BEAR+BI Org Special Issue

Special Issue: Behaviourally Informed Organizations

MANAGEMENT

SPECIAL ISSUE: Behaviourally Informed Organizations

Features

6

Harnessing Behavioural Insights: A Playbook for Organizations by Bing Feng, Melanie Kim, Jima Oyunsuren, Dilip Soman and Mykyta Tymko

14

Thought Leader Interview: Richard Thaler by Dilip Soman

20

Changing Behaviour for Good by Angela Duckworth and Katy Milkman

32

Consumer Behaviour Online: A Playbook Emerges by Dilip Soman, Melanie Kim and Jessica An

50

Managing Mental Health: A Behavioural Approach by Renante Rondina, Cindy Quan and Dilip Soman

Bridging the Organ Donation Gap with Behavioural Science by Nicole Robitaille and Nina Mažar

60

Behavioural Science in the Wild

by Michael Sobolev

66

The Behavioural Science of Online Manipulation

by Elizabeth Costa and David Halpern

72

Cash Transfers to Address Homelessness

by Kenneth Ong, Daniel Daly-Grafstein and Jiaying Zhao

78

Leaning In or Not Leaning Out?

by Joyce He and Oceana Ding

104

Defaults Are Not the Same — By Default by Jon Jachimowicz, Shannon Duncan, Elke Weber and Eric J. Johnson

92

Banking on Behavioural Science: Commonwealth Bank of Australia by Jingqi Yu, Sanskriti Merchant, Bing Feng and Dilip Soman

98

PwC Consulting: Strategic Value by Sanskriti Merchant, Ankita Suresh and Jingqi Yu

110

The Between Times of Applied Behavioural Science

by Dilip Soman, Jingqi Yu and Bing Feng

Idea Exchange

26

POINT OF VIEW:

POINT OF VIEW: The Elements

Context by Selina Yang, Meghan Yeung, Nathaniel Barr, Chang-Yuan Lee, Wardah Malik, Nina Mažar, Dilip Soman and David Thomson 88

EARLY

90

EARLY

POINT

47

POINT

57

115

BI-ORG BOOKS:

Rotman Management

Behaviourally Informed

Published in January, May and September by the Rotman School of Management at the University of Toronto, Rotman Management explores themes of interest to leaders, innovators and entrepreneurs, featuring thought-provoking insights and problem-solving tools from leading global researchers and management practitioners. The magazine reflects Rotman’s role as a catalyst for transformative thinking that creates value for business and society.

ISSN 2293-7684 (Print)

ISSN 2293-7722 (Digital)

Co-Editors

Oceana Ding and Dilip Soman

Contributors

Elizabeth Costa, Angela Duckworth, Bing Feng, David Halpern, Joyce He, Matthew Hilchey, Eric Johnson, Melanie Kim, Nina Mažar, Katy Milkman, Piyush Tantia and many other members of the Behaviourally Informed Organizations partnership.

Contact Us/Subscribe:

Online: rotmanmagazine.ca

(Click on ‘Subscribe’)

Email: RotmanMag@rotman.utoronto.ca

Phone: 416-946-5653

Mail: Rotman Management Magazine, 105 St. George Street, Toronto, Ontario M5S 3E6, Canada

Subscriber Services:

View your account online, renew your subscription, or change your address: rotman.utoronto.ca/SubscriberServices.

Privacy Policy:

Visit rotman.utoronto.ca/MagazinePrivacy

Design:

Bakersfield Visual Communications Inc.

118

POINT OF VIEW:

Copyright 2024. All rights reserved.

Rotman Management is printed by Mi5 Print & Digital Solutions

POINT

121

Rotman Management has been a member of Magazines Canada since 2010.

The BI-Org Nerve Centre

Liz Kang, Cindy Luo and Yanyi Guo

WHAT WAS IT LIKE TO WORK ‘behind the scenes’ at the Behaviourally Informed Organizations (BI-Org) office at the BEAR research centre? What most people don’t usually get to see are the human mechanisms and processes that play an instrumental role in supporting the activities and outputs from the partnership. The three of us would like to share our reflections on behalf of everyone who was part of the BI-Org core team.

Our mandate was simple: to make it (the partnership) work! We were responsible for numerous tasks, including facilitating collaborations, reporting requirements, budgeting and managing finances, running events, coordinating multiple book projects (plus this magazine issue), and providing research support. This happened through the first global pandemic in a century, as well as several changes to personnel as colleagues here and in our network moved on to new opportunities and new members came aboard.

Each of these activities and tasks came with their own trials and tribulations at times, but we were surrounded by team members who were truly supportive and reliable in every way. We would argue that the origin of this supportive organizational culture stemmed from the behaviours modelled by our leaders at the helm. Leaders who provided guidance and feedback — both positive and constructive — actively encouraged continuous learning and career growth, promoted a healthy work-life balance, often showed appreciation and attributed successes to

the team with humility (even when they really deserved the spotlight), and had everyone’s (idiomatic) backs.

This empowered us to confidently tackle our responsibilities and assist the BI-Org faculty, student researchers and partner organizations with their research projects and related activities. Knowing that support was readily available allowed us to ask for suggestions when we were creating videos, animations, presentations, or infographics without any hesitation. It also meant that our interest in taking educational courses or workshops would be accommodated. We were granted the flexibility to work in a hybrid environment, and our well-rounded team learned to adapt and approach challenges by turning them into opportunities.

Under the leadership of humans who really understand the nuances of human behaviour, we are thankful for the boundless support that enabled us to meaningfully contribute to the partnership. We learned immensely from these experiences and will undoubtedly apply these learnings to other endeavours. It is bittersweet to see this partnership come to an end, but we know we can never really leave, especially after forming such memorable experiences and bonds at BEAR.

Liz Kang (BCom 2016), Knowledge Translation Specialist. Cindy Luo (HBSc 2015), Research Officer. Yanyi Guo (Rotman MBA 2023), Research Associate, Behaviourally Informed Organizations.

FROM THE EDITORS Oceana Ding and Dilip Soman

Behaviourally Informed Organizations

FOR THE PAST FIVE YEARS, the Behavioural Economics in Action at Rotman (BEAR) research centre has been home to an initiative known as the Behaviourally Informed Organizations (BIOrg) Partnership. BI-Org is a collective of academic researchers, government units, for-profit entities, consulting firms, and consumer groups across the world working on how best to embed behaviourally informed thinking and practices in organizations. The articles in this special issue of Rotman Management magazine are some of the outputs produced over the course of the partnership. This issue has also benefited from inputs of the BI-Org core team, as well as from Iris Deng (for conceptualizing the cover), Sophia Khan (for editorial assistance) and Sarah Lazarovic (for the Thaler Rules illustration).

Highlights from this issue include an interview with Nobel Memorial Prize winner Richard Thaler, a short history of BEAR, and an article by Dilip Soman, Bing Feng and Jingqi Yu on the challenges to the adoption of applied behavioural science in organizations. We showcase two case studies from our partnerships with PricewaterhouseCoopers (PwC) and the Commonwealth Bank of Australia that illustrate what it takes to build in-house behavioural science capabilities. Nicole Robitaille and Nina Mažar, former BEAR members and now professors at Queen’s and Boston University, respectively, present an initiative that helps Ontarians who intend to sign up to be organ donors do so. Jiaying Zhao discusses what behavioural insights can teach us about creating successful cash transfer programs, and Joyce He shares findings from her research that show how certain choice architecture interventions can help close the gender promotion gap.

Beyond these highlights, you’ll find articles that discuss whether budgeting really works, whether choice architecture interventions are ethical, and how frequent changes in policy can be better framed to build public trust.

We recognize that most organizations are in the business of behaviour change, so understanding how behaviour change comes about is critical to organizational success. We hope that this issue teaches you something new about how to apply behavioural insights and inspires you to approach your work with a behaviourally informed lens.

Oceana Ding, Co-Editor

Dilip Soman, Co-Editor

HARNESSING BEHAVIOURAL INSIGHTS: A Playbook for Organizations

Some of the smartest organizations are moving behavioural insights up the value chain and embedding them into their everyday processes.

by Bing Feng, Jima Oyunsuren, Mykyta Tymko, Melanie Kim and Dilip Soman

WHILE THEY GO ABOUT THINGS in very different ways, at their core, every organization is actually in the same business: behaviour change. Whether it is a for-profit firm encouraging consumers to switch to its product; a government agency trying to get citizens to pay taxes on time; or a health agency interested in improving the consumption of medication, behaviour-change challenges abound.

Many organizations struggle to make behaviour change happen due to a fundamental empathy gap. ‘Econs’ — as depicted in economics textbooks — are hypothetical individuals who have well-defined preferences, are able to accurately predict the future consequences of their actions, have immense computational abilities and are unfazed by emotion. Humans, on the other hand, are cognitively lazy, impulsive, emotional and computationally constrained. The empathy gap occurs when organizations design products and services for econs, when in fact, the end-user is a typical human being.

Not surprisingly, the day-to-day behaviour of humans differs

significantly from that of econs. Factors including context, cognitive laziness, procrastination and social pressure play key roles in human decision-making. The emerging field of behavioural insights works to connect the psychology of human behaviour with economic decision-making to explain these phenomena.

Over the past 10 years, we have seen a great deal of progress in the application of behavioural insights (BI). With thousands of trials being run by hundreds of public and private-sector organizations around the world, human behaviour has become a key focus of activity in the policy, welfare and business world.

Inspired by the growing interest in BI from all sectors — as well the absence of any formal guidelines for embedding them within an organization — Behavioural Economics in Action at Rotman (BEAR) has published a ‘playbook’ to help interested leaders navigate the realm of BI. In this article we will summarize its key messages, beginning with a description of the four roles that BI can play to create value for virtually every organization.

We can proactively design a ‘choice context’ to nudge users in a particular direction.

ROLE 1: BI as Problem Solver Behavioural insights can enable an organization to address problems arising at ‘the last mile’ — those moments where an end-user directly interacts with your organization or its product or service. Whether your last-mile issue is a low take-up rate, poor sales or low conversion rates, BI can be harnessed to make subtle changes to better align your product or service with human behaviour.

One of the key tools that enables problem-solving here is the concept of ‘choice architecture’, whereby organizations proactively design the ‘choice context’ in such a way as to steer or nudge users in a particular direction. Following are the four main types of decisions that choice architecture can influence:

COMPLIANCE. Getting people to act in accordance with a regulation set by a government or agency (e.g., tax deadlines, regulatory paperwork requirements).

SWITCHING. Getting people to convert from one choice to another (e.g., brand switching, replacing soda with water at meals).

FOLLOWING THROUGH. Getting people to follow through on commitments that they themselves have made (e.g., completing a weight loss regimen, or just acting on intentions).

ACTIVE CHOOSING. Getting people to break undesired habits by converting passive, mindless decisions into active choices.

It is fair to say that the biggest successes for BI to date have come in the domain of choice architecture. For example, a few years ago, Ontario’s Behavioural Insights Unit set out to increase organ donations in the province. Working closely with project partners ServiceOntario and Trillium Gift of Life Network, the team’s goal was to increase the number of people who register as organ donors.

Initially, several ‘barriers to registration’ were identified, including the length and complexity of the registration form, failure to ask every customer if they wanted to register and asking customers to complete yet another transaction after they had waited in line and completed other paperwork.

The result: By removing these barriers — and thereby enhancing the choice architecture — registration rates increased

by 143 per cent. The organ donor registration process in Ontario was redesigned based on the following choice architecture principles:

• Provide different versions of the donor registration form;

• Change the timing of when the form is handed out; and

• Offer additional information to help people make their decision.

ROLE 2: BI as Auditor

At the end of every design process for products, services or processes, a behavioural scientist can be tasked with auditing the outcome and evaluating it for human-centricity. In this role, BI is used to evaluate and provide suggestions for further ‘humanizing’ organizational outputs.

The federal government’s Impact and Innovation Unit (IIU), along with BEAR, worked with the Canadian Armed Forces (CAF) on its goal to increase the percentage of women in its ranks from 15 to 25 per cent by the year 2026. The team used a BI lens to audit the following stages of the recruitment process:

THE APPLICATION PROCESS. BI-informed changes were made to the Armed Forces’ application form to increase clarity and understanding. Furthermore, the Department of National Defense (DND) made improvements to the appointment process and recruitment follow-ups.

RECRUITMENT MARKETING. The IIU conducted a social media marketing trial aimed at understanding ‘what works’ in engaging Canadian women with a career in the Armed Forces.

POLICY AREAS AND GUIDELINES WITHIN THE CAF. A number of policy areas, including deployments and relocation, leave without pay, childcare support, and long-term commitments were identified for consideration and BI-inspired improvements.

The result: The collaboration was successful and has resulted in changes to the CAF’s marketing and recruitment efforts and an increased appetite for experimentation in these areas. Key success factors included buy-in from executives at both the CAF and the DND. This was essential in allowing the IIU access to the relevant information and in implementing recommendations. In addition, the IIU was equipped with well-

rounded staff that included both quantitative and qualitative researchers. The relationship between the IIU, CAF and DND has resulted in continued engagement and testing. The DND has also recently established its own BI team: Personnel Research in Action (PRiA).

ROLE 3: BI as Designer

In this role, a behavioural scientist is involved from the outset, ensuring that the design of a product, service or process is behaviourally informed from the start — which means that the process always begins with observing actual users in a natural setting.

One of the key concepts in this domain is ‘behaviourallyinformed design’, an approach that combines the principles of BI with those of ‘design thinking’. The BEAR team has worked with a couple of government branches to employ BI to improve consumer protection. At these agencies, BI is now used to better understand the behaviour of consumers and agents in different marketplaces. The agency continues looking for opportunities to apply BI to both its policy design and its daily operations.

Three Truths About Behavioural Insights

1. PEOPLE ARE ONLY RATIONAL TO A CERTAIN EXTENT. We often assume—sometimes implicitly—that people balance the pros and cons and assess risks on the basis of all the available information, and thereby make well-considered and consistent choices. The behavioural sciences teach us that the choices people make are only rational to a limited extent. One example of this is that people are more sensitive to a loss than to a gain . As a result, they put in more effort and take greater risks to avoid loss than to win the same amount.

People also follow certain rules of thumb or shortcuts in processing information. One example of this is the availability heuristic : for events that we can recall more easily—for example, because they were distinctive or emotional or recently in the news—we overestimate the likelihood that they will occur again. These deviations from rationality are partly predictable, which means we can take account of them in policy-making.

2. PEOPLE HAVE LIMITED SELF-CONTROL. It takes effort to resist temptation and suppress impulses, and people have limited available resources for this. They want to eat more healthily, exercise more often or save for their retirement, but in practice it turns out to be harder than they thought. One consequence of this is that there is a major difference between planning to do something and actually doing it — between intention and action. Everyone has experienced this conflict at some point: planning to do some chores but ending up slumped on the sofa; starting a diet but ending up in a burger bar after just a week; wanting to save but still going out for dinner.

Another example of this role is the growing realm of selfcontrol products. Products and services are being created to enable customers to close the ‘intention-action gap’ — whereby people intend to do something positive but fail to act on it. Examples include Clocky, an alarm clock that literally runs away when the sleepy user hits the snooze button — propelling itself off of the bedside table and across the floor, out of reach of the dozing user; and stickK, a website that uses incentives and peer-effects to encourage people to stick to their goals.

