In October, we described an experiment that tried peer pressuring Americans into conserving electricity during peak hours. (Here’s why that’s important.) On high-demand days, a company called Opower telephoned people to remind them to cut back on their energy use. Each time, it followed up with e-mails telling people how much they saved compared with their neighbors.
Perhaps it was peer pressure, or perhaps it was the telephone calls themselves that did the trick. Either way, people on those days used almost 3 percent less electricity on average, which is a pretty big decline as far as these things go.
Recently, the results of a similar experiment in Japan came out. That program posted similar numbers, but with an even milder nudge.
All they did was ask politely.
“If you go back in the history of energy policy, the most, kind of, popular government intervention has been pure moral suasion,” said Koichiro Ito, an assistant professor at Boston University and one of the study’s authors. “Just asking people to please conserve your energy.”
The lesson from Japan
In the aftermath of the 2011 Fukushima earthquake, most of Japan’s nuclear power plants were shut down and the nation’s grid was running dangerously close to capacity. Public campaigns urged citizens to turn off light bulbs and crank down their AC units. The efforts were “ambitious and strikingly successful,” The New York Times Editorial Board raved in September that year.
To an economist like Ito, this highlighted just how effective simple messages could be. There are, of course, programs like peak pricing, which have been shown to reduce energy use by charging customers more during peak hours. But Ito and his colleagues wondered if polite requests could be anywhere near as effective.
Beginning in 2012, they signed up 1,659 Kyoto households and randomized them into three groups. The first group was simply asked to use less electricity during peak hours on certain days. The second group was assigned into peak pricing: they would be charged double to quadruple the price for power during those hours. The control households didn’t get assigned peak pricing, and received no messages about their energy use. Everyone got a smart meter to measure their electricity use, and an in-home display to show them how much electricity they were using.
The researchers ran the experiment in the summer of 2012 and the winter of 2013. They looked at the weather forecast to identify peak days when electricity usage might spike. They sent text messages the day before, and the morning of the peak day. The reminders also showed up on the in-home display.
Some households got a polite message like this:
“In the following critical peak demand hours, please reduce your electricity usage: 1 pm – 4 pm on Tuesday, August 21.”
Others were told they would face a peak price:
“In the following critical peak demand hours, you will be charged very high electricity price, so please reduce your electricity usage: 1 pm – 4 pm on Tuesday, August 21. Price will be 85 yen per kWh.”
The customers who were surcharged during the peak hours responded just as researchers thought. Depending on the price, households reduced their afternoon energy use by 15-18 percent compared to the control group. These results are on par, Ito said, with similar experiments conducted in the United States. People respond to prices, especially if they are constantly being told about the higher prices.
But the polite reminders also proved effective — surprisingly so. There was no financial incentive here, yet households that received the basic reminders still reduced their afternoon energy usage by around 3 percent compared to the control group. This is about the same percentage that households conserved during Opower’s peer-pressure experiments, but without any of the peer pressure. All Ito and his colleagues did was make phone calls.
The power of the polite reminders, though, seemed to diminish over time. In the first couple of days, people cut back by a whopping 8.3 percent on average. (This is a big number. At their best, Opower’s experiments only ever saw a maximum reduction of 5 percent.) By the end of the summer, though, the text messages weren’t working at all. (The 3 percent average comes from averaging over the entire season.)
The researchers waited to try once more that winter, when they hoped the novelty had returned. Again, at first, people cut back by 8.3 percent — and again, after a couple of times, they started to ignore the reminders. The average reduction for the entire winter was about 3 percent. As before, most of those reductions happened in the beginning of the season.
Households facing the surcharge didn’t exhibit this problem of attention fatigue. Every time the price spiked, they responded by cutting their energy use, no matter if it was the beginning of the season or the end. Even three months after the program halted, these households still were using about 7-8 percent less electricity during peak times. Somehow, the higher prices had taught them the lesson of using less power.
“This is one of those papers where you’re like ‘shoot, why didn’t I think of this?’” quipped Steven Puller, an associate professor at Texas A&M, after Ito presented the paper at the American Economic Association conference in Boston on Monday.
Puller said one key question is whether these results can be replicated outside of Japan. “How much of this is institutional or cultural?” he said in a phone call on Wednesday.
It’s really anybody’s guess, Ito said. On one hand, we might think that Japanese households at this time were particularly amenable to energy conservation requests given the trauma of Fukushima. If the experiment were repeated in the United States, customers might be less responsive.
On the other hand, he said, Japanese households were already cutting back a lot, and this experiment asked them to conserve even more. In a less energy-conscious place like the United States, there may be more potential for people to improve their energy habits.
Regardless, what Ito and his colleagues discovered is somewhat remarkable. Simple, polite text messages, sent at the right time, can cause people to decrease their energy use by up to 8 percent. That’s nothing to scoff at, in any context.