How ‘behavioural insights’ are changing public policy

Sasha Tregebov is the director of The Behavioural Insights Team, Canada. The organization originally formed within the U.K. government to apply behavioural insights to policy challenges. In effect, making better policy through science. Tregebov was a recent speaker at PPF’s Policy Leadership Program, offered to public servants who are directors or aspiring directors, in collaboration with the Telfer School of Management. He talked to PPF about the rise of behavioural insights, where it fits in the policy process and why it comes with some important ethical considerations.

This interview is edited for length and clarity.

What is “behavioural insights” and how did it become something tied to public policy?

Sasha Tregebov: Academics over the last 30 to 40 years in a bunch of different fields — things like social psychology, behavioural economics, which are collectively referred to as behavioural science — have been doing all this really fascinating research on all of the hidden factors that influence how we process information, how we make decisions. There was this idea that started to gain momentum in the early 2000s that we should bring that understanding out of academia and into the work of government. There was this idea that if we could have a more nuanced and realistic model of how people behave, how they make decisions, we could transform public policy and service delivery, make it more intuitive and easier for people to do.

So in 2010, the Behavioral Insights Team was set up within the heart of the British Government in its Cabinet Office and they coined the phrase “behavioural insights” as a way to differentiate the applied work that they were doing from the academic work that they were drawing from and being inspired by.

What do people in a behavioural insights team do?

Sasha Tregebov: We’ll look at either a policy issue or an operational implementation issue and we’ll just bring a new lens to that problem. One of the things that got us really famous in 2010-2011, and that indirectly led to me getting the job that I’m in now, is some work on getting people to pay their taxes on time that were late paying. We added one sentence to the beginning of letters that were sent to taxpayers and we just said: ‘Hey, 9 out of 10 people with a tax debt like yours pay their taxes on time,’ which was completely true. Simply referencing what is normal behaviour brought forward hundreds of millions of pounds in revenue at basically no cost.

Today, we’re working on issues that are far more complex with interventions or new approaches that are much more sophisticated than adding a sentence to a letter. Like: How do you equip community organizations in Canada to help build vaccine confidence within historically marginalized communities? How do you help people receiving social assistance navigate transitions towards employment? How do you share labour market information in a way that can help newcomers or the recently unemployed or students chart their career and educational pathways? There are many applications of this approach.

What is it about a single sentence that makes people adhere to the norm? What is it cognitively that you’re tapping into?

Sasha Tregebov: There’s this idea that at the very heart of behavioural science, which is that when people make a decision, they don’t always – and in fact they don’t usually – weigh up all available information, assess the long-term pros and cons and take the choice that creates the most utility in the long run. That is the classic rational economic approach to understanding decision making.

Our brains are constantly taking shortcuts – and thank goodness for that. One of the shortcuts that we take that’s, generally speaking, very helpful for us, is to look around at people who are like us or in a similar situation to us and see what they do and use that as a cue for what we should do. So we were tapping into the fact that people like to get cues from their neighbours, their peers, etc., on how to behave.

In the case of tax filing, that information was completely absent, because it’s not what you talk about when you’re hanging out at the park with your kids and your friends. You’re not like, “Hey, have you paid your taxes?” So, that sort of social reference was really missing.

Is that what you would call a nudge?

Sasha Tregebov: Yeah, a nudge is basically a change to the environment in which people are making choices that does not actually constrain their choices or create economic incentives around those choices. That letter fits in perfectly to the classic definition of a nudge. We weren’t increasing fines or creating an incentive for people to pay their taxes. We were just shifting the mental – or in this case, social – environment in which people were making that decision to pay or not.

Where do you fit into a policy process? Say a government comes in and says we’ve got this mandate, are you there day one and saying: this is how to implement it? Or do you come in when policy has already been made to suggest how to action it?

Sasha Tregebov: You can absolutely bring some behavioural scientists in early. Let’s say the government is looking to increase the number of girls interested in STEM. We could say: from a theoretical perspective, from an evidence perspective, here are a bunch of approaches informed by behavioural science that are effective and that you should consider.

What’s probably more frequent is that we’re brought in a bit later in the process. For example, the government may have identified a piece of legislation or regulation or programming that they’ve developed at a high level that they think is going to achieve a policy goal. They might ask a behavioural scientist to come in and audit it and say: Okay, what are the assumptions that are implicit in this approach about how people are going to behave? Or, ‘hey look, you’re assuming here that people are going to want to sign up to this this program because it creates a lot of long term value, but actually we know from a psychological perspective that the way they have to sign up for this program – they have to maybe go to their bank and open up this new type of savings account or something like that – is really something people try to avoid at all costs, so you should rethink your fundamental policy mechanism.’

Unfortunately, the most frequent case when we get called in is actually post-implementation, where the whole thing has been implemented and now a piece of the puzzle isn’t working right. You had such a great program, but people keep dropping out and you don’t actually get the full benefit of the program, unless you’re enrolled for six months. We would do some research, talk to people, look at the data, look at the evidence base and say: we think there are some behavioural factors here. Or we might come in and say, you know what, this actually isn’t a behaviour change problem.

What about the ethics of this approach to public policy. Are we manipulating people in a certain way?

Sasha Tregebov: There are important ethical parameters and considerations at the core of our work. But the first thing I’ll say is that there is no neutral environment that we can create for people to make choices in. When we create a new policy, when we shape a new environment in which people are making decisions, even if we’re not conscious of the behavioural science, we are making choices related to the behavioural science. Maybe we’re making it hard for people to sign up. Maybe we’re making it easy for people to sign up. Or we’re creating a default option where people just go with the flow. Are you registered in your workplace pension plan, if you’re lucky enough to have one by default, or do you need to sign a form? Those two options are going to create really significant differences in outcome – opt in or opt out – and you have to pick one. There is no neutral model that is neither opt in nor opt out, right?

Government is in the business of behaviour change – and that can either be in an evidence-informed way or a less evidence-informed way. They can’t actually get out of the game and just let people decide for themselves. It’s impossible. My view is that at least knowing what you’re doing, it’s probably more ethical than kind of not knowing what you’re doing.

The second thing is I think your work should be guided by certain principles. So for example, our tagline that we were founded on and still used to this day is: Helping people make better choices for themselves. I think that’s really important that in the vast majority of cases, your interventions, the sort of approaches that you take, preserve autonomy and they boost the ability of people to make the decisions that they would make if they did have the time and bandwidth and knowledge to really think through all the pros and cons. In my mind, a good behavioural scientist is just making it easier, making it more likely for people to make that choice.

Learn more about the Policy Leadership Program