How Do We Solve the Last Mile?

This article was originally published on The Psych Report before it became part of the Behavioral Scientist in 2017.

You sign up for a free flu-shot but forget to schedule the appointment. A qualified high school student is accepted to college but gets tripped up applying for financial aid and doesn’t enroll. The government creates a great program to help low income families fund their children’s education, but only a handful of people sign up.

Some of society’s stickiest problems aren’t a failure of intention, importance, or value. They’re the result of a failure to understand human behavior at the last mile—the final stage where desires and plans must turn into action.

Last mile problems range from the frustrating—signing up for a gym membership and failing to go as often as you’d like—to the fatal—knowing how to treat diarrhea, yet seeing hundreds of thousands die annually because they misunderstand the reasoning behind the treatment.

In his new book, The Last Mile: Creating Social and Economic Value from Behavioral Insights, Dilip Soman, a professor at the University of Toronto’s Rotman School of Management, explains how governments, businesses, and individuals can use behavioral science to tackle the stubborn and persistent problems that trip us up so close to the finish line.

We spoke with Soman, who advises the behavioral science units in both the Canadian and United States governments, to discuss the steps organizations can take to understand and solve their last mile problems, the different approaches to behavior change, and areas where behavioral science can play a key role in finding solutions.

Evan Nesterak: Can you describe an example of a policy or product that ran into trouble at the last mile?

Dilip Soman: Something that was salient to me a few years ago here in Canada, our Canadian government introduced a welfare scheme called the Canada Learning Bond. Without diving into details, it was essentially $500 to eligible low income families, with the goal of educating your kids. When the program was being put into place, I vividly recall an economist saying, ‘Who would not accept the Canada Learning Bond?’ [The] take up rate should be like 100 percent. It turns out [take up was] only 16 percent for the first few years.

The challenge wasn’t that people didn’t want the money, it was a great product, but you needed a bank account, and a particular kind of bank account to accept the money. These low income people, for whom it was designed, didn’t have the time to go to the bank or didn’t want to go to the bank. So the solution wasn’t in promoting the bond or increasing the amount of dollars, it was actually making it easy for people at the last mile to sign up.

That in essence is what we mean by the last mile—the fact that people inside of the organization have a particularly hard time relating to what’s going to happen at the last mile.

EN: What can an organization do to understand and solve its last mile problems?

DS: We think about three different activities that organizations need to do to be really good at the last mile. Activity number one relates to the process of scanning the literature—making sure that you convert all of the insights from psychology and behavioral sciences into simple digestible forms. There’s a lot of stuff that’s been written, but organizations typically don’t ever see it. They don’t really know what it means because of the way our research is typically written and presented. So one activity is taking all of that literature and condensing [it] into simple behavioral insights written in the form of guidelines.

The other one is more of an auditing process. When does the customer first touch the organization? What information do they get there? What is the next stage? Can we start documenting what percent of people go from stage 1 to stage 2 to stage 3 and so on and so forth? Once we do that, we can start seeing where the problems are. For example, if your flow rate from stage to stage is consistently 90 percent, then all of sudden you reach one stage and half the people drop off, there’s something going on there. And that something could either be an issue with the process itself, or it could be an issue with the product. If it’s a process problem—if information is hard to find or there’s too many options being presented or you give the opportunity to procrastinate on making the decision—that’s something you can fix. Once you audit each of your processes, you can get a better idea of exactly where the bottlenecks are and start fixing them.

The third thing that you need to do as an organization is to be experimental. Try different interventions, document the success, and keep iterating with time. That’s one of the other things that I am very passionate about—making sure organizations, be they governments or for-profits, learn to be experimental in the psychology sense. [When] you talk to companies and you stress to them the need for being experimental, they are thinking of that word as meaning “Oh we try different things” and that’s not the idea. The idea is to do controlled testing, make sure you collect the data, and you’re very rigorous about the whole thing.

If you have an organization that does the translation process, that does the auditing process, and then does the intervention and testing process really well, to me that’s a fantastically well behaviorally informed organization.

EN: In your book, you explain how a lawyer, an economist, a marketer, and a behavioral scientist each approaches the task of changing behavior differently. Can you describe the different approaches?

DS: Suppose you want to get people to change behavior—you want to get somebody who is doing hypothetical action A to move to hypothetical action B. What are the different approaches to doing that?

There’s the restriction approach, which says you as an organization can simply withdraw A. If you make A unavailable, people have no choice but to move to B. Or as an economist you could think about the role of incentives. You could incentivize people to pick B or penalize them for picking A. We do this with taxes [and] with subsidies all the time. [Then] there is the marketing approach, which is all about information and highlighting the benefits of the product. Finally, there is the choice architecture approach. The idea that we could use insights from behavioral sciences [to] make sure the way in which information is presented is consistent with the mental model that people are using to address that particular problem.

I’m actually not about to suggest that any one of them is better than the others. I’m also not going to suggest we need to be very rigid with the way we operationalize behavioral change—that we should either be just economists or either be behavioral scientists. But I do think that it’s a useful way of thinking about different approaches to behavioral science.

EN: What do you see as society’s most solvable last mile problem? One that we haven’t yet been able to crack.

DS: To be honest, I don’t know what the most solvable one is, but I’ll give you some flavors of the genres of the kind of problems. I actually don’t think behavioral science itself is going to crack any major nut. I don’t think we are going to be able to change choice architecture and get everybody to stop smoking, get people to stop doing crazy things, or stop wars. I don’t think it will do any of that. But what we are going to do is we’re going to make differences at the margin and I think those are going to be important differences.

[In] education one of the big challenges is how do we get kids who complete high school to go to college. This is an interesting problem because throughout the first twelve years of our schooling, there’s always a default. There’s nothing you need to do, in particular to get to the next grade. But then you get to high school and now of all sudden you have to make a bunch of active choices. These active choices include applications and financial aid and forms and so on. Could we simplify that process? Could we actually create a pattern of defaults that allow people to seamlessly go to the next level and can we create processes for that? One of my colleagues, Philip Oreopoulos, has been working with a bunch of organizations to try and do just that. Can we prepopulate your application forms with information that we already have from your schooling, so that it’s not such a daunting thing for you to fill up all those forms.

Health is the other big one. I think there’s a lot of dollars being spent on curative medicine, but I think we need to [spend] a bit more attention on preventive medicine. What can we do to get people to go see their doctor once a year? What can we do to make sure they get their flu shots and wash their hands? These are all simple things that I think behavioral science can address really well.

Finally, a lot my work is in the global south, in poorer countries. I spend time in India and Thailand, and Malaysia. There, financial well-being is a big thing. We talk to people routinely who earn cash incomes, [have] unpredictable wages, don’t have access to banking, don’t have access to formal loans. I think behavioral science has a massive role to play there, in terms of making sure at the margins these people save enough that they don’t have to live their economic lives one day at a time.