Solutions Cast is a podcast series hosted by Innovation-AccLab Pakistan. In this series, we explore conversations on the disruptive impact of COVID-19 and also highlight local innovation and solutions that are accelerated by the crisis.

In our second episode, we talk about how Behaviourial Insights, especially while communicating in a crisis can bring a shift in decision making. How can we use perception polls to gain deeper insights into people’s behaviours? What are Behaviourial Insights? And how can we use this approach to improve lives? These are some of the questions we explore with our guests — Dr. Umer Taj, who is a Behaviourial Scientists and Bilal Gilani, who is the Executive Director at Gallup Pakistan.

This episode is hosted by Javeria Masood, Head of Solutions Mapping at Innovation-AccLab Pakistan; and Tabindah Anwar, Communication and Knowledge Management Consultant, at Innovation-AccLab Pakistan.

To listen to our first episode on trends and disruptions that COVID-19 brought in Pakistan, click here



[00:00] Dr. Umar Taj: Within behavioural science, the research has shown again and again that people often make decisions intuitively without a lot of effort or without a lot of thinking, and quite often, with quite little conscious awareness.

[00:26] Tabindah Anwar: Hi, I am Tabindah Anwar.

[00:28] Javeria Masood: And, Javeria Masood.

[00:30] Tabindah Anwar: Welcome to Solutions Cast, Podcast Series hosted by Innovation Acc-Lab, Pakistan. In this series, we want to understand the disruptive impact of COVID-19 and also talk about the local innovation and solutions that the crisis has accelerated. In this episode, we will talk about how Behavioral Insights, especially while communicating in a crisis, can bring a shift in decision making. How can we use perception polls to gain deeper insights into people’s behaviour? What are behavioral insights? And how can we use this approach to improve lives? These are some of the questions that we will be exploring with our guests. With me, I have Dr. Umar Taj who is a Behavioural scientist. Dr. Umar, if you could please introduce yourself briefly.

[01:16] Dr. Umar Taj: Thank you so much for having me here, Tabindah. I am Umar, I am a Behavioural Scientist and I work at the intersection of academia and practice. My interest is in applying behavioural insights into practice which means working with governments around the world or international development organisations, to try and bring social impact.

[01:43] Tabindah Anwar: Thank you. And we have Bilal Gilani, who is an Executive Director at Gallup Pakistan. Bilal, please introduce yourself briefly.

[01:52] Bilal Gilani: Thank you so much. I’ve worked at Gallup for about 12 years and I have background in Political Science and International Relations but my experience at Gallup has been running survey works for different organisations, including a lot of third-party evaluations, political polling, and generally social research.

[02:19] Tabindah Anwar: Okay, thank you. Welcome to the Podcast, Dr. Umar and Bilal.

[02:22] Javeria Masood: So maybe, you know, Bilal, we can start with probing a little into what BI means and how a lot of the times, its dependent on people’s perceptions and how a lot of the times, you know, behaviour, psychology, social psychology experiments are as taken something very minute and not really incorporated in taking decisions. So recently you guys at Gallup did an opinion poll which was around COVID-19. Can you tell us a little about why you did it and how you initiated the poll?

[02:55] Bilal Gilani: Yes so, we as a public opinion polling company really believe in the primacy of giving voice to the public. So, I mean, the public opinion part and the public is really crucial. For us, democracy cannot function till the voice of the people is heard. In the COVID times when everything becomes in a crisis and everything is held in a constant, the voice of the public is difficult to be heard. In this particular scenario, and in general scenarios as well, public opinion polling provides a tool for policy makers, for the government, for the people in general as well to hear what a collective voice of the society is on issues that matter. We started the project on perception tracking of the general Pakistani population related to COVID-19 as a project to give voice to the voiceless Pakistanis who were in the quarantine, who were spread across 200 million people and therefore, their voices were not being properly conveyed to where the power corridors were. This was the major rationale but the ancillary objective was to also provide decision makers with facts and figures, a tool for empirical decision making. And also to provide a continuous stream of data that can act as a monitoring and evaluation tool, a third party evaluation, voice for different initiatives in the economy, in society, in health, in COVID-19 itself through vaccination and other interventions that were happening. So these were the three key objectives: 1) giving voice to the people, 2) making decision making empirical, and 3) providing a continuous monitoring and evaluation framework for the government initiatives.

[05:26] Tabindah Anwar: Okay, thank you. I think our listeners would be interested to learn also about what are behavioural insights, or BI, for short and application of BI. Dr. Umar, my question to you is, can you tell us a little bit about that? And you have been working across the globe, also do share some examples of BI work that you’ve been engaged in our life before COVID.