The BEAR team also consulted with a government agency that set out to raise awareness of consumer protection issues, targeting adult consumers of all ages who were highly educated, but with limited knowledge of behavioural science. Working together, we designed its information brochure using the following BI guidelines:

• Ensuring that the colour and visual appearance were appealing to the public.

• Balancing image and text ratios to optimize comprehension and readability.

A psychological phenomenon relating to this is that people perceive a reward in the near future as more valuable than a reward further in the future. This plays a role in trade-offs between present and future rewards, such as pensions and savings, but also with respect to health.

3. PEOPLE ARE INFLUENCED BY THEIR ENVIRONMENT.

In order to determine what the ‘right’ behaviour is, people often look at the behaviour of others, particularly in new or uncertain situations. In a classic example of this, researchers conducted a study in which a group of confederates unanimously gave the wrong answer to a very simple question. The participant—who was unaware of this— then conformed by also giving the wrong answer. The news and government communication often show undesirable behaviour in order to stress its objectionable nature; however, often this actually gives the subconscious signal that this is ‘normal’ behaviour.

It is not just the social environment but also the physical environment that can have a major influence on behaviour. If fruit is within closer reach than chocolate, it makes it easier for canteen visitors to make a healthy choice. And optically narrowing stripes on the road prompt people to drive more slowly of their own accord. The physical environment can also communicate a particular social norm. An environment with a lot of litter on the street provides a lot of information about other people’s behaviour, and can therefore lead to more litter.

– Courtesy of the Behavioural Insights Network Netherlands

• Managing the aesthetics of the cover to maximize the likelihood of the brochure being read.

The website for this agency was also improved by using appropriate language to maximize engagement. For instance, making sure that the negative connotations of words such as ‘hazards’, ‘risk’ or ‘unhealthy’ do not reduce attention by creating negative emotions; and testing the balance between visual and text information to minimize the risk of information overload. The results were impressive and included increased traffic to the website; an

increase in public debate and discussions on the site; and greater public reporting of marketplace frictions.

The following research findings indicate the breadth of opportunities in this domain:

• Telling people about positive, pro-social things that others have done (as opposed to asking them to stop doing a certain negative thing) can produce a fourfold increase in the number of people being more pro-social;

• Showing people smiling or frowning faces — a small emotional trigger — on their electricity bills improves resource

The ABCD Framework for Using Behavioural Insights

A: ATTENTION – Make it relevant, seize attention and plan for inattention

Attention is the window of the mind. However, attention is scarce, easily distracted, quickly overwhelmed, and subject to switching costs. Practitioners will often find that attentional issues have been overlooked in the design and implementation of traditional public policies. For this reason, when practitioners find a behavioural problem with attentional issues, it may prove more effective to design policy interventions that are more relevant, seize attention and, if this is not possible, think about how to plan for inattention.

B: BELIEF FORMATION – Guide search, make inferences intuitive, and support judgment

While there is no such thing as too much information in a traditional public policy perspective, information overload has become a serious problem for the people inhabiting the real world. For that reason, problems in belief-formation usually go hand in hand with the vast amounts of information and possibilities that are put on offer. In this perspective, it is not surprising to find that some of the biggest companies today are companies built around information search engines and consumer comparison platforms. What is perhaps more surprising is that traditional public policy interventions with regards to problems of beliefformation have been slow in copying what these companies do, but instead often try to approach problems in belief-formation by offering even more information.

C: CHOICE – Make it attractive, frame prospects and make it social

When making a choice is difficult, people are likely to be influenced by biases and heuristics in their decision-making. ABCD

suggests that researchers look into making preferable choices more attractive, use framing of prospects and leverage social identities and norms. The framing and arrangement of prospects is perhaps the most famous, but also the most technical area of BI as applied to public policy. In facing a series of choice-options, a person also faces a series of possible futures, i.e., prospects. While making it attractive provides reasons for choosing, the framing of prospects influences people to choose one or another option in subtle ways independent of what is chosen and why. That is, one option may be chosen over another simply due to the way that choices are presented – either as a matter of arrangement or as a matter of formulation.

D: DETERMINATION – Work with friction, plans and feedback and create commitments

Most people know that it is easy to form an intention of doing something. It is much harder to get it done. However, we do not always anticipate this and tend to systematically overestimate our own ability in taking the small steps to accomplish our goals. Thus, choosing to do something is not the same as succeeding. The world is complex and when any one person has to juggle multiple goals at once, even relatively small obstacles may become a reason for postponing taking action. As a result people tend to procrastinate leading to inertia and staying with the status quo.

–Courtesy of the OECD Behavioural Insights Toolkit

conservation by 25 per cent; and

• Exchanging similarity-identifying information before a negotiation can boost the successful outcome rate from 50 to 90 per cent, with an average 18 per cent increase in perceived final value across parties.

ROLE 4: BI as Chief Strategist

In these cases, every touchpoint with an internal or external stakeholder is run from a behavioural perspective. The organization is human-centric in everything that it does, with behavioural insight as its main operating principle.

In these cases, firms can create BI-informed decision tools that help agents make better choices by providing feedback, rules of thumb, computational support, decision support or peer comparisons. For example, with a mandate to improve financial literacy and facilitate positive behaviour change, the Financial Consumer Agency of Canada (FCAC) started offering financial tools and calculators to educate Canadian consumers and help them make better financial decisions: Its Mortgage Calculator helps determine a mortgage payment schedule based on user inputs and allows the user to input different pre-payment options to show them how they can save money; its Budget Calculator helps consumers get a portrait of where their money comes from and where it is going; and its Financial Goal Calculator helps people figure out how to pay off their debts or reach their savings goals.

Another example is Evree, whose app connects to your bank account and helps you save money for things that really matter to you. It basically employs many of the same behavioural science tactics that other companies use to make you part with your cash — but uses them to help you make smarter financial decisions instead.

All members of the Evree team, including management, content designers and engineers, are trained to understand the basic principles and applications of behavioural science. The objective is for the entire team to resolve problems using a scientific method and to approach daily operating roles with a BI lens. According to Stephanie Bank, a behavioural economist at Evree, “We look at behavioural science as our foundation, not just as a tool for a designer or a problem solver. All of our staff members are trained to look at business problems and opportunities through a behavioural lens.”

How to Get Started

Before deciding which of the four domains can create the most value for your organization, there are two key decisions to be made:

Embedding BIs in Organizations: Four Approaches

Concentrated Expertise

Diffused Expertise

CAPACITY BUILDING

INTERNAL CONSULTING BEHAVIOURALLY INFORMED ORGANIZATION

1. THE LOCUS OF EXPERTISE: whether to set up a concentrated team/unit within your organization OR to diffuse expertise across the enterprise; and

2. THE LOCUS OF APPLICATIONS: whether to use BI in a narrow application where they are applied to a specific geographic location or department OR to use BI in broad applications where it is applied across domains, geographies and departments.

These two decisions create four main approaches to embedding BI in an organization.

THE FOCUSED APPROACH. Employment and Social Development Canada has its own Innovation Lab comprised of behavioural scientists, data analysts, designers and policy analysts. The Lab works on projects with internal partners to tackle problems using a combination of human-centred design and BI methods. Its full-scale design project for 2017 was the Canada Learning Bond, which found ways to increase uptake and better understand perceptions of education and financial decision making among low income families.

FIGURE 1

Four Roles for Behavioural Insights

The valuecreation chain of product development

Roles that BI can play

Pre-marketplace: Within the Organization

Stage 1: Define goals, strategy and operating principles

Stage 2: Design and develop product, program or service

Stage 3: Prepare to go-to-market

In the Marketplace: The Last Mile

Stage 4: In-market: End-user interacts with product, program or service

THE CAPACITY-BUILDING APPROACH. The federal government’s Impact and Innovation Unit (IIU) houses expertise in four areas: innovative finance, partnerships and capacity building, impact measurement and behavioural insights. It offers services through a core unit at the Privy Council Office and through the fellowship model, in which scientists are deployed to other Government of Canada departments and agencies to provide behavioural science expertise and run behavioural insights trials.

THE INTERNAL CONSULTING APPROACH. The World Bank’s Mind, Behavior and Development Unit (eMBeD) is a behavioural science team housed within the Bank’s Poverty and Equity Global Practice. The team works closely with World Bank project teams, governments and other partners to diagnose, design and evaluate behaviourally informed interventions.

THE BEHAVIOURALLY INFORMED ORGANIZATION APPROACH. These examples and cases demonstrate that behavioural science is foundational to the entire organization and to all streams of its work.

In closing

The early years of BI as a field were marked by a need to score quick wins and find proof-of-concept for this approach to engineering behavioural change. Now that the field has gained broad acceptance, we expect organizations to start using it to tackle

more complex behavioural and policy challenges going forward. In our view, BI can — and should — play a key role in important societal domains such as the environment, business sustainability, preventive health, and diversity and inclusion. Consider this article as a nudge to get you started.

Bing Feng (MBA 2019), Jima Oyunsuren (MBA 2019), Mykyta Tymko (Rotman Commerce 2019), and Melanie Kim (MBA 2016) were researchers at BEAR.

Dilip Soman is the Rotman School’s Canada Research Chair in Behavioural Sciences and Economics. Soman is the Founding Director, and Feng and Kim previously served as Associate Directors.

Editor’s Note: The BEAR Playbook has contributed to the thinking on behaviourally based efforts at various organizations, including the OECD, in their work on the application of behavioural insights to public policy. The Western Cape Government in South Africa and the Impact and Innovation Unit in the Government of Canada aim to use its content to strategically embed BI into their work.

BI as Designer BI as Auditor

BI as Problem Solver

BI as Chief Strategist

FIGURE 2

Develop a deeper understanding of behavior and decision-making with Behavioral Scientist magazine.

The latest insights from leading behavioral scientists delivered to your inbox each week.

Sign up for the weekly email edition today: behavioralscientist.org/ subscribe

The Nobel Memorial Prize winner, co-author of Nudge , University of Chicago professor reflects on the field after the publication of Nudge: The Final Edition.

Thought Leader Interview: Richard Thaler

Interview by Dilip Soman

Many years since the publication of Nudge, there is still some debate about how to define a nudge. How would you define it? The way we define it is that a nudge is some feature of the environment that alters the behaviour of Humans but not Econs, which means it does so without significantly changing incentives (since Econs care about prices or imposing limiting ‘choices’. According to that very narrow definition, incentives don’t matter but improved messaging could. But I don’t think we should get too hung-up on the definition of a nudge.

That title, by the way, was not our original idea. It was suggested by one of the many publishers who declined to publish the book. He thought Nudge might be a good title, which it was, but its widespread usage has become a mixed blessing. It did help sell books, but I think of the book as a book about choice architecture. I think that the real answer to your question is, everything is about choice architecture.

What I worry about is that people think that nudging is just about tweaks, adding a word, subtracting a word, changing the colour of the package or billboard. It should be much more that; it can be a complete restructuring of the choice environment.

One of the examples that caused much confusion from the original Nudge was the example about opt-in versus opt-out organ donation. The newer version of the book goes into more depth on the topic.

The most frustrating thing after the publication of Nudge was people were convinced that we were advocates for ‘presumed consent’, the policy that makes agreeing to be a donor the default. It’s true that when we decided to write this book, we had seen the data from Eric Johnson and Dan Goldstein’s Science paper from 2003. But when we started reading the larger literature, we ended up advocating something else that we called ‘prompted choice’, which is an opt-in with a nudge. In other words, people have to volunteer to be a donor, but they are asked periodically until they do agree.

After the original edition came out, every time a country switched from opt-in to opt-out, I would get notes like ‘Hey, way to go, Thaler. Another country has done what you have wanted’, and I would start tearing my hair out. We took a very deep dive on this topic in Nudge: The Final Edition and that chapter is completely rewritten. Here is my take on it. If you look at

Sometimes the results of field interventions do not replicate because the context might be different.

that Science article, the fact that almost no one opts out should tell you something important. It probably should tell you that people are not paying very much attention to it. There are lots of situations where people override the default option. How much weight should we as a society place on the fact that 98 per cent of people did nothing? Should that be enough to infer the donor’s actual wishes?

Of course, the real question is, ‘Which system works better?’ That turns out to be a difficult empirical question. My conclusion is that an opt-in system, like we have in the U.S. and Canada, gets roughly 25 per cent more organs donated. Part of the problem is that in many studies, Spain is misclassified as being an opt-out country, which is not the case. That is important because Spain is an outlier. They are a world leader in providing organs for transplants, but they achieve that because they have a very carefully thought-out and effective organ harvesting choice architecture. As soon as a patient looks like they might be a potential donor, wheels start spinning. It is true that Spain once passed a law adopting presumed consent, but a year later reversed it. They are still classified in most analyses as being in the ‘presumed consent’ category. If you take them out, the results change. I think morally, and in terms of getting the most lives saved, my answer to the Do Defaults Save Lives question is — yes, but you should make the default that you have to opt-in [Editor’s note: See an article on using behavioural science to improve organ donation consent rates using active choice on page 50 of this issue].

What do you say to those who argue that the death of behavioural economics is imminent?

I wrote a paper 20 years ago that was called The Death of Behavioural Finance. I predicted that it would die because finance would become as behavioural as it needs to be. I think that is what is happening in economics. When I came to the University of Chicago, I was an outlier. Now we have a weekly behavioural economics workshop (different from the long-standing behavioural science seminar) that attracts 40-50 people each meeting. More important, there are lots of people who have behavioural economics as part of their toolkit, just like there are people who run experiments, or do econometrics or economic theory.

Should we be worried about the replication crisis?

There is no replication crisis in behavioural economics. Now, there is a big replication crisis in other parts of psychology. The basic principles of behavioural economics — that people have self-control problems, care about fairness, exhibit loss aversion, are overconfident, and so forth seem to hold true. I know this because every time I teach managerial decision making, I start by doing a survey of the students on all the classics. It is always the same. Always, 90 per cent of the students think that they will be above the median. Always, the most popular decile is the second one because of modesty.

Here is a good distinction between a concept and its operationalization to keep in mind. Loss aversion, for sure, exists. Now, does loss framing always have a significant effect? No. If we go and replicate the mugs experiment (demonstrating the endowment effect), it will work. But somebody who does an experiment trying to describe something either as a loss or a foregone gain — that’s somewhere between loss aversion and priming, and priming is not a robust phenomenon.

In Misbehaving, I wrote about ‘Supposedly Irrelevant Factors’ (SIFs). Sometimes the results of field interventions do not replicate because there might be an SIF in one case but not the other; the context might be different. That is not a true replication failure but is something we should be careful about when we do interventions in field settings [Editor’s note: See article entitled The Elements of Context on page 84 of this issue].

Would you say that the results of behavioural economics generalize across cultures and nations? Would you say that there are more similarities or more differences in people across the world?

There is a lot of research on cross-cultural differences. I do not want to minimize or disregard it, but if you ask the question the way you did — which I think is a good way to ask it — ‘do I view people as more similar or different?’, then I do have a clear opinion. I definitely believe that people are more similar than they are different. For example, there was a series of experiments on the ultimatum game [a famous experimental game played in pairs; one player gets to allocate a sum of money between themselves

and a second player, and the second player can accept the allocation, or reject it such that no one gets anything] in various countries. The results do vary somewhat from one country to another, but the basic finding is robust: If you offer less than 20 per cent of the pie anywhere in the world, you have a very good chance of getting turned down.

Maybe you have to offer 30 per cent in this country and 10 per cent in another, but nobody likes to be treated unfairly. Now, there are customs and norms about what people think of as fair that will vary around the world, and that’s going to matter a lot. But again, we see more similarity than difference. There’s an old fairness study that Daniel Kahneman, Jack Knetsch and I did that has been replicated all around the world.

Sometimes when organizations are trying to implement choice architecture to ‘nudge for good’, there might be some internal conflicts due to misaligned incentives. For instance, a financial institution that is trying to design tools to help clients avoid taking on too much debt might face pushback from other groups more interested in profit. Any advice on how to deal with that using choice architecture, or any other approach?

I think this is a big question — how can organizations do good and yet be financially successful? I actually spent some time with the board of a large Australian bank, Commonwealth Bank, which I believe is a partner of BEAR. They are trying to be what I call the ‘good bank’ [Editor’s note: See a case study on Commonwealth Bank of Australia on page 92 of this issue]. There is an open question as to whether a good bank can compete successfully against an evil bank. The reason why this is a question is that the good bank might look more expensive.

If the evil bank says, ‘Here’s a free credit card and a free checking account’, but there are a lot of sneaky fees, it being inexpensive will be a bit of an illusion. It is like some ‘discount’ airlines that may have a low-ticket price but if you have to check a suitcase, good luck. The only way to succeed as the good bank is via trust. You have to be selling trust and you have to earn it.

Here is an old story. My wife, France, and I were in Thailand and needed a cab ride to a restaurant for dinner. The tradition

there is that fees are negotiated. As a good behavioural scientist, I tried to use all the techniques I had learned and managed to get the fare down to something like $3 for a 20-minute ride. When we arrived at our destination, the driver said, “Would you like me to take you back?” I said, “Sure. What would the fare be?” continuing my tough negotiating posture. He said, “Same fare.” Now, we have a dilemma of, ‘Will he come back and get me?’ and ‘Will I be there if he does?’ He proposed a contract to solve the problem that no economist would ever suggest: He said, “Don’t pay me anything now. I will be here waiting for you.” Isn’t that good? He is telling me that he trusts me and I should trust him.

When we got done with dinner, of course, we went to look for him. Then we said, “On the way back, we want to stop at this night market.” He said, “Fine. I’ll wait for you. Again, don’t pay me anything now.” When we were done there, we walked 15, 20 minutes out of our way to go find him because we now owed him $9. I think there is a deep lesson there that if you want to be the good bank business (or any other type of ‘good’ business), the only way you can succeed is by being the place that people trust.

On the one hand, your research on mental accounting says that people are more likely to splurge impulsively on luxury purchases when they receive an unexpected windfall. On the other hand, you seem to advocate for policy that withholds more money from the taxpayer to ensure larger tax refunds at the end of the year, as I understand, because the refund feels like a windfall and will lead to increased savings. How do we reconcile this?

The bigger the windfall, the larger proportion will be saved, though some of it will get spent possibly on luxuries, possibly on durables. You can see this among academics, at least in the U.S., where there is this odd tradition that you get a nine-month contract and then additional ‘summer money’ on top of that. In most places, the ‘summer money’ takes the form of two months of salary. People live on their nine-month salary because they have to pay the rent every month and then they get this lump sum. You see both splurging and durable investments.

A study that I think would be very interesting to do is to compare people who get paid once a month with people who get paid

Remember to nudge for good!

every two weeks. The reason is if you get paid every two weeks, there are two months out of the year that you get three paychecks. Those will feel like little windfalls. If you are affluent enough, it won’t matter, but if you’re constantly checking whether you have enough money to buy shoes for the kids, then I’m guessing that behaviour would be different in those two situations.

You have said, both in the book and elsewhere, that nudges are usually not enough for solving complex problems and that mandates and stronger shoves are sometimes necessary. This calls for a judicious mix of nudge and shove strategies which might not be easy for a practitioner to perfect. What advice can you offer, from a behavioural perspective, on how the practitioner can choose between nudging and stronger options?

When we wrote the book originally, the idea was to show that there are some things we can accomplish even if we tie one hand behind our back and don’t force anybody to do anything or even bribe them to do it. That was like changing defaults and little nudges. When it comes to climate change, we are not going to get there if we rely only on nudges. Like every economist in the world, I am in favour of a carbon tax or a cap-and-trade program because we’re not going to reduce emissions as long as they’re free.

Our chapter on climate change starts with that. First, get the prices right. Now, there are nudges that can help, and you get effect sizes like two or three per cent, which sounds small and cannot solve the problem alone but is nothing to sneeze at. We quote our former University of Chicago colleague and former president, Barack Obama, who liked to say around the White House, “Better is good.”

You have often said by way of advice to researchers; “Make your research about the world, not about the literature.” What do you mean by that?

What I mean by that is people should get ideas by looking at the world rather than by only reading journal articles. I think that there is a trap that graduate students and young faculty members start reading papers and they end up writing the 25th paper on

some topic because there’s some gap in the literature that no one has filled. I have rarely written more than two papers on any topic because I have a short attention span.

I’d rather write the first paper on mental accounting than the hundredth. People were doing mental accounting before I gave it a name, just like they were nudging before we gave it a name. Likewise, there have got to be a number of phenomena, of behaviours and marketplace observations that no one has written about and studied. Study those.

Any final thoughts?

Thanks to all my friends in Canada and around the world. As I always write when I’m asked to sign any of my books, remember to nudge for good!

Richard H. Thaler is the Charles R. Walgreen Distinguished Service Professor of Behavioral Science and Economics at the University of Chicago Booth School of Business. He is the 2017 recipient of the Nobel Memorial Prize in Economic Sciences, the co-author of Nudge (Penguin Random House 2008, 2021) and the author of Misbehaving: The Making of Behavioral Economics (WW Norton, 2016). He is widely considered as the founder of behavioural economics and applied behavioural science.

A multi-day program for talent leaders

Inclusion by Design

Learn how to use behavioural insights and business design to advance your diversity, equity, and inclusion goals.

CHANGING BEHAVIOUR FOR GOOD

Creating new habits might just be the most promising avenue for making behaviour change stick.

by Angela Duckworth and Katherine Milkman

WHILE WE SELDOM THINK ABOUT IT, our life outcomes are powerfully determined by seemingly trivial, repeated acts. Our health, for example, depends on thousands of daily choices — to eat well and exercise regularly, to avoid smoking, and to take medications as prescribed. And yet, 40 per cent of premature deaths each year result from suboptimal behaviour in this domain: Tobacco is responsible for 435,000 of those deaths, while poor diet and physical inactivity account for 400,000. Cardiovascular disease — the leading cause of mortality — is largely treatable with anti-hypertensive medicines; but just one year after receiving a prescription, only about half of patients are still taking their medication as directed.

Our bad habits run the gamit from health to personal finances. One in three American families has no savings at all, and 52 per cent are under-saving, even though most would need to save just 15 per cent of their earnings to prepare for the future. Academic success requires an array of good habits

at any age: attending classes, studying and engaging with challenge on countless occasions. Yet tellingly, 23 per cent of high schoolers and 49 per cent of college students drop out before earning diplomas. Sadly, all of these challenges to life outcomes disproportionately harm disadvantaged members of society.

In recent years, behavioural scientists have learned a great deal about the underlying situational and psychological factors that determine our daily decisions, leading to many successful and scalable interventions to change short-term behaviour. The problem is this: behaviour change rarely endures. In this article, we will review a small but growing body of research suggesting the most promising approaches to changing behaviour — for good.

The Power of Habits

Perhaps the most promising avenue for making behaviour change ‘stick’ is by changing our habits. Habits are automatic,

Adding a desired behaviour onto the end of a routine that is already habitual is an effective way to create a lasting habit.

effortless and repeated actions that develop when the following cycle is repeated multiple times: A situational cue triggers a behaviour, and that behaviour triggers a reward. Exactly how many repetitions of this cycle are required to generate a habit remains an open question, and it is likely context-dependent.

Past research suggests that intervening to creating sustained behaviour change in the form of new habits requires the deployment of two kinds of strategies:

Targeting the situation. Strategies targeting the situation insert behaviour-triggering cues that are both interesting and obvious, making beneficial behaviours easier and more rewarding.

Shifting peoples’ cognition. The second type of strategy shifts cognitions, equipping people, for example, with ‘beneficial beliefs’. This allows them to forecast the consequences of their behaviours more accurately and act more adeptly.

Importantly, changing cognitions without changing the situation puts an inordinate burden on individual willpower — and as a result, there should be synergy in applying these approaches in combination. We will now review some of the most promising research-based strategies for changing behaviour for good.

1. TEACH PEOPLE TO CUE THEMSELVES. Ensuring that cues are established to reliably trigger desired behaviours is key to creating habits. One way to ensure cues are present when they are needed is by teaching people to cue themselves. When people form ‘ifthen’ plans about the behaviours they intend to engage in following a given cue, this robustly increases follow-through. An example of an if-then plan is: ‘If it is a weekday and I am about to leave the house for work, then I will make my own coffee instead of going to Starbucks’. If-then plans reliably promote followthrough on one-time behaviours, but they can also promote

sustained behaviour change. One study found that teaching students goal setting and planning skills improves attendance and grades in the following marking period. Together, this research suggests that to facilitate durable behaviour change, it is helpful to coach people to make if-then plans that put cues in place to trigger desirable behaviours.

2. PIGGYBACK CUES. Another way to ensure cues to trigger desired behaviours are reliably present is by ‘piggybacking’ desired behaviours onto existing routines. For instance, adding a desired behaviour (e.g., flossing, eating an apple a day) onto the end of a routine that is already habitual (e.g., brushing your teeth, having a cup of coffee) can be an effective way to create lasting habits. In one study, flossing habits were more effectively generated by encouraging people to floss after brushing their teeth, rather than vice versa.

3. CHANGE BELIEFS ABOUT BENEFICIAL BEHAVIOURS. People will do things that they believe to be valuable, and accordingly, one of the most powerful cognitive interventions available is to change beliefs about the likelihood of success. For example, teaching students that their abilities — including their intelligence — can improve with effort and experience has been shown to improve report card grades and course completion among at-risk groups. And showing students that ‘deliberate practice is difficult, but it is both doable and effective’ changes how they interpret the necessary frustration of attempting skills they have not yet mastered — particularly among low-achieving students.

Beliefs about ‘norms’ — the attitudes and behaviours of other people — also powerfully influence behaviour. For example, in one study, learning that many people were reducing their consumption of meat prompted more cafeteria patrons to order meatless meals.

The Keys to Changing Behaviour for Good

4. MAKE BEHAVIOUR CHANGE EASY. Making it as easy as possible to sign up for valuable programs that facilitate behaviour change dramatically improves outcomes. For example, letting people enrol in a retirement savings program (so that a portion of every future paycheque is automatically directed to a retirement account) via a stamped postcard increased participation by 20 percentage points, and allowing sign-ups after a future pay raise produced a 78 per cent sign-up rate — boosting enrollees’ savings by 388 per cent over 40 months.

In other studies, providing high school seniors’ parents with help completing the Free Application for Federal Student Aid (FAFSA) while they receive assistance with tax preparation increased the rate at which those parents’ children completed two years of college by eight percentage points over a three-year follow-up period; and providing community college freshmen with reminders and encouragement to renew their FAFSA increased sophomore persistence by 14 percentage points.

5. MAKE GOOD BEHAVIOUR MORE ENJOYABLE. Research suggests that finding ways to make beneficial — but often unpleasant — behaviours (e.g., exercise, studying) more immediately enjoyable has the potential to promote sustained behaviour change. For instance, one study found that when people were only allowed to enjoy tempting audio novels while exercising at the gym, they visited the gym more frequently than a control group over seven weeks. This so-called ‘temptation bundling’ strategy helped make the act of exercising more fun by pairing it with an engaging audiobook.

More generally, combining good behaviours that can be unpleasant with enjoyable activities (e.g., scheduling get-togethers with a challenging relative at a favourite restaurant, doing household chores while listening to a favourite podcast) can promote behaviour change for good. Complementary research has shown

that persistence is increased on challenging-but-important goals by encouraging people to pursue those goals in fun ways. For instance, encouraging gym goers to choose a workout that is fun (e.g., a dance class) rather than pursuing the workout that is most effective promotes more persistent exercise. Similarly, playing music in a high school classroom to make studying more fun increased persistence on schoolwork. Overall, making good behaviour more enjoyable facilitates the association of a ‘reward’ with the behaviour, and such repeated rewards are key to habit formation.

6. REPEATEDLY REWARD GOOD BEHAVIOUR. Perhaps the most promising stream of research designed to promote habit formation has shown that repeatedly paying people to engage in a valuable behaviour or otherwise encouraging it (e.g., by conveying its popularity) for as little as a month can increase the target activity for many months post-intervention. In one study, paying students to visit the gym eight times over the course of a month rather than just once or not at all produced behaviour change that lasted long after that month (and the intervention period) ended. The students who had been rewarded for repeatedly visiting the gym worked out nine times, on average, in the following seven weeks, while other students went roughly half as often.

Several follow-up studies have since replicated the finding that rewarding repeated exercise over a period as short as one month can lead to sustained habits detectable up to a year later. Incentives aren’t the only reward that has yielded this result: Repeatedly alerting people to how their energy usage compares to that of their neighbours in comparable homes (‘social norms marketing’) also has sustained benefits after messaging is discontinued. Monthly energy reports with social norms information sent to residential homes for two years continued to reduce energy consumption for years after

Situational Interventions

CUE

CUE BEHAVIOUR

discontinuation, with only a 10–20 per cent attenuation of their benefits per year.

Finally, similar results have been generated by offering ‘regret lotteries’ — lotteries requiring measurable behaviour change by a given date to win. Regret lotteries have been effective at encouraging sustained weight loss and smoking cessation.

7. INTERVENE AT OPPORTUNE TIMES. Finally, the timing of an intervention matters. Research on the ‘fresh start effect’ shows that people are more prone to engage in beneficial behaviours like exercising, searching for information about dieting on Google, and creating a commitment contract on the website stickK.com at the start of cycles like the beginning of a new week, month, year or following birthdays and holidays.

How to Induce Future-Focused Thinking

Present-bias, or the tendency to over-weigh immediate gratification while under-weighing the long-term implications of a choice, is responsible for many errors in judgment. Specifically, presentbiased thinking has been blamed for societal problems ranging from obesity to under-saving for retirement. Following are a series of nudges designed to promote future-focused thinking in order to reduce the pernicious effects of near-sightedness.

CHOOSE IN ADVANCE. One means of reducing biases related to near-sightedness is prompting individuals to make decisions well in advance of the moment when those decisions will take effect. This strategy is impactful because of people’s tendency to make less impulsive, more reasoned decisions when contemplating the future than the present. Choosing in advance has been shown to increase support for wise policies requiring sacrifices, to increase charitable giving, to contribute to increases in retirement savings, and to increase the healthfulness of consumers’ food choices.

Another result of choosing in advance is that choices are made in a higher construal level mindset, which means they focus more on abstract (e.g., why?) rather than concrete objectives (e.g., how?). This has other by-products, however — for example, higher construal leads to greater stereotyping. Therefore, an important caveat to choosing in advance is that it may lead to greater discrimination against women and minorities, as demonstrated in a field study of decisions on whether to grant prospective graduate students requests for meetings.

