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 Umar Taj.
Umar is an Associate Professor of Behavioural Science. He teaches executive education at London School of Economics, University of Oxford and Warwick Business School. He is also the Co-founder/Director of Behavioural Decision Making at CogCo, an organisation helping institutions grow using behavioural science, design and data science. He previously founded the ongoing ‘Nudgeathon’ programme at Warwick Business School, which helps organisations to rapidly learn and apply concepts from the behavioural sciences to addressing their own goals.
Who or what got you into behavioural science?
After my first year of undergrad, I went to a summer school which had a behavioural science element to it. That's when I got first introduced to some of the concepts, which was more around behavioural insights that relate to decision making. I fell in love. I really enjoyed it. And from there I decided that I want to do something around it. By then I still had to finish my course at Warwick (MORSE – Mathematics, Operational Research, Statistics, Economics) and I enjoyed the ORSE bit, but not the Mathematics. I liked the quantitative aspect, loved the approach towards experimentation. So I decided to do a Master’s in Decision Science – this no longer exists, unfortunately, but it was amazing. And then after that I decided to do a PhD in Behavioural Science, focussing on applications of behavioural science.
What is the accomplishment you are proudest of as a behavioural scientist? And what do you still want to achieve?
At the moment, what I'm proudest of is work that I did in Pakistan during the pandemic. We were able to work with the telecom authority to be able to put an audio message in place of the waiting tone (the phone ringing) when you’re just waiting for someone to pick up. We tested a variety of messages, promoting vaccination and mask wearing and we reached 113 million people doing this. This is to my knowledge the biggest behavioural intervention run during the pandemic. If not ever. I am very proud of that.
And of course I am proud of Nudgeathon too, I know you’re hinting at that (directed at Merle), but I want it to be bigger still. I want it to be a core part of universities across the world. I’d want there to be a global Nudgeathon. But that will require more time and resource investment. So I’d still want to achieve that.
In a similar vein, because it’s all about impact, I have a personal goal to positively impact at least 1 billion people through my work. I’m hoping to reach that number, cumulatively. So with having reached 113 million people in Pakistan I’m about 11% through to my goal. My projects now are closer to affecting hundreds of thousands, but still, we’ll get there!
If you weren’t a behavioural scientist, what would you be doing?
At one point in time, I wanted to become an architect. If I hadn’t enjoyed the quantitative approach so much I think I might have gotten bored and I might have thought about doing a degree in architecture.
How do you apply behavioural science in your personal life?
So my background is in Decision Science (Strategic Decision Making), rather than behavioural science, which is different. So I use strategic decision making in my personal life. And this can take many forms, but one I apply a lot is making sure I don’t consider binary options. I don’t approach options from an either/or standpoint. Either I do this, or I don’t. According to research it’s better to have at least three options. In your case (directed to Merle) you had the option to remain in the UK, move to Australia or move somewhere else. It wasn’t Australia or bust right? So it’s things like this. Also important, we tend to frame decisions in terms of options, but it’s better to frame decisions in terms of objectives. Before deciding on the option set, figure out what objectives you’d like to achieve, and base your decision on that. My third and last point is that often we think we’re making decisions, but we aren’t. I see this a lot with the executive ‘students’ I teach. They tell me that they have a big decision to make, such as: ‘what should be the strategy for the next five years?’ That’s not a decision. A decision is only a decision when you know what you're choosing from. It'll only be a decision once you've put down options, otherwise you are at a pre-decision stage.
With all your experience, what skills would you say are needed to be a behavioural scientist? Are there any recommendations you would make?
The skills that you need to develop are all non-behavioural science skills. You need to improve on how you manage projects, on how you manage stakeholders and on negotiation. I'm not talking about negotiating deals, but when you're applying behavioural science into practice, you are negotiating because more often than not, you are not the one who will actually implement it and you need to get and keep all your stakeholders on board – that’s negotiating and relationship maintenance at its finest. I think these are the skills that people underestimate the most. You don’t necessarily need to have a behavioural science degree to come up with good behavioural science based interventions. Sometimes, you can get inspiration from reading behavioural science books. You can even get them from ChatGPT now (not that I recommend that!). What you do need to be a good behavioural scientist is the skills to make sure these interventions actually get implemented. We miss this point quite often.
