Skip to content

Nudging Pitt Students (Pt. 2 – The Results)

September 26, 2011

In the previous post I talked about the idea behind the LANDFILL label, and why I thought that it could affect behavior. In this post I will talk about our actual experimental results.  Below is a picture of our label on a campus trashcan (click to enlarge):

The Experiment’s Format

  • Observed whether specific recyclable items were trashed or recycled
  • Observed 2 trashcans in separate areas, one adjacent to a recycling bin (“Store Area,” above), the other close to, but not next to, two recycling bins (“To-Go Area”)
  • 4 weeks of data collection (M-F, 2 hours each day)
  • First 2 weeks = control, raw observational data (no LANDFILL label on trashcans)
  • Last 2 weeks = observational data with LANDFILL labels on trash cans
The Results

The greatest part about this experiment involved the fact that none of our subjects had any idea that they were actually participating in an experiment.  This is an issue that experimental economics labs tend to have (people that choose to participate in a study may be more likely to respond to the study’s queues than an average person, for example).

Additionally, the most rewarding part of the experiment was actually being able to see individual reactions to the LANDFILL sign.  And while we did not collect any data on this, we noticed that females were significantly more likely to respond positively (stop to think about their action, or recycle) to the new LANDFILL sign than males.

Here are four graphs of our results (there are eight in the full study – I’m only displaying four to keep this post manageable):

Figure 1 details the base (control – weeks 1 and 2) recycling behavior, because this data was collected before the LANDFILL label was added to the trashcan.  In Figure 2, the data collected was from weeks 3 and 4, and shows a slight yet distinct increase in recycling rates.  In this example specifically, recycling rates increased by 26% after the addition of the LANDFILL label.

Figure 7 details the base behavior with respect to the number of recyclable items that were thrown away.  Paper dominates the graph because the trash can was in the “Store Area” (small convenience store located in the dining hall), and every student that bought something received a receipt, which was typically either thrown away or recycled upon exiting the store.

Similar to Figures 1 and 2, a slight change in behavior is visible.  In this specific example, the rate of discarding recyclable items decreased by 27% after the addition of the LANDFILL label.


  • Overall recycling rates increased by 29%
  • Overall discarded recyclables decreased by 23%

In a future post I’ll take time to provide some analysis, but for now I simply wanted to place all of the data into one, somewhat concise post.  The basic result was that the LANDFILL sign affected behavior, and led to an increased recycling rate (with some limitations).  In future experiments, we will try to tease out the specific effect that the IMPERIAL, PA (17.3 MILES) label has.  I have a feeling that attaching regional responsibility to the label was very important.

– James

Note: Special thanks to Richard Thaler and Cass Sunstein, the authors of the 2009 book “Nudge.” Without their inspiration I may never have founded this organization.  Also thanks to John Balz, the great editor of Nudge’s online companion,, and tolerator of all of my dumb tweets.

From → Experiments

  1. Steve Willard permalink

    Just stumbled across this via that FastCompany article – no wonder you were featured! Quite a stunning experiment, and a great result! I wonder if that had anything to do with the demographics of your subjects?

  2. This is a great experiment. Someone in a partnering lab is literally in the process of doing the same type study of with a landfill trash can next to a normal recycling one.

    What might be great if you decide to try this again is doing what’s called a multiple baseline experimental design. Essentially, you’d pick three areas to put the trash cans instead of two. Then, you start the baseline (or no landfill trash can) for the first location for say two weeks and then put in the landfill sign. While this is happening, you take baseline on the second location until say 3 weeks until you go into the experimental condition. Then, say four weeks of baseline for the last location before putting the landfill sign on the next can.

    I personally don’t think the demographic had any effect but with a multiple baseline you can help show that you had control as the experimenter in changing behavior.

    Great post!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: