In New Orleans, like a lot of places, a fair share of low-income residents who are eligible for subsidized health insurance don't take full advantage of it. The city decided to try a social science experiment to see if there was a way they could encourage more people to visit a doctor or other caregivers.
The city thought people might be more likely to make appointments if they received reminders via text messages. Behavioral scientists from a program called What Works Cities designed three different text messages (below), each one going out to about 7,000 people.
These three texts, each one going out to about 7,000 people, were part of a social science experiment to encourage more people to see a doctor. (Behavioral Insights Team)
They didn't get a huge response, but at least one of the messages -- the "ego" approach -- encouraged a few dozen more people among the thousands enrolled in Medicaid to actually take advantage of free medical appointments.
The value of this exercise for the city was figuring out exactly what types of messaging actually resonate with people, and what doesn't. New Orleans is already running similar experiments in other program areas. Just thinking in terms of running such tests and evaluating quickly what works is a useful endeavor for a city.
New Orleans is one of 21 cities participating in What Works Cities, a project financed by the Bloomberg Philanthropies, the foundation started by former New York City Mayor Michael Bloomberg. One reason cities are joining is the free access to the Behavioral Insights Team, a private firm that specializes in taking research on human behavior and using it to improve the operations of a government program.
This kind of experimentation is catching on. The Obama administration has funded a series of “nudge” experiments designed to improve outcomes in child care, child support and other human services programs.
New Orleans is the first to announce results from its What Works Cities experiment. Oliver Wise, the director of the New Orleans Office of Performance and Accountability, spoke with Governing about the results. The interview has been edited for clarity and length.
What was the reaction in your office when you saw that even the most successful intervention generated responses from maybe 70 people, or less than 2 percent of everyone texted?
In any text message campaign, or with direct mail, you often get fairly low response rates. But in the event that you are already trying to do one of these campaigns, you might as well inject some behavioral insights into them, and you might as well test them in a rigorous way so you can build off that for next time. I don’t think this text message campaign is going to, by itself, change the culture of how New Orleanians seek health care. But it represents incremental progress and a whole lot of these incremental nudges can lead to a culture change.
Do you think that you’ll continue to try nudge experiments? Would you be willing to do it with city funds as opposed to philanthropic dollars?
Absolutely. This experiment has opened our eyes. Government communicates with residents all the time, and we do it because we want something to happen. If we can inject some experimentation as we’re doing it, just to see if we’re getting a little better each time, then we should do that.
Doing these kinds of trials allows you to see which ones resonate with your residents. It gets government to do a better job of thinking through what services look like from a resident’s perspective, not just a provider’s perspective.
How might you use nudge techniques in the future?
We have similar experiments going on in other areas now. One is a text message to parents of kids who participate in our recreation department’s programming to see if we can compel them to volunteer as coaches.
We’re also trying to recruit more police officers. What we’re finding is that only about 20 percent of police officer applicants actually show up to their first required test, so we’re trying different email messages to applicants to see which ones are better at getting people to take their civil service test.
We have another trial going on in code enforcement where we are sending warning letters to owners of properties that have 311 license plates. We want to see if we can get them to comply before bringing them through a lengthy legal proceeding.
This kind of incremental work is very different from what I’ve seen with the mayoral innovation teams, another Bloomberg Philanthropies initiative, where the city sets big goals like drastically reducing homelessness or the homicide rate in a few years.
I think incremental innovation is how 95 percent of innovation occurs. The tech sector is about that, too. It’s not like Microsoft 1.0 is the breakthrough. It’s the yearly update and learning from users what works well and what doesn’t, and making iterative tweaks based upon that.
In terms of this type of work, yes, this is an incremental improvement in itself, but a local government having the capacity to pilot different interventions and then rapidly test and evaluate the responses to those pilots is in itself a breakthrough innovation.