Estimating the economic value of reliable electricity
It was hot when Hurricane Beryl knocked out power for millions of Houston customers in July 2024. People willing to spend $1,000 on a portable generator had trouble finding one on the shelves of local hardware stores. Some residents who already owned one even locked it to their fences to prevent theft.
Georgia Tech’s Bobby Harris found similar hurricane-related spikes in consumer demand in his nationwide analysis of two million generator sales. They helped him estimate the economic value of grid reliability, or the willingness to pay (WTP) to avoid power outages—an important feature of the U.S. electricity market. Harris’s interest was both professional and personal.
“I grew up in ‘Hurricane Central’ in North Carolina and remember losing power for two weeks from Hurricane Isabel,” says Harris, an assistant professor of economics who became an EPIcenter faculty affiliate in fall 2024. “My family bought a portable generator right before Isabel hit our town in 2003.”
His analysis of generators sold at 1,800 major retail stores from June 2012 to December 2020 showed that three-quarters of purchases occurred in the week before and after an outage, and two-thirds in counties with at least one tropical storm advisory during the study period. That period included five of the costliest U.S. hurricanes, from Sandy in 2012 to Harvey in 2017. Harris focused on hurricane-caused outages since they explain 90% of blackouts affecting at least 1,000 people for an hour or more.
Given the strong dependence of sales on recently experienced outages and storm predictions, he developed an analysis model where “outage beliefs” and generator prices predicted consumer demand.
“It’s reasonable to assume that people buy the amount of backup power they expect to need during future outages, based on their county’s blackout history up to the purchase time,” explains Harris. “Since I was unable to measure these outage beliefs directly, I developed a method for estimating household WTP from county-level data on sales, outages, hurricane tracks and demographics.”
This estimation involved three steps: First, a generator’s running wattage—a measure of how much power it can provide—determined its expected power output during an assumed 10-year lifetime. Next, Harris used 60 years of historical data to estimate baseline outage beliefs per county and month, similar to how the federal government estimates a county’s hurricane risk index. He updated these monthly outage beliefs for all counties that experienced hurricane-caused outages during the study period. Finally, he combined household demographics with the price and power output of each generator and with outage beliefs at the purchase time to estimate household WTP as $1.57 per kilowatt-hour (kWh).
The average annual outage duration of 6.6 hours from 2012 to 2020 resulted in an annual WTP of $10.37 per household, or $1.2 billion for all U.S. households—equivalent to raising the average monthly electricity bill by $0.86.
Regional variation in average outage durations caused regional variation in WTP. For example, average 60-hour outages increased the annual WTP to $94 in parts of Florida, South Carolina and Georgia. The WTP to avoid winter outages is likely higher, says Harris. Such outages are more common in the Northeast and Midwest and have a greater impact on human lives since they affect home heating.
Purchasing generators is a household decision, similar to installing rooftop solar systems with battery storage or joining a community microgrid. But the WTP estimate also helps assess potential utility investments in improved reliability, says Harris, noting that these have better economies of scale. (Utility-scale solar plants are more cost-efficient than residential systems for the same reason.)
Downed distribution lines, which deliver low-voltage power to homes and businesses, cause 94% of all outages. Grid hardening strategies include replacing wood with concrete poles, trimming trees near power lines and undergrounding lines, the most expensive option. The cost of undergrounding existing lines ranges from $158,000 to $5 million per mile and far exceeds household WTP. This effort might only make sense in densely populated areas with frequent outages, says Harris. However, households are not the only grid users.
“Similar research to estimate WTP for commercial and industrial users will be important to help decide how to pay for infrastructure upgrades,” says Harris.
Such decisions involve consumers, state legislators, utility companies and public service commissions, which oversee electricity rates. These monthly rates include a volumetric charge for the consumed amount of power and a constant fixed fee for all grid users. To support investments in improved reliability, California has been discussing restructured electricity bills with a lower volumetric charge and a higher fixed fee that depends on household income.
“Our growing electrification and the rise of AI data centers make grid reliability more important than ever,” says Harris. “As an EPIcenter faculty affiliate, I hope my energy-related research will inform discussions about the future of Georgia’s electricity market.”
News Contact
Published on: September 2, 2025
Story Written by: Silke Schmidt
Priya Devarajan || Research Communications Program Manager