[05:55] Dr. Umar Taj: Sure, sure. Most definitely. So behavioural science or behaviour insight is a term that’s becoming increasingly popular but its roots started in around 1970s and a lot of insights and methods, they are drawn from research that has already taken place in fields of psychology, economics and neuroscience. And fundamentally what behavioural science has done is it has challenged this view that people always behave in their rational self-interest. So within behavioural science, the research has shown again and again that people often make decisions intuitively without a lot of effort or without a lot of thinking, and quite often, with very little conscious awareness as well. So behavioural insights in short you can think about it as a body of knowledge that is informing what are the factors that influence behaviour change.

Just to give you an example of the kind of things that have been done and I will share a few. So let’s start with this one insight about, its related to salience. We are more attracted to whatever attracts our attention. And here’s one example where an organisation wanted to increase response rate to surveys, and that’s something which I think across all domains is something we would want. And the way they used the insight was that they just added a post-it note to the survey itself and just a handwritten message and that just in itself, this post-it note with handwritten message doubled the response rate. One other example that we can use on this is really interesting, so this is what behavioural insights offer which is a different lens. We all know that incentives generally do work. But here’s the interesting thing, there’s this behavioural insight which is called ‘loss aversion’. The insight is that we tend to value losses twice as much as the gains so imagine losing, you know, let’s say 100 rupees versus you’re getting 100 rupees, but the pain of losing 100 rupees is almost twice as much as the joy of gaining 100, so just using this phenomenon, there was this one study where they tried to incentivize teachers based on the performance of their students. And what they did was to one group they said ok, if the students do well, we are going to give you X amount of money and to the other group of teachers, they said well here’s this X amount of money, we are already giving it to you. If your students reach a particular target then you’ll just get to keep it and if they don’t then sadly, we’ll have to take it away. Now in essence, if you think about it, they’re exactly the same. Only those teachers are going to get the money, who meet the certain performance requirements, but in one, they are only going to get after and in the other, they got it first but if they don’t reach, and just this, improved the performance of the students 4 times. So again, there was same amount of money that would have been given but just thinking about when to present that, that brought about a big difference. And I think, maybe, just one more that I can use to give you an idea about this. Here’s an example on vaccination, getting parents to vaccinate their kids and quite often the intention is there but somehow people don’t end up carrying out that behaviour and in this piece of work, the behavioural insight was around getting people to plan. So they created this calendar and they went home and gave that calendar, and the person who would give the calendar to the parents would ask (them) to mark down the date and the time when they think they would be able to do and they would request them to hang that calendar on their door. And just this act of planning that yes, next Sunday I think I am going to go around 10 and then marking that down, that itself increased the number of people, number of parents who then went on. These are just a few examples. The basic principle is understanding the way our mind works, trying to leverage these insights and develop very low-cost solutions easy to implement solutions, which themselves are what people would want to do.

[11:13] Tabindah Anwar: And these interventions that we have based on these behavioural insights we call them nudges.

[11:20] Dr. Umar Taj: and yes, a nudge is quite often a label that is used for interventions that are using behavioural insights.

[11:28] Tabindah Anwar: Okay, thank you. Bilal, my question to you is, that you’ve mentioned that Gallup Pakistan has weekly trackers of opinion polls. I am curious to know mostly from the recent polls, what has been some of the deepest insights and trends?

[11:48] Bilal Gilani: Thank you so much. Apart from just doing public opinion polling, we have also been polling other segments of the society because their voices are also crucial. For example, we have recently completed a survey of doctors in the country, the front-line workers as we call them in the report, we also looked at the voices of students because the student community has also been impacted because the universities have closed. We looked at students who are in the school community and are now either studying from home or exposed to the tele-school initiative. We’ve been working on understanding the pulse of the businesses in the country in order to understand where they stand in terms of their economic situation with respect to employment that they’re offering, and so the public opinion is not just restricted to the public at large but also segments of the public that we have been targeting.

I have four key points that my learning through the public opinion data across the different publics that we have been studying for about 4 months now, and I thought they provided a good insight, a window into understanding the Pakistani public and to some extent, the public experience across the world as well. The first one being that the public follows leadership and is very sensitive to how the environment around them is. This is to some extent also a learning from the behavioural economics field that people are not rational, they are also impacted by the environment that they’re living in. We followed how people’s perceptions on the issues related to COVID changed. Right at the beginning when the government was ambivalent about lockdown, we saw that the support for the lockdown in the public was also low. And when the lockdown was imposed and it seemed that the entire leadership in the country was behind lockdown in a major way, the highest we had in one of our weekly polls when the lockdown was continuing, about 78% of the people saying that the lockdown was something that was required and needed, and although they were suffering economically, this was something they were willing to suffer because this probably offered a long term solution and a quicker solution. And then, we started seeing how generally the lockdown situation and the policy or at least at the government level, the consensus started to move away from the lockdown and towards smarter lockdowns or other forms of implementation of SOPs. And then, the support for lockdown started to fall and in one of our, I think 5th wave, around after the Eid that has just passed, we saw that the support for lockdown had gone to single digits. And most people were not in the favour of lockdown.

