Participant Selection

CENTRAL CHALLENGE: How to select participants that are representative of the whole University, as the carbon charge will ultimately be extended to all buildings on campus?

Energy use is highly diverse at Yale due to different building uses, ranging from teaching and research facilities to residential colleges, athletic facilities, offices and public museums, among others. There are also differences in how buildings and activities are metered and how energy charges enter departments’ budgets.

Administrative units on campus differ with regard to their control over budgets and energy bills. “Self-support” units are those responsible for their own budgeting and fundraising, and therefore are energy ratepayers. These include six self-support schools and a few autonomous entities. Centrally-supported units, on the other hand, are not responsible for their own fundraising or utilities expenses.

Within the set of centrally supported units, there are larger units that, while not paying their own bills, are attentive to and have more control over their budgets. These include the balance of schools, the collections (e.g., museum and art center), the main library, athletics, and the officer units (e.g., Office of the President, Office of Development). Finally, there are units that are either small, have limited budget authority, or are co-located with other units, which makes estimating energy use and carbon charges difficult. These include small departments and the residential colleges. This last group represents the largest challenging for carbon charge application.

An important goal of the pilot was to test how the carbon charge would impact the whole campus. As a result a diverse set of buildings were selected based on three criteria: building and budget type, meter data quality, and carbon footprint.

Carbon emissions were calculated for 300 buildings on campus to establish quintiles of largest to smallest emitters. The 20 selected buildings were then categorized by building type (e.g., classroom, lab, office, dorm, etc.) and budget type (i.e., self-support vs. centrally-support), then assigned to the four carbon pricing schemes until each scheme had a building from each quintile and a profile of building and budget types that is representative of the whole campus.

The 280 remaining buildings served as the control group. It should be noted that an administrative unit can have multiple buildings, but only one building from each administrative unit was chosen, hence “buildings” and “units” were used interchangeably when discussing the pilot. More information on the 20 selected buildings is provided in the Pilot implementation at the building level section.