Interview with Elisabeth Costa
- Merle van den Akker
- 7 hours ago
- 8 min read
Behavioural Science is a rapidly expanding field and everyday new research is being developed in academia, tested and implemented by practitioners in financial organisations, development agencies, government ‘nudge’ units and more. This interview is part of a series interviewing prominent people in the field. And in today's interview the answers are provided by Elisabeth Costa.
Lis is the Chief of Innovation and Partnerships at the Behavioural Insights Team, leading global partnerships and the development of new service offerings and emerging areas of expertise, with extensive experience across the UK, Australia, and the United States and particular expertise at the intersection of behavioural science and AI, market regulation, and economic policy. Previously Managing Director of BIT UK, she led flagship programmes applying behavioural science to economic policy and oversaw dozens of projects and trials, co-authoring influential work on consumer and financial decision-making, online safety, labour markets, productivity, and sustainability, while regularly contributing to public policy debate. She is a Senior Visiting Fellow at the London School of Economics, a member of Ofcom’s Online Information Advisory Committee, and previously held senior roles at the Australian Treasury and completed postgraduate study at Harvard University.
Who or what got you into behavioural science?
I had a relatively indirect pathway. I trained as a lawyer and studied law and commerce, focusing on economics. I was set to become a corporate lawyer and enjoyed working at one of the big firms. But I took a graduate role at the Australian Treasury, just as the Rudd-Gillard government came into power. It was an exciting time for public policy. Over nearly a decade, I worked on tax, climate change, clean energy financing, and emissions trading. After those policies failed in the Senate, I started questioning where I could make the most impact. That led me to work at the UN Environment Programme and later to do a Master’s at Harvard Law. It was there I took Cass Sunstein’s class—he had just returned from the White House. That class connected regulatory policy, environmental policy, and behavioural science for me. In 2015, I joined the UK Behavioural Insights Team to lead energy and sustainability. I took a leap of faith, moved back from Australia to London, and built out work on financial decision-making, gender, consumer regulation. BIT was small then—25 people. Now it’s over 200 in 11 global offices. It’s been a fantastic journey, and I’m so grateful for that class.
What is the accomplishment you are proudest of as a behavioural scientist? And what would you still like to achieve?
I'm extremely proud of how behavioural science has grown globally. When I joined BIT, the field was smaller—now it’s vibrant and global. Personally, I’m proud of our work integrating behavioural insights into regulatory systems. Back in 2015, few regulators had in-house behavioural teams. The UK’s Financial Conduct Authority was leading, but it was rare. In 2016, I co-authored a paper proposing how behavioural science could be part of the regulatory toolkit. That sparked early experiments with the UK’s energy and telecom regulators. Today, many regulators worldwide have strong behavioural science teams. Seeing that shift feels like a real accomplishment. Looking ahead, my top wish is to run a project to reduce online abuse of women in politics and journalism. I think behavioural science can offer meaningful interventions here. There were promising early trials at Twitter, using friction to prompt users before posting abusive content—about a third reworded or deleted. That small moment of reflection has massive potential in a dynamic system. Scaling such interventions could shift norms and reduce harm in online spaces. So, if anyone’s keen to collaborate—get in touch. That’s the project I’d most love to make happen.
What are the greatest challenges being faced by behavioural science, right now?
A major challenge is the superficial application of behavioural science. Its popularity is both a strength and a risk. There are now lots of books and headlines that have popularised the field, which is wonderful—but it sometimes leads to simplistic use. You see it in policymaking: plug in a social norm here, a friction there. But that’s not what good behavioural science is. It’s complex, rigorous, and context-specific. It’s our responsibility as practitioners to champion that depth, push for higher standards, and call out superficiality when we see it. It becomes essential that the field develops a shared language about what good behavioural science looks like—and why rigour matters. This isn’t just about pride—it’s about doing work that really helps people. Another challenge is the framing debate—boosts vs nudges, iframe vs s-frame. While I enjoy a healthy debate (the lawyer in me relishes it), I find these definitional arguments reductive. We operate in complex systems. Sometimes an individual focus is right, sometimes a systems approach is better. We need both. Instead of arguing which is better, let’s talk about when and how to use them effectively. Finally, we need to think beyond single interventions. Sustaining behaviour change requires broader strategies—building people’s skills, changing institutions, reshaping systems. Behavioural science is evolving toward that, and we must keep pushing forward. We’re not just fixing one-off problems—we’re helping people lead better lives.
How do you think behavioural science will develop (in the next 10 years)?
I think behavioural science will evolve in exciting ways. One area we’re exploring deeply at BIT is the intersection with AI. AI has the potential to transform how we do behavioural research—making literature reviews faster, enabling qualitative research at scale, and creating more dynamic, personalised interventions. Instead of static nudges, we can now offer interactive, real-time experiences. Early evaluations show this can be significantly more effective. But it’s not just about using AI to enhance what we do. There’s also a new frontier: understanding how AI systems affect human behaviour. For example, if someone asks a chatbot, “Who should I vote for?”, what is the responsible response? Do we offer impartial information? A list of candidates? Help sort their values? Each option has behavioural implications. These are deep, unresolved questions that demand a behavioural science lens. We’re also seeing challenges with the adoption of AI in areas like mental health—AI therapy bots that meet clinical standards are still met with hesitancy by doctors. That, too, is a behavioural challenge. Beyond AI, I’m excited about scaling participatory methods like deliberative democracy. We recently partnered with Meta to run large-scale online citizens’ assemblies across countries and languages. It retained the benefits of in-person deliberation—better ideas, deeper understanding, trust—but at unprecedented scale. I see more blending of traditional and digital approaches in the future. So yes, AI is part of the story—but so is human connection, participation, and system-level thinking. We shouldn’t see these as opposites. The best behavioural science of the next decade will integrate tools—digital and analogue—to understand and improve decisions at every level of society.