In closing

When our daily, routine choices are sub-optimal, they can undermine just about every conceivable life outcome. And when poor choices accumulate, they particularly harm populations with fewer resources and less ‘slack’ — for whom seemingly trivial mistakes can spiral with remarkable speed (e.g., a small ailment goes untreated and becomes debilitating; the interest on small, unpaid debts accumulates to produce financial insolvency; a failing grade makes it difficult to justify studying for a degree over working full-time).

Sadly, adversity itself exacts a hefty physiological and psychological toll: When individuals perceive that the future is uncertain and threatening, they become biased towards meeting their short-term needs versus working towards long-term goals.

INCLUDE A PRE-COMMITMENT.

Because people tend to make more patient and reasoned decisions about the future, providing opportunities for individuals to both choose in advance and make a binding decision (or at least decisions where penalties will accompany reversals) can improve many choices. Research has shown many benefits from pre-commitment. For example, substantial savings increases result from providing individuals with access to bank accounts with commitment features such as a user-defined savings goal (or date) such that money cannot be withdrawn before the pre-set goal (or date) is reached. Although only 28 per cent of those offered such commitment accounts selected them when equivalent interest was available on an unconstrained account, average savings balances increased by 81 percentage points for those customers of a Philippine bank with access to commitment accounts. Recent research has also shown that pre-commitment can help people quit smoking, exercise, achieve workplace goals and resist repeated temptation in the laboratory. Pre-commitment is particularly valuable in settings where self-control problems pit our long-term interests against our short-term desires. When it comes to food, for example, pre-committing to smaller plates and glasses can reduce consumption.

BUNDLE TEMPTATIONS.

A new twist on pre-commitment called ‘temptation bundling’ actually solves two self-control problems at once. Temptation bundling devices allow people to pre-commit to coupling instantly gratifying activities

As a result, solutions to enduring behaviour change carry universal benefits.

At present, organizations and academics are incentivized to work on behaviour change in isolation, focusing on a single setting, measuring success over the short-term. But there is an enormous untapped opportunity for large-scale, interdisciplinary work combining practical and theoretical insights to enable sustained improvements in daily decisions on a collective level.

As Aristotle said, “We are what we repeatedly do.” If our ultimate destinies derive from our daily habits, then the 21st century may be the first in which humanity learns how to change behaviour for good.

Angela Duckworth is the Founder and CEO of the Character Lab, a non-profit whose mission is to advance the science and practice of character development. She is also the Christopher H. Browne Distinguished Professor of Psychology at the University of Pennsylvania, faculty co-director of the Penn-Wharton Behaviour Change For Good Initiative, and faculty co-director of Wharton People Analytics. Katherine Milkman is a Professor at the Wharton School of the University of Pennsylvania and holds the Evan C. Thompson Endowed Term Chair for Excellence in Teaching. She has a secondary appointment at U Penn’s Perelman School of Medicine.

This article is based on a paper made possible by support from the Chan Zuckerberg Initiative, the National Institute of Health, the Robert Wood Johnson Foundation and the John Templeton Foundation.

(e.g., watching low-brow reality television, receiving a pedicure, eating an indulgent meal) with engagement in a behaviour that provides long-term benefits but requires the exertion of willpower (e.g., exercising, reviewing a paper, spending time with a difficult relative). Such pre-commitment devices can increase engagement in beneficial behaviours like exercise while reducing engagement in guilt-inducing, indulgent behaviours.

ORGANIZATIONAL COGNITIVE REPAIRS. De-biasing can also be embedded in an organization’s routines and culture. Researchers call these de-biasing organizational artifacts ‘cognitive repairs’. A repair could be as simple as an oft-repeated proverb that serves as a continual reminder, such as the phrase ‘don’t confuse brains with a bull market’, which cautions investors and managers to consider the base rate of success in the market before drawing conclusions about an individual investor’s skill. Other examples include institutionalizing routines in which senior managers recount stories about extreme failures (to correct for the underestimation of rare events) and presenting new ideas and plans to colleagues trained to criticize and poke holes (to overcome confirmatory biases and generate alternatives).

Many successful repairs are social, taking advantage of word-of-mouth, social influence and effective group processes that encourage and capitalize upon diverse perspectives. Although cognitive repairs may originate as a top-down intervention, many arise organically as successful practices are noticed,

adopted and propagated.

One cognitive repair that has not only improved many organizational decisions, but saved lives, is the checklist. This tool could easily fit in many of our de-biasing categories. Like linear models, checklists are a potent tool for streamlining processes and thus reducing errors. A checklist provides a list of action items or criteria arranged in a systematic manner, allowing the user to record the presence or absence of the individual item listed to ensure that all are considered or completed.

Checklists, by design, reduce errors due to forgetfulness and other memory distortions (e.g., over-reliance on the availability heuristic). Some checklists are so simple that they masquerade as proverbs (e.g., emergency room physicians who follow ABC — first establish a irway, then b reathing, then c irculation). External checklists are particularly valuable in settings where best practices are likely to be overlooked due to extreme complexity or under conditions of high stress or fatigue, making them an important tool for overcoming low decision readiness.

— From “A User’s Guide to Debiasing” by Jack B. Soll, Katherine Milkman and John W. Payne in The Wiley-Blackwell Handbook of Judgment and Decision Making.

Behavioural Science on the Ground: A Global South Perspective

WHILE THE USE of behavioural science in helping organizations tackle applied problems, like the design and delivery of products, programs and services, was initially largely confined to the Western world, we are now seeing an increased use of behavioural science in the Global South. On the face of it, the infusion of funding into behavioural science research seems like a beacon of hope, promising to nurture a thriving ecosystem of researchers and practitioners looking to solve behavioural challenges in the Global South. A closer look, drawn from the experiences of behavioural science researchers in Africa, reveals a narrative far more nuanced and complex. It is a story of disparities in the distribution of benefits, a tale of local researchers and research institutions bearing the hidden costs (both direct and indirect) of this burgeoning field. Busara is a vocal boutique advisory that unlocks impact for organizations working in the Global South by helping them apply cutting-edge behavioural science specifically for the people and the contexts they serve. Drawing from the experiences of Busara, we examine the true price of progress in the field.

Navigating the Predictably Unpredictable Financial Costs

Unusual circumstances arising during a study can affect the time and financial costs of research implementation. Jennifer Adhiambo (Senior Manager, Busara) once ran a study that required participants complete surveys before and after their payday. How often could paydays exist? Once or twice a month? Jennifer encountered eight different types of pay-

day patterns in a pool of women participants. The schedule of the study had to be overhauled and the study, which was budgeted to be completed in a week, took a month. This was accompanied by an increase in costs.

Behavioural science projects are typically funded on the basis of predefined research goals, budget norms and implementation methods. In Africa, this approach to funding often lacks consideration for local realities. Funders, if they fail to consult experienced local professionals, may also hold misconceptions about the local context and fail to estimate project timelines and budget needs appropriately. The commonly held notion that research costs will be low, in keeping with African cost-of-living standards, can lead to wariness and suspicion in the funders’ mind and a refusal to accept practical advice about actual research costs.

In addition to Jennifer’s story, it is crucial to understand the dynamic cultural landscape in which research activities unfold. In contrast to their Western counterparts, many participants in the Global South often lack prior experience with experiments or surveys, necessitating extensive pre-experiment coaching and guidance in using technical tools. Issues like lack of punctuality and absenteeism are prevalent, influenced by factors such as transportation difficulties, opportunities for higher wages elsewhere, and weather conditions. Delays caused by latecomers can lead to demands for additional compensation from those who arrived on time. Furthermore, the burden of domestic responsibilities on women and gender disparities in some

One suggestion is to invest resources to truly reverse the top-down nature of the planning, design and control of most research programs.

regions can lead to reduced female participation, thereby skewing the research sample.

Although these incidents are predictably unpredictable, the answer to the question of who should bear the higher costs arising in such situations is unclear. Many times, the research agency bears the costs, a burden hard to carry given that research projects are extremely cost-competitive.

Many research agencies, having firsthand experience with the burdensome costs resulting from impractical budget constraints, approach research funding cautiously when dealing with those who hold unrealistic expectations that research costs will be cheap in African countries. When these funders rely on notions of what costs should be, they often resort to working with local organizations who provide data on their terms. This can lead to the production of unreliable data through oversampling of easily accessible urban populations, rather than sampling from the more challengingto-reach and isolated communities that should be the true focus of the research.

The effects of such pseudo-data gathering add up in the medium- to long-term and it is the communities most in need of development-directed research which tend to suffer. Nathanial Peterson (Vice President, Busara) quotes the example of northern Nigeria whose population of approximately 100 million is about twice that of Kenya. Northern Nigeria is highly understudied and very much in need of research that could improve development efforts. Yet Nairobi, given its proximate location and other factors which make it an easy place to sample, has a sampling rate almost 100 times greater than Northern Nigeria.

Clearly then, taking the reality of African life into consideration in the design of research programs is necessary to obviate such adverse and unwanted outcomes.

Much Needed Flexibility to Fund the Future

Funding for research is often insufficient and primarily allocated to direct costs that include operational aspects like data collection. Consequently, researchers and organizations are left to shoulder the substantial costs involved in literature surveys, publishing, and presenting results from their own salaries. The challenge of building capacity and making internal improvements becomes even more pronounced when budgetary inflexibility prevails.

Research funding programs, especially those originating in Western countries for the broader Global South re-

gion, often prioritize capacity building; however, achieving this goal is difficult. Top institutions in the U.S. and Europe enjoy substantial overhead allowances, along with significant endowments, adept project management, administrative expertise, long-standing leadership, and a wealth of institutional knowledge amassed over decades, if not centuries. This starkly contrasts with research centres in the Global South, struggling to provide such resources within their limited overhead allocations, which typically fall around 15 per cent.

Moreover, often absent from project budgets is an allocation of funds toward the improvement of the research process itself. As the move toward greater localization gathers momentum, there is a growing realization of the possible danger that the changes being implemented may actually represent tokenistic rather than effective localization. Effective localization enhances the effectiveness of development efforts by shifting power, resources and ownership from international actors to local communities. Simply employing more local talent may not always be enough, says Megan Grazier (Director, Busara), who, along with others at Busara, is involved in meta-research, or research about research, with the aim of improving research processes. The idea, she says, is “to internally unpack some of the challenges to make sure the research processes are contextually appropriate to all the stages.” This proactive approach to applying behavioural science internally often falls on the organizations themselves.

This leads to a fundamental question: How can emerging research institutions in other parts of the world reasonably aspire to develop similar capacity and quality without the flexibility to allocate financial resources and establish institutional infrastructure as needed?

Budgeting For Intellectual Aspirations

Universities are supposed to both impart knowledge and generate new knowledge through research. Francis Meyo (Vice President, Busara) takes the example of an institution running courses relating to behavioural science. Such centres are thinly spread across African countries and typically have faculty-to-student ratios that are much higher vis-a-vis the West. In addition, academics in the Global North (unlike those in the Global South) might also have access to support like research and teaching assistants, graduate students and administrative assistants. This contributes to a growing

disparity in what can be accomplished with the resources at hand for academics and trainers in African countries. So when program funders, including major institutional funders, have a policy of routing their (often substantial and long-term) funding only via local academic institutions, such funds often absorb the complete resources of these African centres, leaving little opportunity for them to initiate any research with local focus and relevance.

Most of the available resources within research infrastructure get tied into externally funded research programs, leaving academics, non-governmental organizations (NGOs) and research agencies little time or opportunity to dedicate to their own focus areas or pressing issues that require further investigation. In any case, funding for such research is low compared to funding for specific programs. Program funding is usually directed towards the needs of the funding agency. Well-intentioned though such programs may be, they do absorb the bulk of available resources, leaving unattended and shortchanged those sectors more urgently in need of short- to medium-term attention, such as research that would help to improve or modify development policy.

Additionally, local researchers often face significant personal costs in their pursuit of intellectual opportunities. Local universities frequently cannot afford to prioritize these endeavors, forcing researchers to navigate complex political landscapes to gain the freedom to work on projects of intellectual interest. Additionally, when they are fortunate enough to participate in such projects, they often find themselves in junior or subordinate roles alongside prominent Western academics and sometimes multiple Western colleagues. Consequently, local researchers must ‘pay their dues,’ often enduring tokenization in the process, to establish credibility and build a noteworthy curriculum vitae (CV). Only after this uphill battle can they secure their own grants or attain the privilege of contributing substantively to shared research agendas. This stark disparity highlights the broader challenges faced by local researchers in balancing their career aspirations with the demands and expectations of the field.

The Missing Piece of the Puzzle

Despite these realities, it is clear that every stakeholder is doing their part. Little more can be asked from a funder, from a local researcher, from a Western researcher, from a research institution, and from implementation agencies. There isn’t a silver bullet that can realize the full potential of local research resources. One suggestion is to invest resources to truly reverse the top-down nature of the planning,

design, and control of most research programs in order to give greater voice to local researchers, research organizations and communities, to introduce an element of research at the concept stage to study the cultural terrain in which the program will proceed, and to design and build the structure of the program using the knowledge and insights gained from such initial studies.

A more longer-term thought is to reimagine the current market. Foundations are not fickle; many see their mission as putting ideas into the world rather than sustaining or scaling the ideas or institutions. Academics see their mission as generating new knowledge and specialization. Implementation agencies see their mission as providing services to the last mile. Research organizations see their mission as some combination of these. How can we combine the benefits of different models and harness the overlap in roles while each stakeholder still plays their part?

Catalyzing transformative change in the research landscape of the Global South requires a holistic reimagining of the entire system. Drawing inspiration from the proven success of cooperative ownership models, we can address the pressing challenges of misaligned funding and localized real-world intricacies, ensuring that research budgets are set with true ground knowledge. By embracing the structure of public-private partnerships, we can effectively counteract the pitfalls of pseudo-data gathering, harnessing the resources and expertise of both sectors to reach even the most remote communities. The stability offered by endowment models, time-tested in prestigious universities, can provide the much-needed financial safety net, guaranteeing funding for both direct and indirect research costs. Merging this with the collaborative spirit of open-source models can combat token localization, offering a platform where local researchers can refine and share globally proposed methodologies to better fit local contexts. Lastly, by infusing the social enterprise model’s principle of mutual benefit, we can ensure that for every externally funded project, a local research initiative also gets the spotlight. Together, these strategies create an invigorating model that promises a future where research in the Global South is not only adequately funded but is also profoundly respectful of its local realities, ensuring that it remains deeply impactful and collaborative at its core.

Anisha Singh was Vice President and Chaning Jang is the CEO and Director of the Busara Center for Behavioral Economics in Nairobi.

Behavioural Science at Scale

WE LIVE IN

of unprecedented scale. Big data, trilliondollar brands, and astonishingly fluent artificial intelligence (AI) trained on massive troves of language. For those of us at ideas42 and elsewhere dedicated to making social change, how can we use this vast field to make the world more equitable, healthy and happy? ideas42 is a non-profit that uses insights from human behaviour — why people do what they do — to help improve lives, build better systems, and drive social change.

On the one hand, scale offers an opportunity to achieve our goals to a depth that would not otherwise be possible. Good ideas catch on faster: we can learn about other best practices and share our own breakthroughs instantaneously with colleagues many time zones away. And we can use existing infrastructure to bring resources to those who need them most, even in remote corners of the world, so that no one gets left behind.

