To embrace this return to science, I want to write today about a fascinating new working paper by Tufts economist Steve Cicala.
Professor Cicala has been studying the effect of Covid on electricity consumption since back in March, when the Wall Street Journal picked up his work documenting an 18% decrease in electricity consumption in Italy.
The new work, focused on the United States, is particularly compelling because it uses data that allows him to distinguish between residential, commercial, and industrial sectors. This turns out to matter a lot.
Without further ado, here are four facts he uncovers about Covid and U.S. electricity consumption.
Fact #1: Firms Are Using Less
U.S. commercial electricity consumption fell 12% during the second quarter of 2020. U.S. industrial electricity consumption fell 14% over the same period.
This makes sense. The second quarter was by some measures, the worst quarter for the U.S. economy in over 145 years!
Economic activity shrank. Schools closed. Offices closed. Factories closed. Restaurants closed. Malls closed. Even health care offices closed as patients delayed going to the dentist and other routine care. All this means less heating and cooling, less lighting, less refrigeration, less power for computers and other office equipment, less everything.
The decrease in the industrial sector is a little more surprising. My impression had been that the industrial sector had not fallen as far as commercial, but the patterns for both sectors are quite similar with the decline peaking in May and then partially rebounding by July. The paper also shows that areas with higher unemployment rates experienced larger declines in both sectors.
Fact #2: Households Are Using More
While firms are using less, households are using more. U.S. residential electricity consumption increased 10% during the second quarter of 2020. Consumption surged during March, April, and May, and then leveled off in June and July – with much less of the rebound observed on the commercial/industrial side.
This pattern makes sense, too. In Professor Cicala’s words, “people are spending an inordinate amount of time at home”. Many of us switched over to working from home almost immediately, and haven’t looked back. This means more air conditioning, more running the dishwasher, more CNN (especially last week), more Zoom, and so on.
The paper also examines the correlates of the decline. Areas in the U.S. where more people can work from home experienced larger increases. Unemployment rates, however, are almost completely uncorrelated with the increase.
Fact #3: Firms are Less Peaky
The paper next turns to a novel dataset from Texas that makes it possible to measure hourly electricity consumption by sector.
As the figure above illustrates, the biggest declines in commercial/industrial electricity consumption have occurred Monday through Friday between 9AM and 5PM.
The dashed line shows the pattern during 2019. Notice the large spikes in electricity consumption during business hours. The solid line shows the pattern during 2020. Much smaller spikes during business hours.
Fact #4: Everyday is Like Sunday
Finally, we have what I would like to nominate as the “Energy Figure of the Year”.
Again, start with the pattern for 2019, reflected by the dashed line. Prior to Covid, Texas households used a lot more electricity on Saturdays and Sundays.
Then along comes Covid, and turned every day into the weekend. Residential electricity consumption in Texas during business hours Monday-Friday is up 16%(!).
In the pattern for 2020, it isn’t easy to distinguish weekends from weekdays. If you feel like weekdays and weekends are becoming a big blur – you are not alone.
Researchers are increasingly thinking about electricity consumption as a real-time indicator of economic activity. This is an intriguing idea, but Professor Cicala’s new paper shows that it is important to look sector-by-sector.
While commercial and industrial consumption indeed seem to measure the strength of an economy, residential consumption has been sharply countercylical – increasing exactly when people are not at work and not at school.
These large changes in behavior are specific to the pandemic. Still, with the increased blurring of home and non-home activities we may look back on 2020 as a key turning point in how we think about these three sectors of the economy.
More broadly, Professor Cicala’s paper highlights the value of social science research. We need facts, data, and yes, science, if we are to understand the economy and craft effective policies.