Toward Better Decision Making: Q&A with Marc Laitin, Managing Director of the Venture Studio, ideas42
At the Inaugural Market Shaping Conference, participants highlighted innovative and successful applications of large-scale data to improve public policy. The conference was supported by Schmidt Futures and took place on September 21 and 22, 2018 at Stanford University’s Koret-Taube Conference Center.
When it launched, ideas42 was the first behavioral design lab in the world. It applies the insights of behavioral science to some of the world’s toughest social problems, such as increasing savings rates in the Philippines, preventing students from dropping out of college in California, educating teenagers about the risk of HIV in South Africa, or conserving water in Costa Rica. Marc Laitin, managing director of the Venture Studio at ideas42 and the former vice president of product management at Opower, discussed how ideas42 uses behavioral science for social good at the Market Shaping Conference.
What does ideas42 do?
Every day, billions of people make trillions of decisions. These decisions affect the health, lifestyle, and businesses of the people who make them. We are barely conscious of many of them — for example, whether to turn off the lights when leaving the house. Most of the decisions we make are both good for us and good for society. However, some are bad.
These bad choices cost people, businesses, and the governments they interact with hundreds of billions of dollars every year. Ideas42 is a nonprofit that tries to help people make fewer of these costly decisions.
How does ideas42 identify areas to work on?
Ideas42 designs and implements systems that use behavioral science for social good.
To do this, we focus on situations where there is a gap between what a person would choose to do and what that person actually does. This generally happens because there is some impediment that gets in the way of a good decision, such as a lack of attention or our tendency to procrastinate. Many people may not even be aware that they aren’t behaving consistently with their goals — for example, they might want to save energy and reduce their utility bills, but they don’t realize that they frequently leave the lights on or the water running.
For us to get involved, the behavioral gap must be both individually and socially costly. If it is, we look for a way to intervene to help people make a better decision, using a data-powered “nudge” that can change the behavior that created the gap. We are particularly interested in solutions that can be implemented at scale.
Would you say these are market failures?
Yes, I would say that we are fixing market failures. Society could be achieving an outcome that everyone would prefer, but before we get involved there is no structure in place to get us there.
How do you shape markets to overcome these failures?
We have an initiative called ideas42 Labs. It looks to start independent, for-profit businesses that help fix these failures. We need insight into how to close the behavioral gap, and we need a source of data on how it is working. Without this data, we cannot use our software, and this limits our ability to scale the intervention.
There must also be what we call an “enterprise beneficiary” of the behavioral change, which owns the data that will drive the behavior change and has a relevant relationship with the individual. Someone has to pay for the intervention — but no one is going to pay for it unless they will benefit from it, too. Any enterprise beneficiary that we identify would be better off if individuals successfully change their behavior, and so it has a financial incentive to fund our intervention.
Have you seen all this work in the real world?
The best example of a company that has done this is Opower, and I was lucky enough to work there for seven years. We gathered data, which meant we could show consumers their energy consumption. We compared their energy consumption to that of similar homes and also of similar energy-efficient homes. Just seeing this information successfully motivated residential energy consumers to use less energy. This saved customers $360 million in 2016 alone, but the reduction in energy use of 3.6 TWh also benefited the energy companies that we worked with, because they were better able to meet state-level energy efficiency standards.
Where do you see the most promising opportunities today?
There are many other market opportunities like this right now! There will be even more in the future if markets are regulated to support the enterprise beneficiaries. Our first portfolio company is called EveryDay Labs. This reduces the 300 million unexcused K-12 absences every year by motivating parents and guardians to make sure their kids go to school. The service is clearly valuable to students and parents, but also to school districts that face costs of $15 billion from their students not showing up to class. Using traditional methods, it costs $500 to eliminate one absence.
In 21 school districts, EveryDay Labs now delivers absence reports to parents and guardians at critical moments throughout the year that help them keep track of a child’s absences and motivate them to reduce future absences. The program reduces chronic absenteeism by 10 to 15 percent and, thanks to state incentives for reducing absenteeism, returns many districts $5 to $7 for each dollar they spend.
How important is it to have a clear enterprise beneficiary?
One of our projects in development tries to reduce recidivism among released prisoners. We know that half of the people who leave prison today will be back there in less than three years. Often, this is because former prisoners do not have the basic support needed to function in society. We want to provide them with a smartphone with apps. The apps help them to make and keep scheduled appointments with their parole officers, for example. If it works, this will reduce their chances of going back to prison, and so the intervention helps those individuals and also reduces the high cost to society of incarcerating prisoners.
But identifying the enterprise beneficiary with the incentive to pay for this program is complicated. The total cost to cities and states is $37 billion a year, so the potential payback is large, but many separate organizations share the cost, and so the potential benefit. We have a challenge to figure out in order to make this opportunity work.
And what if the individual is not aware of the personal cost of his or her behavior?
We are currently thinking about helping individuals with student loans manage this debt better, because 40 percent of us default at least once on our student debt. Federal student loans have safeguards built in to help debtors who cannot pay, but the paperwork is complex and tedious. We want to provide an intervention that makes it easier to enroll in repayment plans and help reduce this problem.
Defaulting students are a high cost for colleges, and for the Department of Education — it faces $800 million in annual collection costs. So, in this case, we can identify enterprise beneficiaries.
But the debtors themselves do not face immediate negative consequences when they default. It is hard to design an effective intervention that will change their behavior, even though the eventual impact of defaults, which hurt a student’s long-term financial health and credit record, is large.
How do you ensure that an intervention is good for both the individual and the enterprise beneficiary?
When we use behavioral science to design our interventions, we always test them in small pilots to observe the effects. We examine the outcomes for the individuals and, if those are positive, we pursue a scale-up, at which point we approach the enterprise beneficiary. This means we already have a strong business case for the intervention. We try to ensure that there is continual assessment and tracking of the intervention’s outcomes, and we only look at opportunities for which we can access the data to do this.
How can researchers help?
Our interventions have always been founded on rigorous behavioral research. Ideas42 was created at Harvard in 2008 by a team of top academics. We will remain true to our roots and the scientific approach: Our work is also informed by an affiliate network of more than 85 academic affiliates and advisors, including three Nobel Prize winners.
But there is always room to improve a behavioral intervention, and we constantly work with academic researchers to come up with better designs. Researchers can also help us to analyze the large amount of data we collect through our apps.
Learn more about the Golub Capital Social Impact Lab at Stanford Graduate School of Business.
Follow us @GSBsiLab.
Learn more about Marc Laitin.
With writing help from Erin Fahrenkopf and Tim Phillips