On the other hand, scaling adds an additional set of challenges to our work, a layer that often requires quite different skills than does making change at a local level. On paper, it seems easy to copy and paste social programs from one country to another, or even one zip code to another. But as we at ideas42 have learned through more than a decade of disseminating and applying behavioural science interventions, change is rarely that straightforward.

Behavioural science can help us solve these challenges of scale. By methodically exploring the context, solving unintuitive barriers, and accounting for the weirdness of human behaviour, it is a powerful lens for bringing wise

designs to more people. Although there is no universal formula for how to do this, we have identified common principles for scaling up interventions across different organizations and countries.

Scaling Isn’t Automatic

In an easier world, good ideas would, by their merits, catch fire and proliferate on their own. In a more intuitive world, where laws of human behaviour were as predictable and consistent as laws of Newtonian physics, social innovation that worked in one time and place would replicate in another. In the real world, we have found that solutions that work in one context will often not work as intended if we drop them wholesale into a new context.

We say ‘often’ because it’s worth looking for opportunities where scaling can happen smoothly. This can happen, as we found in one of our programs, when we build on national-level infrastructure. This project, an excellent example of collaboration between academic and applied organizations, studied why more people in Mexico aren’t saving for retirement and then piloted ways to encourage more savings. Along with our colleagues, including Avni Shah, Matthew Osborne and Dilip Soman at Behavioural Economics in Action at Rotman (BEAR), we brought designs for increasing retirement savings to more than 21 million Mexicans.

But scaling isn’t always so fortuitous. In that Mexico project, we were lucky to work with the central government to implement these designs, and we were scaling within a

AN ERA

Piyush Tantia, Alissa Fishbane, Kate MacLeod and Cameron French

The last yard problem is the need to adjust many details of an intervention in order for it to work as intended.

single administrative framework. That’s an uncommon path. More often, the fundamental behavioural insight (say, comparing ourselves to peers) may transfer as is, but how it is packaged and delivered does not. We call this the last yard problem — the need to adjust many details of an intervention in order for it to work as intended.

Reducing Unnecessary Warrants

One of the most common reasons Americans are issued arrest warrants today isn’t because of violent crime or drinking while driving; often, it is simply because they missed their court date. Most often, these court dates are for traffic and other low-level, common offenses — offenses for which arrest or jail are disproportionate penalties.

The (Un)warranted initiative at ideas42 has been working to reduce these unnecessary warrants across the U.S. Improving court appearance depends on addressing the real reason people miss court. A misconception about missed court dates is that people do not think their court dates are important and intentionally miss them. Through research in jurisdictions around the country, we have found that people do want to meet their court obligations. Overwhelmingly, people miss court dates because of behavioural and structural reasons — such as confusing court forms, an inability to find necessary childcare during the date, or conflict with work schedules.

To improve court appearance, we have investigated ways that courts are already developing new practices for overcoming behavioural barriers, and we encourage jurisdictions to replicate this success. We now focus on scaling the most impactful and cost-effective of these new systems: effective court date communications. Our work has led to a 36 per cent reduction in missed court dates. We’re now helping other jurisdictions across the country redesign the information people get about their court date, from the form they get at the time of arrest (a citation or jail release form) to reminders (like those for doctor appointments) sent just ahead of the court date.

Unfortunately, scaling these communication designs is made thorny by each jurisdiction having different processes for handling different parts of a case. For example, there is no standard across the U.S. for which agency ‘owns’ a citation. In some places, a state court is responsible for its

design; in others, the county or city designs it. Even then, the form given at the time of arrest could be owned by police, a sheriff, or a court. This has meant that scaling isn’t as simple as working with local courts around the country. Rather each jurisdiction requires specific mapping to understand which additional agencies are involved in court date communications. While the communications themselves can be tailored and implemented quickly and endure automatically as they are built into existing systems, a good part of scaling involves identifying, engaging and coordinating across the different system stakeholders.

As we continue this work, it’s increasingly clear that scaling is often high-touch in unexpected ways. This effort is worthwhile for interventions like court communications, where the social return on investment is so high, yet the path to scale is not as linear as one may expect. To date, (Un)warranted has worked with jurisdictions in seven states and is actively seeking more funding to meet the demand from others and further reduce mass incarceration. Our goal is to scale this work so that in the next five years we help 500,000 people escape unnecessary warrants.

Scaling problems can also hinge on frontline workers. For people to get reminders about their court dates, courts need current contact information, including phone numbers, email addresses, or mailing addresses. Collecting this information, which is rarely listed on a driver’s license, means law enforcement officers must seek additional details at the time of arrest or jail booking. Doing so creates more efficiency in the long run for officers, who will not need to arrest and jail as many people for missed court dates. But in the immediate moment, it is one more step to take. The hassle (of collecting new data) and the change in status quo (from how they’ve handled arrests in the past) are classic behavioural challenges. The generalizable lesson here is that scaling can be particularly difficult when it relies on client-facing staff to act differently.

Scaling Anti-Poverty Programs

There is a deceptively simple tool at the core of the international anti-poverty movement: give money to people who lack critical resources.

Programs like these are called cash transfers, and they differ from historical anti-poverty programs that provided

We must adapt to the local context, but those adaptations are often not the obvious ones.

services or in-kind resources such as food. An overwhelming body of research from the behavioural sciences and other disciplines suggests that people are sophisticated money managers, and they allocate money more effectively to address their needs than if the same amount of cash were spent on what program administrators think people need.

Our earliest work involved government cash transfer programs in sub-Saharan Africa. In Tanzania and Kenya, for example, cash transfer participants had long-term financial goals, such as starting a business, improving their home, or paying their children’s school fees. However, they reported that they struggled to save enough to reach those goals. To solve this problem and make programs even more effective, we developed decision making aids and tools that prompted people to consider their long-term financial goals, make concrete plans for how much they want to save from each transfer, and then follow through on their plans.

Based on early success, we started bringing these programs to other countries in the region but were met with scaling challenges. For instance, when we brought this work to South Sudan, the value of the local currency was fluctuating and had lost as much as 67 per cent of its value in a year. In this context, saving cash was not an effective strategy. We needed new models for what ‘investing in your financial future’ looked like. We deployed the same mental accounting and goal setting concepts that encouraged people to invest in materials to grow business or improve their home, but added prompts like, “What can you purchase now to work toward your goal?” These adjustments helped people allocate money in ways that matched their unique situations.

Our cash transfer work underscores how social and cultural differences call for different designs for the same intervention. For instance, the social norms around gender and financial behaviour differ between Kenya and Tanzania. In Tanzania, a single design of one of our savings tools (a physical pouch in which to allocate cash to immediate needs versus future goals) worked well for both men and women. In Kenya, because of the association in people’s minds between men and financial activities, we needed to design ‘male’ and ‘female’ variants of the pouch. Had we not usertested carefully, we might have wound up with a less effective design in Kenya, one that unintentionally excluded half of the population.

Scaling behavioural designs for cash transfers is not a case of simply making tools available; we need to customize the behavioural principles for each program’s unique context so that the designs actually work as intended. We are now looking to acquire funding to bring this work to even more programs and countries. For instance, we’re currently partnering with the government of Ethiopia on a program that helps people precommit to allocating a portion of their money to business startup activities. And after our early success in Tanzania, we’re continuing to work with the government there with the goal of bringing cash transfer designs to more than half a million people.

Lessons For Practitioners Wanting to Scale

The overarching lesson for scaling is that we must adapt to the local context, but those adaptations are often not the obvious ones. To learn what those adaptations should be, we need to try out designs in the field, with people, and sometimes even run pilot experiments. This often reveals new barriers that are not apparent on the surface.

In this light, scaling might seem overwhelming. But scale does not require starting from scratch in each new context. Instead, the adaptations are most commonly in how a behavioural intervention is packaged and delivered, rather than in the fundamental behavioural insight. Scaling then requires multiple levels of design, from core insight to actually delivering it to real people. Often, if you’ve hit on a good idea, the essence of that idea is worth preserving. The work of scaling is finding out how to help it shine in a new light.

Piyush Tantia is the Chief Innovation Officer at ideas42. Alissa Fishbane is a Managing Director leading the (Un) warranted initiative and the Safety and Justice team at ideas42. Catherine MacLeod is a Principal Behavioral Designer at ideas42. Cameron French is a Senior Communications Associate at ideas42 and editor at Behavioral Scientist.

MART

CONSUMER BEHAVIOUR ONLINE: A Playbook Emerges

As consumer touchpoints migrate online, it is increasingly important to understand how human nature manifests itself in a digital environment.

by Dilip Soman, Melanie Kim and Jessica An

IN THE NOT-TOO-DISTANT PAST, the term ‘consumer’ triggered images of people in brick-and-mortar stores, touching products, physically making comparisons and receiving face-to-face assistance. Back then, product and price comparisons entailed physical transportation costs, and the act of purchasing often involved waiting in queues and using payment mechanisms such as cheques or cash.

To state the obvious, the shopping experience has dramatically changed. Consumers everywhere are embracing ecommerce options and expanding the variety of products they purchase online. At the same time, the in-store shopping experience now includes elements of technology — either offered by the retailer (in-store shopping kiosks or information display screens) or by third parties (recommendation apps or productcomparison tools).

Today’s omni-channel reality is evident well beyond the retail sector. Banks, credit card companies and insurance providers are using digital channels for various purposes including sales, marketing and customer relationship management. At the same time, ‘pure play’ digital companies — those that rely solely on digital channels — are cropping up in the financial sector.

In this article we will attempt to answer some of the questions that are vexing service providers in finance and elsewhere: Why do consumers become paralyzed in the face of abundant choice online? How does the consumption of online information differ? And how and why do the choices of other people matter in decision making?

Behavioural Patterns

In their book Nudge, Nobel Memorial Prize winner Richard Thaler and Harvard Professor Cass Sunstein make a distinction between two types of agents: ‘Econs’ and ‘Humans’. Econs are the mythical beings that inhabit Economics textbooks. Perfectly rational in the economic sense, they can process infinite amounts of information, are forward looking, unemotional, and always act in their complete self-interest. Humans, on the other hand, procrastinate, are cognitively lazy, freeze in the face of complexity and are highly influenced by context.

The problem is that many offerings today are designed for econs rather than humans. For instance, a privacy policy might provide far more information than an impatient customer cares to read; or a retirement plan might offer too many funds and end

Consumers are more honest when admitting sensitive information to a screen than to a human being.

up confusing the consumer. This common misunderstanding often results in programs and policies that do not produce the desired behavioural change.

Five concepts from behavioural economics are particularly relevant to understanding online behaviour.

BOUNDED RATIONALITY. Given the limited capacity of our cognitive apparatus, consumers are ‘boundedly rational’. While standard economic theory assumes people have an unlimited capacity to process information, the everyday behaviour of humans suggests otherwise. Consider a relatively complex financial decision: Figuring out how much you need to save for retirement. A rational approach would require forecasting of future income, expenditures, inflation rates and other unknowns, and the calculation of net present values under several scenarios. The fact is, most consumers do not possess the computational apparatus to complete this analysis. Instead, they take mental shortcuts to approximate the normative approach.

Given the reality of bounded rationality, it is not hard to see why the challenge of information and choice overload is magnified online. Take online stock trading, for example. By simply entering a keyword into a search engine, consumers have access to all kinds of information, including aggregate data on historical performance and peer investors’ opinions and choices. The sheer volume of information to sort through and options to choose from can be paralyzing, and consequently, consumers are likely to ignore the information, choose not to choose, or rely on decision shortcuts.

DECISIONS BY HEURISTICS AND SHORTCUTS. As indicated, humans are cognitive misers who do not like to expend much effort on thinking. More often than not, we resort to simplifying ‘heuristics’ when making complex decisions. For example, we stick with the default option provided, look to the behaviour of close peers, or rely on information that comes to mind readily. This search for an ‘easy way out’ holds even for important decisions, like choosing a retirement savings plan or an investment product. Due to the number of available choices online, consumers are even more likely to rely on decision shortcuts. In the example of online stock trading, consumers may base their decision on peers’ opinions, or, as one study showed, on how often they have seen the stock in the news. Such shortcuts often lead to erroneous or biased decisions.

PROCRASTINATION AND IMPATIENCE. A common theme in behavioural economics has to do with the manner in which humans deal with decisions whose consequences are spread out over

time. The financial domain is filled with such decisions, as most saving, investing and insurance decisions have to be made at a point in time where the benefits will only occur far into the future.

This stream of research can best be illustrated by the seemingly inconsistent ideas of procrastination (the tendency to delay tasks) and impulsivity (the tendency to act immediately). People tend to procrastinate on tasks that yield long-term value but shortterm pain. For instance, completing arduous forms and meeting with a wealth manager to plan for retirement yields long-term benefits, but demands time and effort in the present. Conversely, people are often impulsive in domains where the product yields short-term benefits, but might not be good for the long term. For instance, spending $5 on an indulgence might yield immediate pleasure, but if done habitually, will deplete future savings (and perhaps have an adverse effect on one’s health). By making transaction costs lower, it is likely that the online environment both reduces procrastination and magnifies impulsivity.

CONTEXT-DEPENDENT CHOICE.

A basic tenet of behavioural economics is that behaviour and choice are context dependent. The way options are presented — how they are framed, the order in which they are presented, and whether a default exists — influences consumer preferences. For example, research shows that the majority of people prefer a savings account to a life annuity when the choice is framed as ‘an investment decision’, but this pattern reverses when the choice is framed as ‘a future consumption decision’. Another study reports that people select riskier portfolios when stock-portfolio data is presented as aggregates, instead of as a list of individual stocks. In an online environment, these subtle-but-powerful ‘contextual effects’ can be manipulated much more easily — for good or for bad. For example, online platforms can influence decision making by modifying the set of choices that are presented alongside the recommended alternative or by choosing which product attributes to make salient.

PEER INFLUENCE. The effects of other peoples’ choices on human decision making have been shown in a number of domains, ranging from tax compliance to herd behaviour in financial markets. As the connected world makes it even easier for consumers to observe the preferences of others in real time, the human tendency to conform to peer behaviour is likely even more pronounced online.

Our Research

In our own research, we have identified three factors that differentiate the online decision-making environment from the offline environment.

When choices are complex, picking the popular option is a common decision shortcut.

1. The Screen Effect

Screens — whether on an in-store kiosk or an Internet-enabled device — allow consumers to view information differently. For example, a consumer can compare several mutual funds along key attributes rather than having to flip through prospectuses one at a time — and this can change decision making from an ‘alternativebased’ mode to an ‘attribute-based’ mode. Research by Rotman Professors Avi Goldfarb and Brian Silverman shows that the impersonal nature of online transactions reduces social frictions in purchasing. In particular, consumers are less shy about being honest on a screen, and this lack of ‘social oversight’ can lead to indulging in irresponsible and embarrassing behaviours.

The following three points elaborate on the different ways in which screens influence the way consumers process and evaluate financial information:

INFORMATION DISPLAY. The way screens display information is very different from the physical world, and one key difference lies in the simultaneity of information presented online. In a retail store, an appealing display might catch a consumer’s eye, leading her to examine the product closely, find it to her liking, and then look for fabric and pricing information. This sequential availability of information has the potential to create not just impulse purchase situations, but also quick appraisals that colour the interpretation of the information that follows. On the other hand, consumers shopping online are often presented information on the product and its price at the same time.

