I’ve been thinking about the future of behavioural science a lot lately. This is not very surprising, given that I am a behavioural scientist, and I continuously ask my interviewees, both those from the regular series and the Next Gen, what they think the future of behavioural science will be like.
In addition to all of that, this is also a question Sarah and I ask our guests on the podcast. There’s really no escaping it. The future of behavioural science is on my mind.
I’ll be very specific about what triggered this post. Sarah and I recorded an episode with Samuel Salzer, who will be our guest on the first episode of our second season. He’s great for opening the new season for us. We had a really great time talking to him. A topic that came up was the journey he, as many others before him, had into behavioural science. People like Sam looked at their peers who moved into more traditional careers, which had very clear pathways as to how to get where you wanted to go. Behavioural science didn’t have that. About a decade ago you’d struggle to find anything vaguely related to behavioural economics and behavioural science. Those jobs didn’t exist, or were well-hidden behind terms such as marketing, management or psychology. It was difficult to rock up to a research unit indicating you wanted to study “behaviour.” This is not something I imagine happening in the past. This is what my interviewees have told me.
But things have changed. There is now also a pathway for behavioural science.
I should know this pathway quite well. I did a liberal arts degree merging both economics and psychology, then moved onto an MSc in Behavioural and Economic Science. Now I’m in a PhD at a business school, which is renowned for having one of the best and largest behavioural science groups in Europe, if not the world. I’m an excellent example of a cookie cutter behavioural scientist.
And as always, when things seem to be going well, or there seems a clear progression, I get worried.
I don’t like clear pathways and cookie cutter output. I don’t like one size fits all approaches and I don’t like
Now it’s not that having a clear pathway and the option to choose behavioural science degrees makes life “more difficult”, it does quite the opposite. I think I’d have a very hard time figuring out my “career” ten years ago. I’m not snowflaking on the fact we now at least have the options and that the field is being taken seriously. I’m wondering what this change in approach and training is doing to the fundament of the field. Hear me out.
How did behavioural economics start? Well, technically it has existed for a very long time, as psychology and economics have always been very intertwined, until economics wanted to get all science-y, and cut out all the “soft stuff.” This happened around the 1920s and Pareto is a known proponent if not propellor of this. Then some psychologists rocked up and then some experimental economists rocked up and proved the very foundations of this new (neo-classical) economics to be wrong, or at least very doubtful. Cool cool.
The field started to attract more interested people. Economists found themselves talking to psychologists, neuroscientists, sociologists, statisticians, mathematicians etc. etc. And I’m not saying these were easy conversations. According to Colin Camerer, these were actually quite difficult conversations, as each field has its own language, and a lot of words get interpreted differently, or don’t exist at all. It’s like me trying to communicate in Russian. Not pretty.
But what it ended up doing was to throw together people with a wide variety of different backgrounds and different training. Backgrounds that sometimes clashed and had nothing in common. Training that seemed to move in opposite directions. But that was, with some give and take, a good complement for approaching similar problems from a completely different angle, and getting new results, needing different explanations.
It was not taking one training to be superior to the other, and applying critical thinking not only to the results, but also to the method, and the assumptions that method was grounded in.
Now we are seeing a second wave of this. We are trying to merge in data and computer science into behavioural science to better understand Machine Learning and AI. To build better models of cognition and align predictions with what we are seeing in the real world, as it’s happening.
But at the same time we are training behavioural scientists increasingly to know statistical analysis, coding and programming languages. They got interested into the field by pop. science books like those of Sunstein, Ariely and Gladwell. They need to learn expected utility, theory, prospect theory, and several other variations of these theories, the ever increasing number of biases, and frameworks such as EAST, COM-B, MINDSPACE and ABC.
It’s a core part of academia, or even life, to celebrate the successful theories and frameworks and discard all the others. But also discarding the evidence against those “core” theories. Prospect theory has had a lot of failed replications, and let’s not forget the initial sample size of that study (it was small, it wouldn’t get published today). Loss aversion, one of the core aspects of prospect and a key bias/heuristic doesn’t continuously replicate either. A lot of “must have” biases do not replicate.
So where is that discussion? Where is the module discussing why things didn’t replicate? How to better approach certain phenomena? Where are the courses on teaching methodologies that aren’t necessarily associated with how we do behavioural science now? Our field became great on diversity, so why are we so keen to produce cookie cutter behavioural scientists?
I’m not saying there’s no individual differences between countries or even universities. I don’t think you’ll get the same training between LSE, UCL, Warwick and Nottingham, to be quite frank. Especially given that some of these programs are Behavioural Economics programs, and not Behavioural Science programs (there’s a difference!). But there are still things we should make sure are in place. The number one being knowing where our field came from, what made it great, and as a result of that continuing to stimulate critical thinking. Just because we think one theory or method is great now, doesn’t mean it’ll be great forever. And when teaching things such as prospect, we also need to discuss the ample evidence against it, and the other theories and explanations that have been proposed.
I love my field, I really do. But I don’t think progress is going to come from us all having had the same (similar) training. Accepting the biases, theories and frameworks as they are, and seeing them as “the fundaments of behavioural science.” Because if we look closer to the evidence as is, they might not be so fundamental. They might not be supported by evidence at all. I know how the confirmation bias works, but come on. We made that mistake once before. Let’s not make it again.
