Skip to content

Knowledge is (Less) Power.

Economists love prices. In fact, when we want people to reduce their consumption of something, we argue we should make that thing more expensive. This is usually done via a tax. My undergrads know, of course, that how effective a tax is at reducing consumption depends greatly on how elastic demand is. What we have learned from literally dozens of papers is that the residential demand for electricity is not very elastic at all. This means that the tax required to significantly reduce electricity consumption in the short and long run is probably pretty high (even more so if we carefully think about block rate pricing and how rates are set in practice). The literature has focused almost exclusively on measuring the price elasticity of demand.

Enter David Rapson and Katrina Jessoe. In their new working paper “Knowledge is (Less) Power: Experimental Evidence from Residential Energy Use” they describe a field experiment they ran in Connecticut. They randomly selected customers into a peak pricing experiment, where they would send households emails, text messages or phone calls informing them that their price of electricity would rise sharply for a few hours the next day. While these peak pricing experiments are not new, they gave half the group in home displays, which informed customers about their electricity consumption in real time. They of course had a control group of households who neither received peak pricing, nor information.

Their findings made my big neoclassical heart skip a few beats from excitement. Households exposed only to the temporary price increases reduced their demand by 0 to 7 percent, which is a very small response given the magnitude of their price increases. However, for the households with the in home display they observe reductions of 8 to 22 percent. They go to great lengths to show that this difference in response is consistent with  learning about one’s consumption patterns. Their evidence also shows signs that once customers have changed their consumption patterns, they stick to them.

Why am I such a big fan of this work? Information is power. We know from a number of Opower type studies that shaming reduces consumption slightly. Matt Kahn told us that liberals can be shamed into conservation more easily than conservatives. From a policy perspective this is of limited use as we neither could nor should assign political preferences in the name of energy conservation. What we can do however, is provide people with real time consumption data. That fancy PG&E smart meter in the back of my home should be unlocked so I can have an app on my phone tell me in real time how much electricity and gas my house is using. The cost of that app would be lower than the $40 in home display and have a payback period of weeks or months rather than the 7 years of the display.



Maximilian Auffhammer View All

Maximilian Auffhammer is the George Pardee Professor of International Sustainable Development at the University of California Berkeley. His fields of expertise are environmental and energy economics, with a specific focus on the impacts and regulation of climate change and air pollution.

15 thoughts on “Knowledge is (Less) Power. Leave a comment

  1. Good work. I tried for a while to get EBMUD to give customers
    more information about current water use, but nothing has happened
    with that.


      • Its bad enough getting bi-monthly (every other month) water usage from the water utility, I doubt they are setup to do ‘instant’ usage smart metering yet. there was a leak in the main line near the shutoff valve on my just-moved-into-older house. It took 4 months (2 bills) of 75ccf/mo for 2 people at no yard watering (rainy season) to figure out that there HAS TO BE a leak. What a waste of water AND of course money. Why would the water company tell me that there was a leak, they make more profit when the usage (incl wastage) is higher. Really, the same goes for all electricity except the very peak usage periods in summer.

  2. We know from marketing behavioral science that consumers are motivated by many things besides price. Every marketer learns the Four P’s: Product, Place, Promotion, and Price. An advertising maxim says that consumers respond to three things: fear, greed, and sex. There are many other factors including convenience (why do people buy water bottles?) and prestige (why do people buy Rolls Royces?). The supply and demand curves of economics only tell us about price.

    In-house displays communicate information, but I suggest that the underlying behavior trigger is *fear*, exceptional self-selecting pilot customers notwithstanding. That is powerful motivator, but does not result in a positive experience for most customers in the long term. It has been suggested that the poor “benefit” more from dynamic pricing, because their fear threshold is lower with respect to pricing. When consumers are given a real choice in the form of competition as in ERCOT, retailers almost universally are marketing the “peace of mind” of fixed-rate or TOU pricing to acquire new customers. That doesn’t address the underlying service delivery problem of the utility, however.

    Consumers don’t actually consume electricity, machines do. If certain communicating machines with timing elasticity automatically shaped their collective load and yielded to higher-priority demand, the puzzle of how to get humans to care about their power consumption timing may become moot. A similar problem was solved in telecom by requiring devices to conform to “automatic congestion control protocols”, and today we have neither busy signals nor an in-home display that tells us the current price of making a phone call. The machines automatically figure out how to share the resource in an orderly way instead of blindly competing for it.

    • So if there was a real price signal to limit demand, it would not be that hard to develop a control system to turn off the major appliances at the appropriate time to limit demand. There rest of the time, energy could be used at very little cost.

  3. I think knowledge of real time energy and prices would be helpful. But more important is the demand component. Most of the year, energy is not very expensive and there is very little difference between on and off peak energy prices. However, we build power plants (or expensiver battery systems) for 4-8 hours per day, 10-20 days a year. Spreading the cost of peaking units over relatively few kWh is expensive. This will be even worse in time as renewable energy in the form of wind and solar plants increase. Soon (2016-2018) power requirements from conventional resources (ie non- wind and non-solar) will peak twice a day for 4 hours each. There will be a lot of fixed costs over relatively few kWh. Ultimately, the key inexpensive energy will be to mimimize those peaks cost effectively. And to the point, send a proper price signal to limit demand, not energy.

    • Jim, I agree that what we would want is real time pricing and full information. But for now I will take an unlocked real time meter. Max

  4. Actually have a meter IN the living space where one can see the usage in kwh/ $ for the last 5 minutes would be even more helpful. This applies to both electricity-gas as well as water usage. There are times where one might forget about having left an oven on, for example, or a faucet running, or even a faucet-line leak of water. With smart electric and gas meters this should be easy, just a matter of the utilities-smart meter manufacturers allowing a secure access to the data, may even be wired-in rather than wifi (for those who are concerned about privacy-security). For water one would need an attachment to the meter (convert it to smart meter) or something inline just this side of the meter. Both these can be sued by crooks to determine if someone is not at home, and break in … but ‘thems the breaks’.

  5. it is pretty outrageous that the CPUC has not required PG&E to unlock the local access to the smart meters – bill payers paid billions and the information is locked up that we, and application developers are waiting to access.

  6. PG&E doesn’t yet have an app for the real time display, but according to the PG&E April 2012 news release, customers have easy access to some data.
    “SAN FRANCISCO, Calif. — Pacific Gas and Electric Company (PG&E) customers can now share and compare their energy use with friends around the nation on Facebook with a new social energy application. The app, created by Opower in partnership with Facebook and the Natural Resources Defense Council (NRDC), provides insight into individual energy use while fostering friendly competition.

    To get started, PG&E customers can visit or search for the “Opower” app on Facebook and connect their utility account.

    Customers who sign up will see how their home energy use compares to others with similar homes. As friends are invited and join in, people will then be able to engage in savings competitions. PG&E customers with a SmartMeter™ and My Energy account can get a leg up on their competition by logging onto to view hourly electric data, gaining insight to further conserve energy.”
    See the PG&E news release at this link for the full story:

    • Hi Betsy, I like the Opower app on facebook and the monthly mailings. But this is a different program in that it compares my monthly consumption to that of my neighbors’. It works. I have ben shamed into becoming more efficient. Also, on the PG&E website I can look at yesterday’s consumption. What is lacking is real time consumption information. You can get it by having someone stand at the meter behind your house, while someone else turns on and off appliances. This is more than most people are willing to do and the information on the meter itself is not presented in a very intuitive fashion. A simple app on the mobile phone or in home display which links to your smart meter would be optimal. Thanks for reading! Max

%d bloggers like this: