Do Smart Thermostats Save Energy? New Research Makes Surprising Findings

The report says it may actually increase energy consumption.

Smart thermostats

Marvin Samuel Tolentino Pineda / Getty Images

It's that time of year when Las Vegas is full of everything electronic at CES (Consumer Electronics Show), and every site is pitching smart home stuff. Many of these devices promise energy savings, with Treehugger noting that "the U.S. Department of Energy says a smart thermostat alone can save 10% a year on heating and cooling."

However, a new study, "The Human Perils of Scaling Smart Technologies: Evidence from Field Experiments," comes up with a different result. Where most previous research used information from the companies producing the smart thermostats, the researchers of the aforementioned study found 1,365 households to track and outfitted half of them with Honeywell two-way smart thermostats. They gathered data for 18 months, collecting 16 million electricity use records and 7,000 daily observations of natural gas consumption.

"Our experimental estimates provide several insights into whether the petri dish estimates of engineers hold when technology is scaled beyond the lab. First, we find that smart thermostats fail to deliver the expected energy savings; our results show that such technologies have neither a statistically nor economically significant effect on energy use."

They found that, in some cases, smart thermostats increase electricity and gas consumption. Apparently, users frequently override the settings and when they do, "the override settings are less energy efficient than the previously scheduled set-point."

"We find that while some user types realize significant savings, engineering models fail to capture how the majority of people actually use smart technologies and this limits the usefulness of their estimates in real-world settings. More specifically, people adopt the smart technology but use its features in ways that undo the purported benefits, suggesting that human behavior is a peril to scaling such technologies."

The authors suggest that engineering models fail to account for how people actually use their smart thermostats and therefore represent an upper bound on potential savings. This has been discussed on Treehugger before. While Treehugger's Sami Grover loves his Nest thermostat, he admits the device had trouble adjusting to his family's schedule.

Honeywell, I'm home!


When I reviewed the Honeywell Lyric smart thermostat, which talked to your smartphone and raised the temperature when you drove home, I suggested that it was designed for a "Leave It to Beaver" family from the '50s where everyone followed the same schedules and dad controlled the thermostat from his car.

But people don't live like that now. Some work from home, some have kids that come home at different hours, and some don't have traditional families. Grover has pointed out the difficulty of scheduling when people are in and out. I had a friend who ripped her Nest Thermostat out of the wall, she got so frustrated with its stupidity.

But the fundamental problem with smart thermostats is that people want comfort, and the temperature of the air is just one part of what makes us comfortable. Comfort comes from a combination of air temperature and the mean radiant temperature (MRT).

I have written before that "many architects don't get it, mechanical designers don't get it (they will just sell you more equipment), and the clients don't get it," and "since there is always someone who will talk up the comfort potential of a smart thermostat or a radiant floor, it is hard to convince people that it is really all about the quality of their wall or window."

If your walls are cold, you will radiate heat from your body and feel cold. If they are hot, you will be hot, and occupants will adjust the thermostat to try and compensate.

I noted more recently that you can't get comfort from a thermostat; it's more complicated than that. People are expecting too much from smart thermostats—it's hard to fight the second law of thermodynamics.

Treehugger's Energy Conservation Pyramid


The study authors concluded that any funds directed toward smart tech should go elsewhere "on more promising ways to promote energy conservation and associated reductions in greenhouse gas emissions." I would suggest this should be directed toward air sealing and insulation, as per our pyramid of conservation, which consciously didn't include smart technology.

In 2014, when the Nest thermostat was introduced, I wrote about how, if a house were built properly, then there would be no need for smart thermostats. I was promoting instead the Passivhaus (or Passive House) design, where the temperature barely changes. I called it a "dumb home" and the term has caught on.

"Then there is the Passivhaus or Passive House. It's pretty dumb. A Nest thermostat probably wouldn't do much good there because, with 18" of insulation, and careful placement of high-quality windows, you barely need to heat or cool it at all. A smart thermostat is going to be bored stupid."

This most recent study shows that nothing has changed. Even in the era of heat pumps and "electrify everything", you still need "fabric first"—to stop the heat from leaking in or out. A smart thermostat won't do it for you.

Many of the studies we cover are difficult and technical, but this one is a joy to read and is more like an article in a magazine. I highly recommend it.

View Article Sources
  1. Brandon, Alec, et al. “The Human Perils of Scaling Smart Technologies: Evidence from Field Experiments.” NBER Working Paper Series. Sept. 2022, doi:10.3386/w30482