Skip to content

One university’s attempt to reduce energy waste at work

If you work outside your home, chances are you don’t pay (directly) for the energy you use at work. At my place of work, the UC Berkeley campus, most employees never see – let alone pay – their energy bills.

Of course, there are plenty of pro-social reasons to be conscientious about my energy consumption at work (climate change and tight university budgets, to name a few). But these “split incentives” (i.e., the fact that I bear none of the costs when I increase campus energy use) beg the question: How much less energy would we use at work if we were all responsible for paying our own energy bills?

windows Source:

This seems like an important question when you consider the quantity of energy consumed each year by commercial buildings (which include office buildings, retail space, restaurants, hotels, hospitals, schools, and universities). The commercial sector now accounts for over 18 percent of total U.S. energy and close to 40 percent of U.S. electricity use.



Do split incentives cause energy waste at work?

A recent paper by Matt Kahn, Nils Kok, and the late John Quigley sheds some light on this question. These authors track the electricity consumption of a large sample of commercial buildings in the Western U.S. In particular, they examine the association between lease incentive terms, occupant characteristics, and electricity consumption in commercial buildings.

They find that commercial tenants whose utilities are bundled into the rent consume significantly more electricity than tenants who pay their own bills. They also find an increase in occupancy by government (versus private sector) tenants is associated with a significant increase in the energy consumption. The authors suggest this correlation could possibly be explained by the fact that government tenants face relatively soft budget constraints.

We can’t be sure that these correlations indicate a causal relationship. But if split incentives in the commercial building sector are associated with higher energy consumption, could an un-splitting of these incentives lead to cost-effective efficiency gains? An amazingly dedicated team of energy efficiency enthusiasts here on campus set to find out.

Un-splitting incentives on the UC Berkeley campus

Back in 2012, energy costs were rising and there was a general sense that energy was being used inefficiently across campus. Campus electricity bills – on the order of $20 million annually – are managed by central campus. How can you motivate individual employees to focus on energy efficiency when they have no direct financial incentive to do so?

The Energy Incentive Program (EIP) was introduced in April 2012 to provide individual departments and units with the information – and the financial incentive – to make cost effective changes to their energy consumption. Each operating unit was assigned a baseline for each building based on electricity use in academic year 2010-2011. Units are rewarded (or charged) 10 cents per kWh for consuming below (or above) their baseline. Importantly, this financial incentive is (approximately) equal to the price the university currently pays for electricity. Units were also provided with real time feedback and improved support for building managers.

The figure below plots electricity consumption on the Berkeley central campus (in blue). The figure shows a notable drop in campus electricity consumption in the first two years of the program (FY 2012-2013). To put these trends into some sort of context, the red line plots student enrollment over the same time period. Electricity consumption has dropped below 2007 levels while student enrollment has increased. Building square footage has also increased since 2011.


(Huge thanks to Kevin Ng and Lisa McNeilly for providing UC Berkeley central campus electricity consumption data.)

This simple graph does not prove that the Energy Incentive Program caused the coincident drop in energy consumption. To credibly estimate a causal impact of the program, we would need to take much more care in constructing an estimate of what UC Berkeley electricity consumption would have looked like absent the program.

A comprehensive program evaluation is well beyond the scope of this blog, but I can offer some anecdata. Although many faculty and staff are oblivious to the incentive change, the program is on the radar screen of the people who manage department budgets and buildings. Every manager I spoke with could point to multiple projects – ranging from email reminders about phantom loads to major lighting retrofits – implemented to keep energy consumption below baseline. Office managers who had been uninterested in building maintenance before the program now closely monitor their daily energy consumption and alert facilities managers when something looks amiss. Much of the $1,869,200 paid out in incentives to date has been re-invested in energy or water efficiency improvements.

Funding for the Energy Incentive Program is scheduled to decrease this year. I hope the incentive/charge per kWh does not. The program budget could instead be balanced by reducing energy consumption baselines in a way that does not disproportionately penalize departments that have made major efficiency investments. Given tight university budgets and ambitious environmental goals,  it is as important as ever to get campus energy prices right.



5 thoughts on “One university’s attempt to reduce energy waste at work Leave a comment

  1. I question the relative value of the program you described on two counts. First, is this the best use of staff time when a cost/benefit analysis of all the different things they could be focused on taken into account? Second, is this the most efficient way to achieve this level of energy reductions? Without these relative measures all reductions at any cost can appear to be progress.

    A couple examples of things done where I work:
    All the lights shut off at 6:00 pm. If I’m working late I need to call building services to have them turned back on in my section of the floor for a defined period of time. I appreciate this type of approach to conservation since it addresses a large part of the problem and allows me to maintain my focus on the great work I’m paid to do.

    We have an annual month-long waste reduction contest. The objective is to change employee behavior to better sort our trash. A core employee team works with a couple individuals on each floor to spread the word and the custodial staff to measure and report results. The winner gets to display the trophy (a beautiful platter made of sea glass with the names of past winners stamped on bronze placards around the edge) for the following 12 months and enjoy a pizza lunch for the floor. This extremely low-cost program has produced some very impressive results both in improved performance at the top end and clustering the pack at much higher levels of performance.

  2. Good article. The split incentive problem exists in at least 3 different forms — commenter Janice noted a second–each with very different mitigation strategies. The International Energy Agency examined these principal-agent problem in a 2008 book, “Mind the Gap: Quantifying Principal-Agent Problems in Energy Efficiency” ( ). It sought to quantify the amount of energy consumption that was in some way insulated from the price signal, from office buildings to vending machines to TV set-top boxes. Our friends at Comcast “rent” us cable boxes but we pay for the boxes’ electricity use; that’s why they make excellent pizza-warmers.

  3. How much natural gas does Berkley purchase to heat these buildings and heat the domestic water and the swimming pools?
    How much of this purchased and combusted natural gas is being wasted, blown up the chimneys as Hot exhaust into the atmosphere?
    The US DOE states that for every 1 million Btu’s of heat energy that is recovered from this exhaust, and utilized back in the building or facility, 117 lbs of CO2 will Not be put into the atmosphere.
    There is Water in combusted natural gas, and with the heat energy removed this water is created, and this distilled water is very useable.
    California is a leader in Energy Efficiency but recognizing wasting natural gas is not as quickly recognized as leaving lights on in an empty room.

    What natural gas is not wasted today, will be there to be used another day.

  4. Another good post. This is relevant on our campus, too. The challenges and obligations are different than for other utilities. However, elasticities, opportunities, and ownership responsibility still matter. In our small office, we already use efficient equipment and have good habits. But we also have no thermostatic controls and outdated windows that belong not to us but to the building. Proper alignment of incentives remains key.

%d bloggers like this: