Power of Measurement and Transparency

Two of my experiences and an article in the Thursday, July 9, 2009, Wall Street Journal entitled “Gadgets Show How Much Power Your House Eats” have convinced me that measuring and disclosing information is highly likely to change behaviors.

One of my experiences was a lunch at a popular New York City restaurant on July 8. When I looked at the menu, every item was labeled relative to the calorie content of the item. I was ready to order the caesar salad as a low-calorie alternative, when I was shocked to find out that the salad contained 790 calories, and that the salad with chicken was 1,325 calories. To my surprise, the 10-ounce filet mignon was only 390 calories, so I ordered it and also had a cup of soup. Undoubtedly, absent the disclosure, I would have ordered the salad.

I also wear a pedometer every day and almost always am able to walk or run sufficiently to get to 10,000 steps per day. On July 8, on my way to the restaurant, I was taking a subway from East 86^th Street in New York to Grand Central Terminal at East 42^nd Street. While I was on the subway, I noticed that I had 20 minutes to spare when the train pulled into the 59^th Street station, and that I needed to log a number of steps. As a result, I exited the train at 59^th Street and walked the rest of the way, because I saw that it would be difficult for me to get to my target.

At Pitney Bowes, we took one further action relative to marketing the benefits of walking. At the bottom of our 3^rd floor stairway in our World Headquarters, we have a sign that informs someone that if they walk up and down these stairs every for a year instead of taking the elevator behind the stairs, they would lose 5 pounds. I watch people make the discretionary decision to take the stairway rather than the elevator.

The Wall Street Journal article notes that the same behavior occurs relative to electricity usage when a home has a power monitor that informs the home owner minute by minute how much power has been consumed. The author, Geoffrey Fowler, cites an Oxford University study in 2006 that found that “people getting direct feedback on their power consumption reduced use 5% to 15%.

There are two implications to these data points:

• If we want to reduce overeating or to increase exercise, measuring, monitoring, and disclosing the quantitative aspects of a behavior will change the behavior.
• Conversely, if we want to change an unconscious behavior by altering the environment that produces it, that will be successful as well. Eating, exercising, and other behaviors, good and bad, have a heavy unconscious, automatic aspect to them.

Reflecting on my behavior in the restaurant and on the subway, I am more convinced than ever that mandatory nutritional labeling works if it is quantitative. Warnings like the Surgeon General’s warning on cigarette packages are much less effective because they are not quantitative. Similarly, disclosures on prescription drug packages relative to side effects are also relatively ineffective because they are not quantitative.

My good friend Dr. Elliott Fisher of Dartmouth’s Health Policy Institute introduced me to a different form of disclosure relative to prescription drugs, a one-page chart that lists every side effect, but that specifically supplements the disclosure by listing in quantitative fashion the results of the clinical trials conducted for that drug. Thus, for example, the chart does not simply say that Drug X has been shown to cause headaches in some people. It specifically discloses that of the 2,500 people who took the drug, 17 of them, or .68%, experienced headaches. There are some indications that individuals confronted with quantitative information react differently. Some pay attention to the disclosure and decide they do not want to take the risk; others are reassured by a low-percentage risk disclosure and decide they will take the risk. In both cases, the quantitative disclosure altered behaviors.

Because of all these experiences and observations, I have become a strong believer in more detailed labeling and disclosure, as well as much more quantitative disclosure, whenever such disclosure is not false or misleading.