Now Reading: Smart Disclosure, Behavior Change, The Green Button

This is an article co-written by Richard Thaler (of “Nudge” fame) and Will Tucker about something called “Smart Disclosure,” which is the US Government term for a concept that Dick advanced in his 2008 book. In the UK, it’s called “midata” and it is:

the “timely release of complex information and data in standardized, machine-readable formats in ways that enable consumers to make informed decisions.”

The basic idea is that the way people are given data affects the way they make choices. This is leveraged wittingly and unwittingly by industries to influence choices that are unfavorable to the individual. This less-helpful behavior by industry is enabled by data that’s locked away or not in machine readable formats that can be presented in an informative way.

I’m obviously familiar with health care data, but I enjoy reading articles like that that take me on a journey into another industry where I can be less tied to my own biases and considerations.

Green Button, for energy

In this piece, that other industry is energy. I had heard about the Green Button but never used it. It’s what the Blue Button is for personal health care data, except for personal energy data. And sure enough as I was able to log in, my utility provider, PEPCO, has implemented it.

I checked it out, again, from the perspective of a customer in an industry I don’t work in, and here’s what I found.

Benefit of Green Button is not Green Button itself, it’s the transparency that Green Button creates

Bad news: When I went to go find apps to process my Green Button data, I was confronted with an array of choices, some unfinished, some not connected to my utility (why, I thought this data is standardized), many not-free, not working, or asking for authentication information that I don’t want to give just to review my energy data. That’s the bad news – if someone can suggest an easy app to process my green button data that doesn’t cost anything, please post in the comments.

Good news: What my utility bundles with my Green Button, is pretty remarkable near real-time information about my electricity use, and I think that’s more than “good enough,” without Green Button download needed. I can see the peaks and valleys of my energy use not just last month, but yesterday. That’s kind of remarkable.

Bad news: The home screen of my electricity use doesn’t really use the best science in comparing me to homes in my neighborhood. It says I’m doing an amazing job. It says, “Congratulations! Your home used less energy than most of the similar homes in your area.” From what I’ve read, saying this to a customer may actually increase their energy use. Better to not say anything unless energy use is significantly higher. Just following what the science says…

I think overall this is a great innovation and it parallels what I think the outcome will be in health care, that the work to process the data is not as significant as the impact of making the data available in the first place, so again, I’m not heartbroken over the lack of easy app access for Green Button, I’m happy that it’s changed the culture.

The promise and the peril

With information display technology, there’s a dark side, which the authors point out:

A bad outcome would be if we were to simply switch the source of obfuscation from seller to choice engine. A primary principle of regulation should be ensuring that choice engines’ business models are transparent. For instance, consumers should be informed if a choice engine gets commissions from service providers. 

This is the part where I worry that my machine-readable XML data can be parsed into anything the parser wants it to be parsed into, and then what decisions am I going to make?

Even if the choice architecture created is accurate and maximizes presentation in my best interest, it’s a little disheartening to read in the same issue of Harvard Business Review that “people are less likely to choose the highest rated option in a quality ranking if it appears first on the list” – they are most likely to choose the highest ranking hospital when it is in the middle of a horizontally presented list (see: It’s Not Necessarily Best to Be First – Harvard Business Review). Strange.

All of this put together makes me less concerned about the machine readability and parsability, more interested in that this process makes the data available.

I still hear people in health care say, “we don’t want to give our patients their data until it’s presented in an understandable way,” and I think it’s a wonderful ideal, however you can’t present the data in an understandable way until you get to see the data at all. I think that’s the most important point in this work, data first, presentation later. If presentation doesn’t come or come perfectly, keep delivering the data, because when people think of this the other way around, somehow the data never gets shared. Our patients/customers are smarter than we think they are, I would rather them work harder to draw conclusions than have no information with which to make decisions at all.

1 Comment

Ted Eytan, MD