5by5: Public Policy and Behavioral Insights with Shrupti Shah, Managing Principal at GovLab
“I fear that this widely publicized abuse of the two sciences (data and behavioral) will set back our efforts to harness their potential in areas where they could genuinely do good. Having said that, I’m grateful that it is shining a light on the ethics of predictive analytics and behavioral design so we can come together as a community and try and develop some shared ethical principles and parameters to guide our work.”
Question 1 Behavioral insights has made headway informing public policy in certain parts of the world. What can be learned from these applications for organizations that are still in skeptical or in consideration mode? (or how can you influence organizations to accept different ways of problem-solving?)
Always starting on the intended public policy outcome is key to being open to different ways of problem-solving. If we are focused on resources or particular inputs or outputs (e.g. how much we spend on health or how many doctors/ hospitals or heart operations that buys), it forces us into a world where there is only one way to solve an issue – usually the traditional way.
If instead, we were to start with the intended outcome (such as reducing mortality through heart disease), we can then take a look at the many factors (diet, exercise; genetics; social determinants of health etc.) that influence our intended outcome to inform the selection of a more effective portfolio of interventions. In short, we need holistic solutions for holistic problems.
Question 2 What are the ethical considerations that go into your planning process? How do you apply a human-centered approach to organizations that haven’t always been traditionally been transparent or planned policy keeping user needs in mind?
I think it’s about two things. Firstly, always keeping the public service ethos that underpins what we do as policy leaders front a center. By that I mean, remembering why we chose public service – to support people and help and empower them in living safer, happier, healthier lives and not to control or restrict them. Secondly, by persuading through results. When done right, applying a human-centered approach to policy and programs can have extraordinary results in terms of improving outcomes. Having said that, we need to be honest about where it is likely to make a difference and where it is less likely to be a game changer. A first attempt that is successful will lay the foundation for wider and more sophisticated applications but the first failure will put them off and feed the cynics.
Question 3 Would you be willing to share instances of failure? What lessons could designers/policy makers learn from applications that didn’t work?
It’s not so much of a failure but a cautionary tale. In combination, data analytics and behavioral insights can really help with some of our most challenging public policy issues in fields as diverse as education and health to crime and criminal justice. But that same combination can also be used to feed our darkest paranoias and prejudices for far less noble ends. We know that Cambridge Analytica (and possibly others) used social media data and provocative messages to divide and/or influence people in democratic processes around the world. I fear that this widely publicized abuse of the two sciences (data and behavioral) will set back our efforts to harness their potential in areas where they could genuinely do good. Having said that, I’m grateful that it is shining a light on the ethics of predictive analytics and behavioral design so we can come together as a community and try and develop some shared ethical principles and parameters to guide our work.
Question 4 If behavioral insights became part of every policy makers’ toolkit what are some best practices that could be leveraged?
Three steps. First, design specific changes to policies and programs based on a combination of human-centered design and lessons from behavioral science. Second, test each separately in the field to understand what works and if possible, why and with whom. Third, share what you find (even/especially the failures) with other public policy leaders so we can scale what works and continue to learn and improve.
Question 5 What are some untapped areas/sectors were behavioral insights can be used to help shape policy?
- Community Cohesion – trying to forge new collective identities that transcend race, culture, religion or class
- Addressing poverty of aspiration and deliberate or unconscious bias – to address inequality in the educational and workplace arenas
- Diplomacy – Improving perceptions of your country abroad and engaging in multilateral arenas.
Interested in this topic? Register now to be part of a larger community at the Design Intersections conference in Chicago May 24-25, 2018.
Interview by Twisha Shah-Brandenburg and Thomas Brandenburg, a collaboration with 5by5.blog
Our charge is – we develop 5 questions and invite 5 diverse voices to answer them. Through this activity we hope to collect the different perspectives to shed light on important questions that need to be raised as design continues to evolve and gain influence. We hope that this blog is place that encourages a communal dialogue and services as a place for inspiration and understanding.