotlight on steps to progress. 

“If we know how power reliability impacts housing prices, policymakers can consider
that when making the decisions they need to make,” Keeler said. “Home buyers can then
factor in things like power reliability when figuring out how much they want to pay
for a house or where they want to locate.”

Ewing and Keeler used data on power from 2013 to 2020 submitted by utilities and power
marketers operating in counties in the Lower 48 states of the U.S. They also assessed
the changes in county-level employment and population growth from 2014 to 2019 and
annual housing prices for the economic growth and housing studies, respectively. 

The power reports detailed three aspects of reliability: the System Average Interruption
Duration Index, measuring the number of minutes of interruption the average customer
experiences yearly; the System Average Interruption Frequency Index, or how many times
the average customer experiences an outage each year; and the Customer Average Interruption
Duration Index, which relates how many minutes it takes to restore service after an
interruption. 

Keeler acknowledged there are numerous ways power reliability and economic growth
can be affected by either outside factors or each other, which the study had to control
for using the cross-fit partialing-out lasso method, or a double machine learning
(DML) approach.

DML helps prevent the problems that arise when multiple independent variables, such
as labor force participation rate or broadband access, are correlated and thus muddy
the picture in terms of determining a single variable’s impact on the model results. 

It also addresses overfitting, when models can’t make accurate conclusions or predictions
due to learning from similar variables, by performing variable selection and regularization.
Regularization reduces features of models and limits variance by shrinking the sizes
of the explanatory variables’ coefficients.

The housing price study utilized several methodologies to check for consistent estimates
and quality assurance. Beyond exploring the differences between urban and rural areas
and analyzing counties’ various natural disaster risks, the study touched on the effects
of migration, such as local movers knowing more about a county’s power reliability
than incomers.

Both investigations illustrate that power reliability matters, Keeler said, but the
conversation should continue onto determining what can be done to be better at restoring
power quickly and preventing avoidable outages.

“I don’t necessarily have those answers,” he added, “but we show that that might be
an important consideration if you want your region to grow or if you want local housing
prices to increase because we see that there’s this linkage there.”

There are direct and indirect effects, which the group will continue to investigate
as they seek to analyze power reliability’s impact on state-level gross domestic product
and income inequality.

Regardless of the exact type of research, Keeler looks forward to communicating with
the larger, public audience, be it economic development officials or other stakeholders,
to transform ideas into tangible progress.

“I think, as scholars, we should try to bridge that gap as best as possible,” he said.
“We want our research to have an impact.”