Another key learning for us was that the impact of COVID-19 was operating in very strong power hierarchies. So the impact was the most felt and the most aggravated for the lowest in the power hierarchy. And it was that particular segment of the population that had the lowest voice in the power corridors. So we looked at how employment in the country was being affected so in our polls about 1 in 2 employed people were saying that they had either been let go of completely or they had been sent on paid or unpaid leave. And we looked into it and we saw that people who were well-to-do or people who were being paid significant salaries, this phenomenon was only 1 in 10. Whereas for the lowest paid employed person, the phenomenon was as high as 7 in 10 were saying that they were unemployed. So for us the learning was that in the public policy operates in significant shades of being successful and the interventions we design, we need to be very conscious of the hierarchies of power that exist within the societies we operate in. And unless we think through those hierarchies that exist and try to penetrate, and to create solutions to those hierarchies, the hierarchies will then merge themselves in the solutions we are creating.

And the last one being that there has been enormous success in media communication with respect to COVID-19. But they were operated in pockets. They were some left out pockets. And those left out pockets were areas that were not digitally connected. And we saw that one shade that really penetrated through all was not being digitally connected so rural people in certain communities, in certain provinces, people above the age of 50, they tended to have less success, the government tended to have less success in being able to reach out to them. So digital hierarchy or digital divide was also a very crucial learning that when we looked through our data, we could see that to be a very important one.

 [18:31] Javeria Masood: Very interesting things you’ve shared, both of you. And Bilal, when you were saying about intentionally designing interventions, to also then tackle these kind of complex problems of hierarchy which obviously has been there forever. These are not things that have just come to surface, these are very systemic elements that have existed. So how can one be intentional when you’re designing these interventions and you’re using BI so how can you intentionally design something where you can actually change something?

[19:10] Bilal Gilani: Maybe, I can talk briefly about this, and Umar can shed more light from international experiences in designing interventions that cater for these. For us over the past 4 decades that we have worked with policy makers in order to design policies that look through and adjust themselves to these existing hierarchies, we advise two significant steps before intervention starts to happen. One of them is to do exploratory research about the subject, and about the target market that one is looking to intervene in. The research doesn’t necessarily have to be primary in the sense of collecting fresh data. So when we are talking about intervention geared towards making people comply to SOPs with respect to COVID we have an identified target market or a target audience and we really need to first understand the target audience in terms of demography, in terms of social indicators, psychographic indicators, and anything that we can do to understand our target market well before we actually start to think about them. So a lot of design thinking essentially, and its principles are a crucial first step into designing interventions. Design thinking for example provides a great step by step guide into thinking along those lines. And the second is: a lot of testing before actual implementation starts to happen. So testing provides an opportunity to make sure that the voices of the actual people who could be exposed to such a message or actual voices of people who would be exposed to these interventions, will have to live with them for example. We spend significant time with them, quantitatively, as well as living through with them qualitatively. It doesn’t necessarily have to be very large samples. So with these two intervention points doing a lot of exploratory research before thinking about intervention and when a prototype of an intervention is available, testing it significantly. So we are able to make sure that we have a good understanding of where we are doing it, why we are doing it, and how it would be perceived. This can help. This can really help. It doesn’t offer a sure solution but it definitely does help to create interventions that are well-tailored to the segments of society and is cognizant of the hierarchies that we would eventually encounter. As you rightly said, these are created, these are persisting over centuries. But I would be curious to hear what Umar has to also contribute to this and how BI solves this puzzle for policy makers.

[22:36] Dr. Umar Taj: I would certainly agree with both the points that you said. I think the couple of additional points that I would like to add are: so first co-creation, so I think this kind of an exploratory thing, I think that itself lands well to this idea of co-creating solutions with the people for whom you are trying to create an intervention for. I think the use of behavioural insights goes into the methods that you are using to understand. So you want people to do X for example and to figure out, well, why aren't they doing X, then what behavioural insights provide is this vocabulary and perhaps some sort of a guide as to what other things that we should look out for. I think that’s one. The second bit is around diversity. When it comes down to generating solutions. So quite often the reason that we still have interventions that work very well for one sector of the hierarchy and not for the other is just because people who are creating the solutions are also probably from that one level of the hierarchy.