In the physical world, financial decision making often involves information that is received sequentially. For example, a consumer can select mutual funds by flipping through options one at a time, assessing each alternative holistically and in isolation (using an alternative-based mode of information processing). In the online world, consumers have the added ability of making side-by-side comparisons. For instance, investment comparison tables (accessible through sites like Morningstar, Fidelity Investments and Vanguard) rate selected funds on individual attribute dimensions, like returns and risk category.

By seeing attributes of different investment funds simultaneously on a screen, rather than having to wade through prospectuses, the consumer’s decision-making process may change from an ‘alternative-based’ mode to an ‘attribute-based’ mode. In an attribute-based mode, options are evaluated directly on how certain attributes compare across alternatives, and consumers are more prone to make substantial comparisons and trade-off analyses among attributes of given alternatives.

Further, research suggests that some attributes might be over-weighted in the decision-making process in a side-by-side comparison compared to a ‘one option at a time’ evaluation. In particular, attributes that are inherently difficult to evaluate in isolation — such as the risk or volatility of a stock — might play a significantly greater role in joint evaluation than in separate evaluation.

VISUAL BIAS. Decades of research shows that many judgments and behaviours are rooted in automatic, non-deliberative processing. A large part of automatic processing is visual, and first impressions are usually retained unless there is strong motivation to change them. Importantly, visual impressions have been shown to influence judgments of completely unrelated qualities. For instance, the greater a website’s visual appeal, the higher consumers rate its perceived usability and trustworthiness.

One experiment had participants evaluate the credibility of two finance websites. Results showed that whereas finance experts focused on information content and source to assess credibility, non-expert consumers relied heavily on overall visual appeal. This indicates that for non-experts, superficial first impressions on screens often disproportionately shape judgment in financial decision making.

THE EFFECTS OF ANONYMITY. Social interactions usually involve a degree of friction, arising from normal feelings of anxiety and self-consciousness of being judged. The online medium removes social friction by making consumers feel anonymous, and the result can be positive or negative. For example, consumers are more honest when admitting sensitive information to a screen than to a human being. Studies show that when asked about their health on a screen, patients tend to report more health-related problems and more drug use than when asked by a human being. In the context of financial decision making, we can imagine a scenario where a consumer embarrassed by his financial situation may not reveal all necessary information to a human (leading the financial adviser to recommend the wrong product), while he may be willing to reveal much more to an impersonal screen.

As indicated earlier, the downside of feeling anonymous is that people become more likely to indulge in irresponsible and uninhibited behaviour. A study done by a pizza franchise showed that sales of unusual, high-calorie orders increased when it introduced online ordering: Consumers ordered double and triple portions of toppings, and bacon sales increased by a whopping 20 per cent. Similar types of unhealthy behaviour, like overspending or buying high risk stocks, may also become more pronounced when there is low social oversight.

2. The Choice Engine Effect

When choosing between options online, consumers are often unable to evaluate all alternatives in great depth. Fortunately, technology allows them to employ choice engines that make decision making easier and more manageable. Following are two types of choice engines.

CUSTOMIZED RECOMMENDATION ENGINES: One group of choice engines called recommendation agents (RAs) conducts initial screening of available products to create a personalized consideration set. These RAs can recommend products based on past behaviour (as with iTunes’ music suggestions based on your current playlists), others’ behaviour (as with posts on Facebook’s newsfeed based on what your friends recently ‘liked’) or consumers’ explicit input of preferences. The result is a ‘personalized consideration set’ that enables consumers to zero-in on options that are of interest to them. Rather than being faced with over 100 pages of random options on Amazon, the choice engine can simplify the decision to choosing among a few options that are of actual interest.

Technology-enabled tools can also help consumers build and manage customized portfolios. Such online operations, called ‘robo-advisers’, started appearing in Canada’s online financial landscape in 2014. Consumers answer questions about their investment goals, time horizon, and appetite for risk; and using an algorithm, robo-advisers spread the consumer’s money into appropriate investments. These virtual advisers are accessible 24/7, and provide automatic adjustments to ensure the portfolio blend stays in line with the investor’s stated ideal mix as market situations change.

PREFERENCE-FEEDBACK ENGINES. Other choice engines allow consumers to solicit feedback on their preferences in real time. For example, in the mobile app FittingRoom, consumers post photos of potential purchases and other app users give feedback using ‘up votes’ and ‘down votes’. Consumers can then use the instant feedback to inform their purchase decisions. Looking ahead, we can imagine the creation of a similar platform to solicit and receive instant feedback on financial and other decisions.

3. The Connectivity Effect

When connected to the Internet, consumers have instant access to an enormous amount of information, including other people’s behaviour. As indicated, consumers look to the choices of peers

to inform their own behaviour. For example, investors find the market more attractive when more of their peers participate, and they are more likely to choose to invest in a certain stock when others have indicated a desire to do so. In short, the connected world highlights the human tendency to conform to peer behaviour by making it much easier for consumers to observe others’ preferences in real time. Two types of access are of particular interest here:

ACCESS TO AGGREGATE

MARKET PREFERENCES. In this connected age, consumers have real-time access to aggregate market preferences. For example, Amazon and iTunes publicize best-sellers on their site, and Kickstarter shows how much funding each project has received. With aggregate market behaviour displayed prominently, consumers find it easier to follow the crowd. As mentioned earlier, especially when choices are complex, picking the popular option is a common decision shortcut.

ACCESS TO OTHER INDIVIDUALS’ PREFERENCES: According to the Edelman Trust Barometer, when it comes to credible advice, people rely on ‘a person like me’ just as much as they rely on experts. The connected world allows consumers to easily refer to peer behaviour for advice. For example, TripAdvisor displays the destinations that have been visited and recommended by the user’s Facebook friends, and many social media platforms allow consumers to see purchases made by their network.

The Way Forward

The digital revolution is here to stay, which means business leaders and policymakers alike need to embrace the following principles of operating in this environment:

1. Increased honesty;

2. Greater ability to make direct comparisons, resulting in a lower degree of appraisal and a greater role of trade-off analyses among displayed attributes;

3. Greater access to information about other peoples’ choices, resulting in a greater likelihood of being influenced by others;

4. Access to an abundance of alternatives and an overload of information, resulting in a search for simpler decision strategies; and

5. Availability of decision-making tools and choice engines that reduce the effects of cognitive burden.

One particular manifestation of these elements takes the form of a decision-making strategy that we call ‘avatar-based decision making’. The traditional approach to decision making can best be ‘algorithmized’ as follows:

a) For each alternative, identify all relevant attributes;

b) Determine the relative importance of each;

c) Score each alternative on each attribute;

d) Scale the importance and the score, and compute the cumulative weighted score; and

e) Choose the alternative with the highest score.

Other models of decision making — such as mental accounting and valuation — use a slightly different algorithm, yet they all share the central idea that choice is driven by the inherent net value of the alternative. In contrast to this view, we found support for an avatar-based approach characterized by the following algorithm:

a) Identify an ‘avatar’ — a role model, a similar other, or an aspirational figure that the consumer looks up to;

b) Retrieve their choices; and

c) Use those choices as an anchor and adjust for personal circumstance.

Consider a bank that sets up a web page to help consumers navigate the complex world of financial products. That web page might have a number of sections, each for a different class of products, and each page might focus on providing the consumer with information and decision tools.

A financial institution that embraces the avatar-based approach would set up this webpage differently. In one scenario, it might present a hypothetical consumer, Justin, as part of a limited number of caricatures that represent different profiles — avatars at different stages in life, career, family, goals and net value. Justin could choose the avatar that he thinks best represents him, and use its choices as the basis for his own. In a second scenario, another hypothetical consumer, Hillary, might be asked a few lifestyle, career and family questions, and an algorithm would generate a ‘closest-match’ avatar.

Our research suggests that an avatar-based approach is more likely than other approaches to result in a robust and meaningful conversation with financial advisers. Likewise, a regulator who believes in the avatar-based approach will recognize that while consumer Stephen — who follows a traditional alternative-based

Avatar-Based Decision Making

Avatars at different life stages with different investment goals.

approach to choosing offline — will be influenced by standard product risk-related disclosures, consumer Kathleen — who relies on an avatar-based approach — will not. Indeed, for consumers like Kathleen, disclosures might be best embedded in the description of the relevant avatar.

In closing

By better understanding consumer decision making online, businesses can help consumers make better choices by providing appropriate decision support tools, and policymakers can design behaviourally-informed regulations. As indicated herein, decision making online is not merely the digitization of decision making in a bricks-and-mortar world. It is a playing field with completely different rules.

Dilip Soman is the Canada Research Chair in Behavioural Science and Economics at the Rotman School of Management, University of Toronto. Melanie Kim (MBA 2016) is a Manager in the Behavioural Insights Practice at PwC Canada and a former Associate Director at BEAR. Jessica An is a Risk Manager at TD Bank and a former Project Lead at BEAR. The full report on which this article is based (“Financial Behaviour Online: It’s Different!”) can be downloaded online.

A customer can choose “an avatar like me,” and use the avatar’s choices as a starting point for making complex financial decisions.

FIGURE 1

Managing Mental Health: A Behavioural Approach

Mental health policy needs to start embracing models and frameworks from other disciplines — in particular, the field of behavioural science.

by Renante Rondina, Cindy Quan and Dilip Soman

WORKING FROM HOME and staying home all day long. Staring at screens. The blurring of work-life boundaries. Inability to meet up with friends and family. Anxiety about one’s future and financial worries. Uncertainty about when — or if — we will ever return to normalcy. Worries that anxiety is a sign of personal weakness. These are just some of the issues facing citizens around the world during the COVID-19 pandemic.

There is little doubt about the massive scale of mental health challenges brought by the pandemic. Unfortunately, mental health was already a looming crisis before the pandemic, accounting for 22.8 per cent of the global burden of disease. COVID-19 had only accelerated the rate at which this parallel crisis has taken root in our society.

As mental health policy experts and healthcare strategists develop plans to deal with this, we call for a radical rethink of how to help people with their mental well-being. The ongoing conversation in the health policy community in Canada and elsewhere revolves around the need to improve access to mental health services. These conversations often culminate in a call for additional resources — training more providers, scaling up service delivery, and improving the quality and quantity of services offered. In particular, there has been a fair bit of work focused

on developing new treatments and therapies, investing in larger programs and policies, designing innovative delivery methods, and increasing the number of clinics, clinicians, and services with an effort to reach a larger percentage of the population. These efforts focus on the supply side of the equation and are consistent with a quote by Ralph Waldo Emerson: “Build a better mousetrap, and the world will beat a path to your door.” This loosely translates to the idea that if we build better products and services, people will automatically flock to consume them. However, as one of us argued in a recent book (The Last Mile by Dilip Soman), the ‘better mousetrap’ argument is fallacious because the builders of the new mousetrap haven’t thought of:

a) whether the value they see in the new product is shared by potential customers;

b) what frictions might prevent the conversion of latent demand into actual demand; and

c) how best to solve ‘last-mile issues’ in communicating the value and facilitating the uptake of the product.

Our collective efforts in mental health care might be falling prey to a similar fallacy. While the work and investment in mental

Unfortunately, culturally-responsive mental health services are rare.

health infrastructure is laudable, there remains a need to take a broader view of the concept of improving access. Specifically, we must ensure that we do not ignore the demand side of the equation.

For one thing, it is important to identify ‘latent demand’, specifically among citizens who fail to access services even when they would benefit from treatment. Many people with serious mental illnesses who don’t seek help show up when in a crisis that could have been prevented — increasing the stress associated with emergency visits and the high cost of hospitalization.

By attending to the demand side, we can increase the ability of citizens to make informed choices for help-seeking. In order to achieve this, our mental health infrastructure must be designed with human fallibilities in mind and there must be minimal frictions that impede access to the appropriate service. In this article, we offer a new framework for matching demand with supply in the realm of mental health.

The Mental Health Marketplace

Like any other market, the market for mental health services will be best served when there is a match between demand and supply. This matching should occur not only for the volume of services sought and delivered, but across geographies and heterogeneous groups. Obviously, different sub-segments of a diverse population — ethnic minorities, Indigenous Canadians, low-income consumers and other underrepresented groups — will require different volumes and types of support and services need to be culturally responsive to be effective.

At the most basic level, the matching process must attend to structural factors that pose barriers to mental health service use. For example, we must help individuals with minimal fluency in the official languages to be aware that services are delivered in their languages; and for Indigenous people, there is a need to decolonize mental health services and to recognize Indigenous healing traditions. Unfortunately, culturally responsive mental health services are rare.

Two main challenges exist for matching demand with supply. First, there is a need for market development. We must work to convert ‘latent users’ into actual users by designing interventions to ensure that citizens access mental health services at the earliest signs of need, before a crisis occurs. Second, a number of

people who recognize the need for services do not access them because of frictions, including complexity of information, clunky processes, lack of access or emotional barriers such as stigma. We must clear these frictions, or, as behavioural scientists say, ‘clear the sludge’. Let’s take a closer look at these challenges.

Challenge 1: Market Development

The act of market development refers to the processes of educating, preparing and organizing the potential marketplace for a new product or service. Traditionally, this includes the development of retail outlets, the communication and education of the value of the product to its potential customers, and the facilitation of sales transactions. The market development for mental health services is no different.

As with many other domains, the principle of ‘A stitch in time saves nine’ also applies to mental health. People that might benefit from accessing services often are unsure that they need it, and therefore procrastinate until they are in crisis. Therefore, one important objective of market development is to encourage people to start accessing services during the early signs of a problem.

Unlike physical health, mental health issues are typically hard to recognize at an early stage. Physical illnesses are preceded by symptoms such as headaches, muscle aches or fevers that people readily recognize. With mental health, these symptoms are typically more ambiguous and not as easy to identify.

For example, if a person experiences stress or anxiety, should they attribute it to a health issue or simply to being unusually busy? Most people do not have the expertise or experience in recognizing mental health problems and may easily misattribute symptoms of psychological distress to external circumstances. Additionally, our society has historically tended to conflate some common symptoms with idiosyncratic or dysfunctional behaviours. Given these challenges, it is easy to see why people procrastinate in accessing services.

There will likely be a further delay in accessing mental health services for the most vulnerable populations if they must choose between treating physical versus mental health problems. People with financial burdens, who live in difficult-to-access areas or do not have paid sick leave have fewer resources to deal with health issues. In the face of constraints, more salient

and visible physical health challenges might get precedence over more ambiguous and invisible mental health issues, especially for individuals who believe psychological distress will disappear over time without intervention.

The behavioural sciences offer two potential ways to mitigate these problems.

STATUS QUO BIAS: The first relates to the ‘status quo bias’ — the tendency to speak to default options. A default choice is a choice that people implicitly make if they do nothing. For example, many people choose not to open retirement accounts, get an annual health checkup or consult with their wealth managers because each of these choices involves effort. However, changing the default (without imposing any restrictions on choice) has dramatic effects on outcomes. For example, randomly assigning an annual checkup appointment but giving people the option to reschedule has been shown to increase the likelihood that people will get a checkup. Likewise, defaulting people into opening retirement accounts but allowing them to close the account later increases savings behaviour.

A similar change in defaults for mental health could have positive impacts. Starting from elementary and high schools, we must build a culture in which people are aware of the high prevalence of mental illness (about one in five Canadians experiences mental illness in a given year) and are comfortable with identifying and addressing mental-health challenges. And just as we routinely recommended annual health checkups, we also need to routinely recommend mental health checkups as a default.

Many people with mental illness have problems with motivation and behaviour. When someone experiences depression, they are often lethargic and unmotivated, whereas experiences with mania are often accompanied by engagement in risky behaviour and impaired judgment. Both extremes are related to lower motivations for help-seeking. Therefore, changing the default will likely have a positive outcome for helping people spot challenges early, and will go a long way in reducing the stigma associated with mental health.

DEVELOP A CONSUMPTION VOCABULARY: A second concept that has value in market development is that of a ‘consumption vocabulary’. In the domain of products and services, researchers have

shown that the provision of a consumption vocabulary — labels to describe why people like particular products and services — improve the learning of those products and therefore the willingness to engage more with that product category.

Consider, for example, product categories like wine. People often know when they like a particular bottle of wine, but they are unable to articulate why. Giving such consumers labels (indicators of, body, sweetness, acidity) helps them better understand their preference and also engage more to experiment and refine their tastes. A similar approach could be of utility in the area of mental health.

Most individuals are unaware of different forms of mental health challenges. In their minds, anxiety, stress and depression are all part of the same large category of mental illness. Giving people labels to attach to discrete symptoms can help them communicate how they feel in a nuanced fashion, thereby allowing them to recognize the challenges and increase their confidence to speak about these issues with service providers.

Modern-day marketing also effectively harnesses the power of social norms, influencers and word of mouth. In a digital era, many people only share highly curated images of themselves. For such individuals, it is particularly important to cultivate a culture of openness and self-reflection about mental health challenges.

There are also some who share openly about their emotions and daily experiences. Educating citizens to learn to identify symptoms in others is particularly valuable for early intervention for this group, as mental health challenges reduce cognitive bandwidth, which may make accurate self-evaluation difficult. It can be extremely helpful to have a trusted friend, family member or confidante identify signs of distress, reach out and suggest seeking treatment.

Challenge 2: Eliminate Frictions

Even after solving the problem of converting latent demand into actual demand, Sludge can prevent people from accessing a service. Sludge is often not intentional, but is a bit like weeds in a garden: Both are initially difficult to spot, need constant clearing up, and ignoring them can result in rampant and uncontrolled growth. Following are four categories of sludge to look out for.

Most individuals are unaware of different forms of mental health challenges.

CHOICE OVERLOAD. A focus on the supply-side has produced thousands of available options for individuals seeking help. However, users often do not know which services can best address their needs. Individuals in distress with limited cognitive bandwidth and motivation may find it difficult to search for and evaluate their options. This state of ‘choice overload’ may lead people to defer making a choice and not seek help at all. Health benefits providers can help their policyholders overcome this obstacle by presenting a tailored list of eligible services, such as a network of eligible therapists and counsellors in a policyholder’s area based on client preferences (e.g., cultural needs).

FRAGMENTED SYSTEMS. When a person has to seek services from two or more providers, care is often not coordinated. A counsellor might recommend consulting for pharmacological intervention to a client, but a consultation for medication must be completed with a physician. At this point, it is not only up to the client to seek the physician but also to keep both providers updated regarding any changes in symptoms, side effects or dosages. Another example may be an escalation of illness severity, whereby a client must seek in-patient care but must provide their own clinical history.

EXCESSIVE PAPERWORK. Another common source of sludge is excessive paperwork. Some plans require a formal diagnosis or a prescription from a physician. For example, Ontario’s BounceBack coaching program — which connects users to a virtual therapist — used to require users to enter their primary care provider’s billing number and professional ID. This can lead to procrastination and raises inequity concerns for individuals without a family physician. Furthermore, perceived stigma may also prevent individuals from contacting their care provider to register for this program. Policymakers should identify similar process frictions and streamline the steps involved for registering in their programs to improve ease of access and increase uptake.

STIGMA. Seeking help usually involves interacting with a healthcare provider, which may be difficult for a person if they perceive stigma from the provider, especially if they are already part of an ethnic or racialized minority. Stigma often leads to the avoidance of clinics or other public places where people might be

labelled. Policymakers should ensure that individuals at risk or with mild symptoms can self-refer to an appropriate low-intensity service. This would allow them to avoid any potential stigma and make them more likely to access those services, thereby preventing further deterioration, and help close the mental health gap for minority groups.

The Way Forward

As indicated herein, we must think about mental health in much the same way as we think about developing any other marketplace. How can we accomplish this? We have five recommendations.

1. IMPROVE COORDINATION AND OVERSIGHT. Many inefficiencies and lack of market development arise due to a lack of coordination in the delivery of mental health services across the nation and across multiple entities. If we had a comprehensive strategy for mental health and resources devoted to market development and sludge-reduction efforts, as well as processes for sharing resources across organizations and provinces, we might move towards reducing some of the identified inefficiencies.

2. EMBRACE DIGITAL SOLUTIONS. Digital tools can overcome many of the structural barriers presented by traditional inperson services, including lack of access, long wait times and stigma. Self-guided digital solutions are suitable for individuals with mild symptoms or those at risk of developing mental illnesses and have been shown to produce modest improvements while protecting the individual from further deterioration. Such solutions might also help to overcome stigma, as individuals do not have to disclose their condition for self-guided treatments. Therapist-guided solutions can be even more effective than self-guided approaches and as effective as traditional in-person services. Therapist-assisted digital solutions may also reduce stigma by removing the individual from the physical presence of the therapist.

3. INCREASE MENTAL HEALTH LITERACY. Campaigns can increase population-level recognition of mental illness and improve

understanding of the benefits of treatment. They can also reduce stigma by normalizing mental illness as a common phenomenon that most people will experience in their lifetime and which can be improved with treatment. Campaigns that inform individuals of prevalence and highlight the biological, psychological and social factors contributing to mental illness can also reduce self-blame and stigma. Using a bio-psychosocial model reframes mental illness as a social issue rather than locating the responsibility within the individual. Mental health literacy campaigns can also instill hope by including examples of individuals with mental illness who are living successful lives post-treatment, motivating consumers to act. Furthermore, by improving mental health literacy at an early age, we can normalize the idea of regular mental health checkups as part of everyone’s regular default routine.

4. REDUCE SLUDGE. As indicated, sludge-reduction efforts can make it easier for people to seek help. Because vulnerable individuals might not know they need help, governments should make validated screening instruments widely available to help citizens determine whether they could benefit from support. Self-administered web-based tools could be completed regularly to help care providers monitor any changes in a patient’s health. Learning from past research, health benefit providers could also automatically assign their policyholders to the next available appointment with an eligible service providing the appropriate level of care, providing an option to reschedule or select a different provider. Finally, as health benefits providers are responsible for monitoring service usage, they are in an ideal position to coordinate such services to enable the integrated and seamless delivery of care.

5. EMBRACE SOCIAL MEDIA TOOLS. Despite the fact that social media has been blamed for a variety of mental health problems, it can also be of help in several ways. We know that people tend to look to other people who are similar to themselves or to social influencers (rather than experts) for advice and information. As a result, campaigns by prominent social personalities (both in the physical world, within communities,

and on social media) advocating for mental wellness and encouraging help-seeking could be beneficial. In an online context, machine learning algorithms could also help to identify patterns of posting that are indicative of particular problems, and peers (both online and in the community) could be trained to identify patterns of behaviour that signal the need for intervention.

In closing

With the wide variety of technological tools available today, there are clearly many different fronts on which to fight the mental wellness battle. Our roadmap is not meant to be comprehensive, but our point is simple: In order to win this battle, we need to do more than simply deploy more resources. By embracing lessons learned from the behavioural sciences and attending to the demand side of the equation, we can increase citizens’ ability to make informed choices for help-seeking — and improve the collective well-being of our society.

Renante Rondina (PhD in Psychology 2019) is a Behavioural Scientist in the Office of the Chief Human Resources Officer, Treasury Board of Canada Secretariat and former PostDoctoral Fellow at Behavioural Economics in Action at Rotman (BEAR). Cindy Quan (BSc 2016) is a PhD Candidate in Psychology at the University of Victoria and a former research assistant at BEAR. Dilip Soman holds the Canada Research Chair in Behavioural Science and Economics, is a Professor of Marketing at the Rotman School and is the founding director of BEAR. He is the author of The Last Mile: Creating Social and Economic Value from Behavioural Insights (Rotman-UTP Publishing, 2015) and co-edited The Behaviourally-Informed Organization (Rotman-UTP 2021)

Segmentation Reimagined

IN THE 1970S, the clothing company Laura Ashley was the pinnacle of women’s fashion. Known for their instantly recognizable and classic feminine floral-print aesthetic, the retail giant had more than 220 stores worldwide, including 94 in North America. They were undisputed leaders in the market and a firm favourite of upper and upper-middle class women, with a fan base that included Lady Diana Spencer. Then, more women started going to work.

Laura Ashley’s target customer was affluent married women over 30 years of age. However, as more married women entered the labour force, their needs changed. They wanted clothing that was stylish and functional for work. But the company failed to adapt. They continued to target the same group of women with the same clothing options. By 1999, Laura Ashley sold all of its North American retail shops in a management buyout for US$1.

Organizations, including businesses, not-for-profits and governments use segmentation to identify their target market and develop products, programs and policies to fit that market. Companies often segment their customers using demographics like age, income or gender. The fundamental idea is simple — create groups of customers, such that the behaviour towards your product is similar for people within a group and dissimilar across groups. The act of segmentation makes sense. It helps companies focus on designing the product and developing a marketing strategy for whom they believe will benefit most. Marketing Laura Ashley’s products to a 62-year-old male retiree is not a good business strategy. Targeting middle-aged married

women is — until their needs change and the company fails to adapt.

So how can companies understand their customer segments and avoid a fate like Laura Ashley? Segmentation has traditionally relied on the beliefs that people are different from each other, that they can only belong to one segment, and that they rarely change segments. Note that while the term segmentation is used primarily in marketing departments, it is a universal phenomenon. Product designers group users into personas, creative artists group people into audiences, and college administrators group students into typologies. Even in social settings, it is not uncommon to group people.

We do not dispute the notion that people are different from each other. Many organizations still lean heavily on this philosophy. But advances in our understanding of human behaviour and the technology at our disposal mean segmentation is ready for an update. We can improve how companies segment their customers, in three fundamental ways.

1. Move from ‘people are different from one another’ to ‘people are different from one another and from themselves over time and across situations’.

2. Move from correlates or predictors of behaviour to observing and using actual behaviour to segment.

3. Move from static, one-time categorizations to real-time, continuous adjustment where customers could belong to different groups at different points in time.

Three Updated Segmentation Features

Traditional Segmentation Segmentation by Usage Segmentation by Behaviour Segmentation Reimagined

Traditional demographic or geographic segmentation. Does not account for differences within a person.

Example

A clothing store segments its audience by their age, gender, and income level. Customers remain in one static segment depending on their demographic.

Segmentation by usage or related behaviours. Includes context as a factor that changes consumer behaviour.

Example

A shoe company segments its customers based on the occasion in which their shoes would be worn. Depending on the context of the purchase, customers can belong to multiple segments.

1. Segmentation by Usage

People are different from each other. We agree! However, research suggests that people might also be different from themselves over time and across situations. In other words, human behaviour is time and context specific. Consider the following examples.

Motivation to accomplish tasks waxes and wanes with the passage of time. People are mostly motivated to take on virtuous tasks like eating healthy or starting a new assignment at salient points of time. Wharton Professor Katy Milkman and her colleagues refer to this as the ‘fresh start effect’. This could be the beginning of a new year, a birthday, anniversary, or other critical life events. The same individual might react differently to a request to join a fitness program if they are asked in January versus in April.

Research on the ‘hot-cool empathy gap’ suggests that people’s behaviours, attitudes, emotions, and judgments are dramatically different when they are in a ‘hot’ state (for example, emotional or aroused) than when they are in a ‘cool’ state. These differences can result in impulsive behaviours on the one hand, and detached systematic thinking on the other hand by the very same person, depending on the state they are in.

Using real-time data to segment individuals based on their actual behaviour rather than cues. Based purely on behaviour rather than individuals themselves.

Example

A non-profit’s website changes its call to action based on the what device the user is on.

Using real-time data specific to various situations, rather than individuals. Contextdependant and data driven.

Example

A bank website could merge its customer data with its browsing behaviour on a given day to customize the layout of the web page.

Likewise, research shows that decision-making online is quite different from offline. The same customer might order extra-cheese and extra-toppings when they place orders online at a pizza place, but would not do the same thing while ordering in person. Rotman Professor Avi Goldfarb, Duke’s Ryan McDevitt, ESE’s Sampsa Samila, and Rotman Professor Brian Silverman show that this happens because in-person purchasing can activate social embarrassment that the very same customer will not experience when ordering online.

There are a number of other examples of people being different from themselves. Rotman Professor Lisa Kramer has shown that investing behaviour is different in the winters relative to summers. Duncan Simester and Drazen Prelec show that the same customer might pay more and spend differently while using a credit card instead of cash.

2. Segmentation by Behaviour

Traditional segmentation relies on correlates of behaviour. For example, a marketer might infer that every Millennial will exhibit similar purchasing behaviour when shopping for furniture simply because they are in the same demographic group, or that all residents of a particular city behave

FIGURE ILLUSTRATES THESE THREE IMPROVEMENTS TO SEGMENTATION — BY USAGE, BY BEHAVIOUR, AND SEGMENTING DYNAMICALLY, RESPECTIVELY.

By understanding that people are different from each other and themselves and using actual behaviour as an indicator, we have the elements to begin approaching segmentation as something that is ongoing and dynamic.

differently from all residents of another city. With advances in data science and the fact that a lot of behaviour can now be observed online, companies can use actual behaviour to understand their customers. Online retailers like Amazon and search engines like Google already use actual behaviours rather than cues — the ads you see, the products that are recommended to you, and in some cases, even the prices that you get, are all a function of their observations of your behaviour. What does all this mean for the practice of segmentation? It means the unit of analysis should shift from an individual to specific instances of an individual’s behaviour (i.e., an individual-situation combination).

For example, an online florist could segment based on whether web visitors were looking for a gift or flowers for a home and not on age, income or geography. Gift purchasers care about the aesthetics of the arrangement and the speed and timing of delivery, and they are relatively less price sensitive than purchasers of flowers for their home, who care more about the longevity of the flowers. A bank website could merge its customer data with browsing behaviour on a given day to customize the layout of the web page. A government website could provide different messages based on whether constituents are accessing the message on a computer or a phone. And a charitable donations website could offer different messages and layouts as a function of when and from where people land on their website. In this sense, segmentation is based on actual behaviour and is dynamic. The web layout that this customer sees today might be different from the one that they see tomorrow if their behaviours change.

And, rather than a rigid view of segmentation in which the same segmentation scheme is applied across the organization, there could be different segmentation schemes as a function of which stage of the value creation process is being considered. In particular, there could be a different segmentation scheme for intervention design, for communication design, and for last-mile distribution, merchandising and sales.