With all your experience, what skills would you say are needed to be a behavioural scientist? Are there any recommendations you would make?
To succeed as a behavioural scientist, you need strong methodological skills and curiosity about human behaviour. A solid academic grounding—whether in psychology, behavioural economics, or another relevant field—is important, but equally essential are practical skills. Working in behavioural science is not just about understanding theory. It’s about applying it—often in messy, real-world contexts. So you need to be adaptable, collaborative, and resilient. You also need project management skills, communication, and the ability to influence stakeholders. I always say, there’s no substitute for doing. Jump into an organisation where you can work hands-on—doing qual research, running field trials, solving actual problems. From my own experience in government, I developed something I call “principled opportunism.” That means being persistent and creative about spotting when the stars align to push a good idea forward. You need to understand policy windows, stakeholder motivations, and when to act. It’s not enough to have great evidence—you need to know how to get traction. Finally, behavioural science often happens in teams. So collaboration and humility are crucial. And don't underestimate the importance of tenacity—especially when experiments don’t go as planned, or data collection is delayed. Fieldwork is unpredictable. The best behavioural scientists are those who stay calm and focused when plans go sideways. They’re problem solvers.
What advice would you give to young behavioural scientists or those looking to progress into the field?
Start with a strong foundation in behavioural theory and methods. There are now excellent academic programmes that didn’t exist when I was starting out. Take advantage of them—but remember that real learning comes from application. Get into an organisation where you can work on real-world problems. Understand not just the behaviour you’re trying to change, but the system around it. That depth of understanding is what separates surface-level nudges from real impact.
Learn to navigate complexity and ambiguity. Behavioural science is rarely about silver bullets—it’s about digging in and iterating.
And then, equally important, learn the soft skills. Project management, stakeholder engagement, influence, communication—these are what make projects happen. You can have the best hypothesis in the world, but without buy-in and good delivery, it won’t matter.
Also, be resilient. Field experiments fail. Timelines slip. You’ll need to problem-solve and keep going. That persistence is critical.
And finally—be part of the community. Behavioural science is global, growing, and collaborative. Attend events, read widely, and share your work. Ask for feedback. Learn from others. This is a field where generosity and curiosity go a long way. Your path doesn’t have to be linear—mine certainly wasn’t—but stay curious, stay rigorous, and stay human.
What is your biggest frustration with the field as it stands?
Honestly, it’s the definitional debates. The constant back-and-forth about whether something is a nudge or a boost, or whether we’re using the i-frame or s-frame. I find these kinds of debates really reductive. They pull us into binaries that don’t reflect how behavioural science actually works in practice. We work in complex systems. Sometimes the individual is the right level of focus; sometimes it’s the system. And often, it’s both. So I’d much rather see us asking: When is each approach appropriate? How do we combine them thoughtfully? What does good integration look like? Rather than wasting energy arguing for one over the other, let’s build bridges. The more productive conversation is about how we can use the full behavioural toolkit in smart, context-specific ways. That’s where the real impact lies.
If not a behavioural scientist, what else would you be?
If not a behavioural scientist, I would’ve likely stayed in government. I’ve always been driven by social impact and evidence-based policy. Early in my career, I worked on economic policy during a turbulent time in Australian politics. I imagine I would’ve continued on that path—focusing on climate, tax, or clean energy. That said, taking Cass Sunstein’s class at Harvard was a real sliding doors moment. It connected all my interests—law, policy, and behavioural science. Without that course, and encouragement from friends at BIT, I might have never made the leap to behavioural science. I'm grateful I did.
How do you apply behavioural science to your own life, if at all?
Yes, I do! Writing our book (Elisabeth is co-authoring a book with David Halpern) I have been working on a chapter about attention—how it’s captured and how to reclaim it. Our phones are enormous attention sinks, so I’ve been experimenting. I tried One Sec, which creates a short delay before you can open apps like Instagram, and a physical device called Brick to increase friction. These tools apply behavioural principles to help regain control. I’ve also been rebuilding habit loops: when I notice I’m about to scroll mindlessly, I get up and do something else. It’s all about creating tiny moments of pause. As with anyone, it's a work in progress. But even as a behavioural scientist, I’m still surprised at how powerful these small nudges can be—when applied with intention.
Who else would you recommend me to interview?
Two people immediately come to mind. First, Iyad Rahwan—he’s doing fascinating work on dishonesty and AI, and also explores the intersection of behavioural science and visual art. I think you’d find his perspectives really fresh. Second, Rasha Attar, who leads the behavioural science group at the Office for Development Affairs in Abu Dhabi. She’s worked on synthetic participants in experiments and is deeply thoughtful about applying behavioural science in the Middle East. Both are not only smart but generous with their ideas. I think your conversations with either would be incredibly insightful and engaging.
Thank you so much for taking the time to answer my questions Lis!
As I said before, this interview is part of a larger series which can also be found here on the blog. Make sure you don't miss any of those, nor any of the upcoming interviews!
Keep your eye on Money on the Mind!