And therefore their view of how the world is, is totally influenced by people around them, people who they know. And I think that, so within the BI domain, we call it the curse of knowledge. This phenomenon that you have been working on it for a while and you have this perception that the actual out there world is very similar to the world that I know. I think this also exacerbates this problem of interventions that might only be helping one level of the hierarchy. So it might just come down to simply realizing that ‘ah’ so we, why are we only thinking about delivering message through X. We do know that there is already this big network that’s out there, why have we never thought about using that network, because it already communicates with that target audience? We miss those quite often just because it’s always the same group who is trying to think of solutions and the new voices usually don’t come in. Who can sort of have this outside view and just listen and say “Oh why haven’t you ever thought of doing this?” So these kind of questions are really important when we are developing interventions.

[25:34] Tabindah Anwar: Okay, thank you. Dr. Umar, I’m glad that you are, you know, sharing about this because I know that you have been working very closely on applying behavioural insights to risk communication for COVID in Pakistan. Can you share some examples of your work with the wider network where you applied BI to shift perceptions and improve decision making?

[25:58] Dr. Umar Taj: Sure, sure. So just generally the application of behavioural insights within curtailing COVID just across the world has all been focused quite a lot on communication.

So it’s stuff like how are we communicating? Who is communicating? And then the things that we want people to do, how are we presenting those options? That’s it. So in Pakistan as well, what I’ve been involved is a little bit into helping people think about all of this communication that is going around. So very specifically, it’s thinking about the content, the delivery, who is delivering, and then the third is the channel, how should it be delivered. Some examples are, there has been a ringtone message that, for those who are in Pakistan they would have definitely, almost certainly would have heard that message, but for those who are outside Pakistan, just for them—when you call somebody on the phone while it’s ringing time you could play a message. And so in Pakistan that’s one thing that has been going around. When you call somebody, a message starts until the other person picks the phone. And so I have been working with a team to try and share some insights on how to frame that ringtone message in itself. So some of the things that we already know from behavioural insights for example, we do know that fear based messaging doesn’t work. It has to be more around self-efficacy. Our discussed base messaging probably works. So we have been trying to avoid fear based messaging and trying to put more idea on what you can control, and these are the things.

Second is about, we are more likely to take action that if as a result of which there is consequence for our family. So if we tell people do this because you’re going to get hurt or if we tell people do this because your family might get hurt, people are more likely to do the action  when we tell them that their family might get hurt. So one of the things we are trying to do is shift this focus and saying well actually we want you to do these things not only for you but it’s actually for your family. The second is around helping the digital campaign that is being run with the help of UNDP and other UN agencies focusing specifically on youth. So again, within that it’s around the content that gets created. So there are videos and with the kind of insights that we could show in those videos, just to give you one example for instance. We need to visualize the virus, right, it’s quite hard, it isn’t visible to show well how it spreads. Somehow bringing this visualization, just through the use of color, that actually helps people. People understand, ‘ah’ so this is how it will spread, this is why I need to do X. Or just explaining to people, people are more likely to do a particular action if we give them reason why not. A lot of times we only instruct but we don’t increase their knowledge. So explaining first this is how it spreads therefore the action is, wear mask, etcetera.

Another one is around robocalls. So where, and this has been happening in some provinces where robocalls, you know, people would receive that with some message and what’s quite important, and this is something we have learned and again Gallup and their surveys are showing again and again which is who are people listening to most when it comes down to messages related to Covid. And it’s overwhelmingly, I think it’s around 70% or so, is doctors. So people are most likely to listen to a message when it is delivered by a doctor, which is really important. So when you are delivering a message, having a messenger, so who is the messenger. It’s a really important point. When you think about robocall, what’s important is who is it coming from. And so we are doing some work to inform, one, who it should come from and then again, what should be the order, so for instance, when it is a robocall you can just hang up right away. So the first thing that we want is for people to just hold on, listen to it. Which then means that the most important information has to go right upfront and it has to capture their attention. So doing some work around or thinking, doing some thinking around what it is that should go right up to capture the attention of people. And I think the other bit then is, is just thinking about communication channels, so which is, we, just to give an example for instance, we already know and again, thanks to Gallup, and this, you know the idea, the thing that Bilal was talking about which is informing. You know, having evidence based decision making. So we already know that communication channels that have the most reach are television and mobile phones, right. Now the issue is if you were going to go with television it’s really expensive, so then comes the question well, what can be done and here is the, here’s an interesting one that I’ve proposed to the ministry of health which is that every day you have a daily press briefing which means that there is already this, you already have free space to put a message out which is at the background of where people are speaking and then also in front. So if you imagine a press briefing, you have the person but then you have the background wall, but then you also have the space in front of the table. And it’s just thinking a little outside the box, which is well, we know TV has the biggest reach, we know it costs, but is there a way, and it’s things like these, which aren’t really behavioural insights but which is just thinking a little outside the box. And there are possibilities definitely that we could work with. And again, the key insights were one, fear based messaging will not work. We already know this from literature. The second is, we need to be very specific, so as soon as you give some sort of an ambiguity, then it would give more of a chance for people not to do something.