What advice would you give to young behavioural scientists or those looking to progress into the field?
Much in line with my answer to the skills-questions, I’d recommend for people to make behavioural science happen. If you want to pursue a career in behavioural science, always find ways to put behavioural science into practice while you are studying. Don't wait for your degree to finish. While studying, you need to find someone to collaborate with, who is running experiments and maybe you can shadow them, learn from them, see how they make interventions happen. I myself have always just done experiments for fun. I remember that once I just tested the status quo bias for the light being on, or off, in a toilet. I’d just sit outside the toilet and seeing how likely people were to switch the light off when leaving the toilet, having entered it with the light having been on. And vice versa. I just have a curious and experimental mind. But this also teaches you about the nuances of running experiments. You need to get your hands dirty in terms of experimentation. Just do it. Do I think doing a masters and a PhD is beneficial? It can be. If you’re also using these degrees to, in addition to upping your theoretical knowledge, increase your network and empirical research experience. Especially a PhD will help a lot with the latter.
How do you think behavioural economics will develop (in the next 10 years)?
So most of the work that we see right now is one nudge for all. that's actually really boring, and very basic. We still see that same text message to everybody. I want us to move towards smart nudging. We see this method in some domains already, for instance, credit risk scoring. Where we have these predictive models where you learn from a smaller initial sample what might be the risk factors of someone defaulting. And then for anyone new who comes in, you run them through the model you built of the initial sample. You could use exactly the same thing in figuring out what Nudge works for whom, and you could be smart about it for the next set of people.
Obviously this does require a sample to be big enough for you to be able to do this. But I still think that this is manageable and yet not enough work is happening in us being smart and more targeted in developing interventions.
What are the biggest challenges for behavioural science??
I think a massive challenge for behavioural science will be the fact that it has been oversold and cannot live up to what some people promised it could do. Most behavioural scientists are very humble, they are very clear about the limitations about their own results and the context those results were obtained in – behavioural science is no silver bullet.
And I think the rapid popularisation of the field has contributed to this false belief that behavioural science can fix the world. But then we get papers showing relatively modest impact (for example, the paper from DellaVigna and Linos showing that across 100+ of nudge unit RCTs they get a relative increase (of desired behaviours) of less than 10%. And we as behavioural scientists know that those are the effects we are working with, but I feel that people outside of the field don’t have those expectations. We’ll need to do some serious expectation management. I think this is already starting to happen now but still a challenge.
What are your biggest frustrations with behavioural science, as it currently stands?
I feel that behavioural scientists aren't really tackling wicked problems. I do a fair bit of work on preventing violent extremism. I hardly come across behavioural scientists who are doing this. There's not many of them. There are these wicked challenges where I think a lot of behavioural scientists are still not ready to get their hands dirty. I think this is a consequence of most behavioural scientists being exposed to a very similar kind of environment (from and based in WEIRD countries) and that's also the environment they're working in. But if you just think globally, there’s a lot bigger problems than tax compliance and pension savings. Don’t get me wrong, I think these are important as well but I'm a bit disappointed that the concentration of behavioural science work is so concentrated on those problems. I'm disappointed that I'm not seeing more breadth in the domains that behavioural scientists are working on.
To solve for this, I think we need to encourage young behavioural science students who come from diverse backgrounds, and when I say diverse backgrounds, I mean from different regions in the world, and I think this problem will slowly start to solve itself. They’ll learn behavioural science wherever they can, and take that information back to a context where it matters. At least I hope so.
Which other behavioural scientists would you love to read an interview by?
I have always enjoyed my interactions with Daniel Read, but you have already interviewed him. I think you’ll really enjoy interviewing Barbara Fasolo, who used to be faculty in the Decision Science Master’s at LSE as she is also a strategic decision scientist!
Thank you so much for taking the time to answer my questions Umar!
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!