[33:22] Tabindah Anwar: Okay. Thank you. Bilal, my question to you is that from Gallup, you know, the perception studies and the insights that have been gained, I want to know more about how, as we say, you know, closing the loop. I know Dr. Umar has mentioned some examples where Gallup’s perception surveys have been used but I want to ask you as well, that were these insights used to inform any decision or response work?

[33:51] Bilal Gilani: So our different studies have been used at different levels. I think one of the crucial areas where the data has been used is the media communications part. Our current work and our past studies around media habits of people have come in very handy for communications purposes in order to understand who is listening to what medium and to what extent and at what times. So if one is targeting someone in the KP rural areas and within the age bracket of 18-30, what’s the most appropriate medium to use? So we’ve seen our works being quoted, for example, by UNICEF which is working around reaching out to populations. We’ve also seen our work being quoted by Ministry of Health and other humanitarian organizations working on this. So we did a couple of studies on how the labour force in the country is being impacted and how the household economy has been impacted. I talked about employment, for example. So we’ve seen the data being used by the Ministry of Economics and we are also seeing ministers using the numbers at the floor of the parliament, citing numbers and justifying the policy decisions that they were doing. So for us that’s one intervention point and the objective being met that decision is being made by the policy makers, not in the darkness of lack of empirical data, but data being used and decisions happening.

[35:58] Tabindah Anwar: Yeah, That’s true. That’s true. So at UNDP Pakistan the innovation acc-lab mainstreams innovation in our country office. One of the workstreams is building innovative capabilities, and introducing innovative approaches in our programmes. My question to Javeria is that, I want to ask you what is Acc-Lab’s vision on including behavioural insights and expanding on our toolkit of innovative approaches?

[36:30] Javeria Masood: So maybe before I answer that question I can hint on some of the things that are already, you know, part of the toolkit before this further expansion. So a lot of design practises like Bilal also talked about design thinking, a lot of our practises and a lot of our deep exploratory work is very heavily embedded with ethnographic practises. Essentially that in itself is, you know, a study and understanding of people and their context to identify why somebody is saying something, where they are coming from, what are some of their influences, not just in terms of environment or people but also a lot of cultural nuances that we follow. So it’s a lot of the times, after these understandings and deep dives into these social psychology domains when we start figuring out that for which project or which stream of portfolio we need which methodologies. So tool kit is just an apt word. There are a lot of offerings like, you know, design practises or design thinking, there is systemic design, there is strategic foresight. And they all interlink, so it’s often not that just one is picked, all of these tools and methodologies have an undertone of each other and specifically the behavioural insights frameworks. But what’s really good now is that we are kind of also taking BI as a separate offering, in the sense that a lot of the testing that Umar and Bilal talked about, there is a lot of emphasis now on that. That even before you are proposing an intervention, which is very well supported by the ethnographic work, you should test it. So whether that is through, you know, perception surveys or that is through focus groups or different tools or even try to figure out that are you even able to reach the larger audience if there is a digital divide or a collapsed or an incomplete infrastructure around that. So then what are the BI tools or what are the design practises that you can incorporate. So, you know, like you said that one of our mandates at the Innovation Acc-Lab is to mainstream innovation and a very strong element of that is to build capacities. So not only are we, you know, reaching out and identifying traditional and non-traditional partners who are doing interesting work, such as what Bilal and Umar shared and they, I’m so glad that they, you know, very clearly, explicitly used a lot of examples to really deconstruct and help us understand what each of those means in terms of technicality, what were the tactical aspects, what were the strategic aspects of it. So this again also helps us largely inform our audiences, our partners, our friends and the larger community. So our vision really is to build a community, mobilize a community and you know, have this community of practise around it.

[39:42] Tabindah Anwar: Okay, thank you. Dr. Umar, Bilal, Javeria, thank you so much for your time. We will continue to follow the perception polls by Gallup Pakistan, and we are working closely with Dr. Umar and UNDP Pakistan.

To our listeners, thank you for listening. This is Solutions Cast!

UNDP Around the world

You are at UNDP Pakistan 
Go to UNDP Global