3. Dynamic Segmentation

Given that human behaviour is time and context-specific, it does not make sense to tether people to a particular segment that is time- and context-independent. By un-

derstanding that people are different from each other and sometimes from themselves and using actual behaviour as an indicator, we have the elements to begin approaching segmentation as something that is ongoing and dynamic, rather than set once and static.

Had Laura Ashley had this knowledge and the ability to collect data over time at their disposal, they could have observed that fewer women would enter their shop during working hours and accordingly produced more practical clothing to suit customers’ needs. Alternatively, their pricing and displays might have looked different on weekends (when working women visited stores) from the weekdays. While this was logistically and economically difficult to do given the world of the 1970s, the technological and behavioural tools in the 2020s make it easier for companies like Laura Ashley to adapt to their customers’ changing tastes and preferences by adopting a dynamic approach to segmentation.

What’s Next

Our science and capabilities have evolved dramatically over the past few years, but the mechanics of segmentation has not kept up. Behavioural science tells us that while people are different from each other, they are also different from themselves. And data science now allows us to observe behaviour and immediately customize marketing interventions rather than rely on surrogates of behaviour. Together they conspire to give us a dynamic approach to segmentation, which might be the difference between understanding your customers and failing to do so.

Kayln Kwan is currently at New York University and was a project lead at the Behavioural Economics in Action at Rotman (BEAR) research centre at the University of Toronto. Dilip Soman is the Canada Research Chair in Behavioural Science and Economics at the Rotman School of Management, University of Toronto.

Does Budgeting Actually Work?

FINANCIAL ADVISORS

frequently implore people to set a budget. The logic behind this advice is that budgeting helps consumers manage their spending and improve their financial well-being. And yet, consumer spending and debt continue to climb, even as financial technologies like budgeting apps have rapidly grown in popularity.

So, what gives? Does budgeting actually work?

Budgeting has been a focus of academic research for decades. However, most budgeting research has examined the psychology of budgeting rather than its influence on people’s real-world spending. This means that we have a deep understanding of how people think about budgeting decisions, but we do not know much about whether those decisions actually impact their behaviour in meaningful ways. Our research addresses this knowledge gap by using budgeting app data and field experiments to directly examine the influence of budgets on consumer spending and determine if budgeting works.

The Psychology of Budgeting

Consumers set budgets for many reasons. Often, people use budgets as a self-control device to reign in their spending. Other times, people set budgets to take stock of their spending priorities, or to simply gain an accurate view of how much they will spend in the future.

Regardless of why someone sets a budget, it subsequently becomes a reference point or benchmark against which they can track their spending. And that is where the potential power of budgeting lies: A great deal of research shows that reference points can influence our behaviour, as long as we actively monitor our behaviour in relation to the reference point.

Budget Optimism

Optimism is the belief that the future will be better than the past. In the context of consumer financial well-being, this almost always means spending less than usual and using the difference to decrease debt or increase savings. Thus, given that optimism is a common characteristic of human judgment and decision making, it is perhaps unsurprising that most people set their budget well below their average past spending.

Recognizing that budgets tend to be optimistic is important for understanding whether budgets actually work because there are two ways to define budgeting success. The first way is strict budget compliance. Under this definition, budgeting ‘works’ if you spend less than or equal to your budget — and it is a failure if you spend more than your budget. The second way to define budget success is simply spending less than usual. As an example, consider someone

After setting a budget, people continued to decrease their spending every month for six months.

who usually spends $500 per month at restaurants. Now assume this person sets a budget of $300 for next month and subsequently spends $400. Are they a budgeting failure because they spent $100 more than their budget, or are they a budgeting success story because they spent $100 less than usual? We will leave that value judgment to you, but we think most people agree that all else equal, this person’s budget has influenced their spending in a positive way.

To determine if budgeting actually helps people decrease their spending, we ran a series of three studies.

The App Study

We first partnered with Money Dashboard (MDB), a popular personal finance app in the U.K. that provides its users with a holistic overview of their financial situation. To accomplish this, the app collects and combines all transactional information across all financial accounts for each user. For example, if a user has three credit cards, a chequing account and a savings account across different providers, MDB will aggregate all inflows and outflows across these cards and accounts and present the user with up-to-the-minute information on when, where and how they are spending their money. Importantly for our research, MDB also has a budgeting function that some people use and others do not. In total, we were able to observe the financial behaviour of approximately 70,000 app-users who made more than 350 million transactions over a three-year period.

We compared how much budgeters spent during the month immediately after they set a budget to their average monthly spending over the three months before they set a budget. In the two most popular budget categories — dining out and groceries — budgeters decreased their spending by an average of 16.71 per cent and 9.42 per cent respectively. However, budgeters also tended to spend a lot more than their budget: in the same pair of budget categories, people exceeded their budget by an average of 38.01 per cent and 22.11 per cent. In other words, budgeters were spending somewhat less than they used to, even though they continued to spend much more than they budgeted.

What do you think happened the next month? For example, if a person sets a budget on June 1st and throughout that month spends somewhat less than usual, but much more than they budgeted, what do you think happens to their spending in July? Some research suggests that failing to stay on budget might cause people to abandon their budget and let their spending rise back up to pre-budget levels. We were therefore pleasantly surprised to see the opposite: after setting a budget, people continued to decrease their spending every month for six months, even though they continued to spend more than their budget.

In our next analysis, we leveraged the fact that some people who use the MDB app set budgets while others do not. This allowed us to compare the spending of very similar people who had either budgeted or not. For example, if Jack and John both earn $3,500 per month and spend $500 on dining out, but then Jack sets a budget of $300 for next month and John does not, who ends up spending less? Sure enough, we found that budgeters like Jack end up decreasing their spending a lot more than non-budgeters like John.

The Field Experiment

The strengths of our first study include a large sample of consumers and precise measurement of their real-world budgeting and spending decisions over a long period of time. One limitation is that the descriptive nature of the data did not allow us to determine conclusively that budgeting caused the spending changes we observe. Although the evidence strongly suggests that that is the case, the results of the first study only let us say that budgeting is correlated with lower spending.

To address this, we next ran a field experiment in partnership with Vancity, a community credit union based in Vancouver, BC. At the start of the month, we randomly assigned 226 Vancity members to either set a budget for their discretionary spending or not. Then both groups tracked and reported their spending on a weekly basis for the rest of the month. The magic of random assignment means that the only meaningful difference between these two groups

Budgeting works for many different types of people, indicating that it is a broadly applicable financial management tool.

was whether they set a budget or not. Therefore, if budgeters subsequently spend less money than non-budgeters, we can conclude that budgeting is what caused spending to decrease.

Sure enough, during the month of the experiment, participants in the budgeting group spent 21.9 per cent less on discretionary expenses than participants in the non-budgeting group. This shows that budgets do cause people to decrease their spending. In other words, this provides clear evidence that budgets work. However, as in the first study, participants in the budgeting group also spent significantly more (36.2 per cent) than they budgeted. This shows that although budgeting works, most people still have room for improvement.

Another finding of note from this study is that budgeting was equally effective for different types of people. For example, budgeting worked just as well for impulsive people as non-impulsive people, for planners as for non-planners, and for optimists as for pessimists. This suggests that budgeting not only works, but it also works for most people.

The Financial Diary Study

In our third and final study, we wanted to directly examine the extent to which an optimistically low budget is beneficial. To set the stage, imagine a consumer who typically spends $500 per week, budgets $300 for next week, and subsequently spends $400 during the week. In essence, the question we wanted to answer in this study was: What happens if this person budgets $400? Do they become a perfect budget success story and spend only $400? Or is it possible that a higher budget causes them to spend more than they would have otherwise?

To examine these possibilities, we conducted a financial diary study in which 340 consumers in the U.S. were randomly assigned to make either a relatively high or low budget for the next week. They then recorded their spending at the end of each day in an online financial diary. As expected, people in the high budget group made substantially larger budgets than people in the low budget group ($255.44 vs $189.88). However, people in the high budget group also

spent significantly more than people in the low budget group ($498.35 vs $374.32). This is important because it reveals that when your budgeting goal is to minimize spending, an effective budget is not always an accurate budget.

In closing

We started this article with a straightforward question: Does budgeting actually work? Our research shows that the answer is yes, although imperfectly. In our studies, budget compliance tends to be weak — meaning that most people spend more than they budget. However, we also see that budget influence is strong — meaning that budgeters spend less than non-budgeters, spending continues to decrease months after a budget is set, and lower budgets are associated with lower spending. Moreover, we also observe that budgeting works for many different types of people, indicating that it is a broadly applicable financial management tool.

Past research has revealed a great deal about the psychology of budgeting, typically by using tightly controlled lab experiments that examine hypothetical decisions and outcomes. Our work shows that important new insights regarding the budgeting process can also be generated by examining budgets and spending ‘in the wild’. As app data and field experiment opportunities become more readily available, we encourage researchers and practitioners to examine budgeting outside of the lab, so that its real-world costs and benefits can be better understood. Finally, to all readers who would like to reduce their spending, we can strongly recommend setting a budget!

Ray Charles “Chuck” Howard is an Assistant Professor of Marketing at the Mays Business School, Texas A&M University. Marcel F. Lukas is a Senior Lecturer in Banking and Finance at the School of Management, University of St. Andrews.

BRIDGING THE ORGAN DONATION GAP WITH BEHAVIOURAL SCIENCE

The study’s success in motivating individuals to become registered organ donors offers hope for addressing the growing demand for life-saving transplants and improving public policy.

by Nicole Robitaille and Nina Mažar

AN ORGAN TRANSPLANT IS A PROCESS wherein a healthy organ from a donor can be surgically removed and placed in the body of a recipient human being to replace a diseased or damaged organ. However, there is currently a worldwide shortage of organs available for transplantation, with demand far outstripping supply. According to the Government of Canada, over 4,300 people currently await lifesaving organ transplants in Canada, while in the U.S., the Health and Resources Services Administration reports that there are more than 100,000 people on the national transplant waiting list, and on average, 17 people die each day while waiting for a transplant.

In the quest to bridge the ever-widening gap between the demand for organ transplants and the availability of donors, there is a simple yet powerful solution: increase the number of registered organ donors. Past research has shown that those who have taken the step to register their consent are much more likely to ultimately donate their organs. This means that each registered donor can potentially save multiple lives through their selfless act of giving.

There has been a growing interest in changing the way we approach organ donation policies, such as moving towards a system where everyone is automatically considered a donor by default but with the option to opt out if they wish. While this idea

seems promising, the impact on actual organ donations has been somewhat mixed. This is due to various factors, including families, who make the ultimate decision, being uncertain about the preferences of their lost loved ones when it comes to organ donation. Therefore, we wanted to explore if there were ways we could improve registration numbers, within the current registration systems.

During our time at Behavioural Economics in Action at Rotman (BEAR), we conducted an experiment together with our Rotman colleagues Claire Tsai and Avery Haviv, as well as Elizabeth Hardy from the Ontario Behavioural Insights Unit, where we explored how seemingly small interventions grounded in behavioural science can make a big difference in the percentage of people who register to donate their organs. Given that one single donor may save up to eight lives and enhance as many as 75 more lives, this research has the potential to have a meaningful impact on the lives of many.

The Gap Between Intentions and Action

The decision to donate organs posthumously is a highly personal decision and is a decision that is not made lightly. Many governments and organizations expend significant effort and resources to try to improve attitudes towards organ donation. Similarly,

Organ Donation Forms

Standard Form in 2014

Organ and Tissue Donor Registration

many academic researchers have studied ways to increase individuals’ intentions to donate. What is particularly interesting about the issue of organ donation however is that attitudes and intentions towards organ donation are already remarkably positive.

For example, 90 per cent of adults living in the Province of Ontario support organ donation, and a recent survey found that 81 per cent say that they themselves would be willing to register as organ donors, yet in Ontario, only 35 per cent are actually registered as donors. Similarly, in the U.S., only 60 per cent of the population are registered, despite 90 per cent of Americans supporting organ donation. These statistics make one wonder that perhaps the organ donation problem is not just about finding ways to improve intentions and attitudes to organ donation. Rather, it seems the bigger issue is actually the gap between peoples’ intentions and their actions. Closing this gap between good intentions and follow-through action could have significant social benefits.

One effective approach to close the intention-action gap often used by behavioural scientists is to first put oneself in the consumer’s shoes to discover what the barriers to action are, and then develop and test solutions to overcome these barriers. When we learned that 85 per cent of organ donor registrations in Ontario were happening in-person at ServiceOntario centres, even though it is possible to register online, we decided to use ServiceOntario locations as our starting point.

The Organ Donation Landscape

Imagine going to ServiceOntario (or a Department of Motor Vehicles service centre in the U.S.) to renew your driver’s license. After taking a number, waiting to be called and finally nearing the end of your transaction, you are asked if you would like to register as an organ donor today. The reality is that organ donation is probably not a top-of-mind decision at that moment, and that the question comes as a bit of a surprise. Consequently, for a decision that feels personal and important, it is not surpris-

Nudge Statements in Brochure

Information Brochure

Nudge Statement

‘Reciprocal Altruism’ Nudge: If you needed a transplant, would you have one?

‘Imagine Self’ Nudge: How would you feel if you or someone you loved needed a transplant and couldn’t get one?

‘Imagine Other’ Nudge:

How do you think people feel when they, or someone they love, need a transplant and can’t get one?

ing that most people say “no” when they are asked. Yet, at the time of our experiment, this was the standard organ donor registration process in Ontario as well as most of North America. Customers who do say “yes” are then handed a lengthy form to complete. The form is full of information and legalese, extending their stay at ServiceOntario even further. You might not be surprised to learn that at that point many customers decide they might just wait to register another time.

Embarking on a Real-World Journey: Our Partnership with ServiceOntario

With these issues in mind, we collaborated with ServiceOntario to design and conduct a large-scale real-world experiment aimed at improving organ donor registrations. This experiment unfolded over a 2.5-week period back in 2014 at one of their service centre locations. We chose to run the study at one of the largest and most frequented centres in the province. The centre’s demographic makeup mirrored that of Ontario’s population,

spanning various age groups, income levels, educational backgrounds, and religious affiliations. This carefully selected setting ensured that our research findings could offer insights applicable to a wide spectrum of people across the province.

During our experiment, we tested a combination of changes designed to improve the registration process along with several behavioural interventions targeting the intention-action gap. First, we created a simplified in-person registration form to increase the likelihood that the form was not only read but understood. To focus consumers on the key action we wanted them to take, we removed all additional information, facts and statistics from the form. Additionally, given that these customers were already at ServiceOntario completing transactions and the centre knew their name, address and other details, we removed the questions that requested duplicate information the centre already had. We then printed the simplified and smaller (i.e., only a half-page) registration form on sturdy cardstock so that it could be filled out by someone standing in line or sitting

90 per cent of adults living in the Province of Ontario support organ donation, and a recent survey found that 81 per cent say that they themselves would be willing to register as organ donors, yet in Ontario, only 35 per cent are actually registered as donors.

in the waiting room and added a colourful banner at the top of the form designed to catch people’s attention.

Secondly, we tested improvements to the registration process itself. The changes included handing out the organ donor registration form, along with the visitor’s waiting number, to all visitors when they entered the service centre instead of at the end of their license renewal process. This gave everyone time to read, process and consider the form and the decision of whether or not to register at their own pace. And because it was printed on cardstock, they could also complete it while they waited which would save time for them, the service agents and the other customers waiting.

Among the service centre visitors who stepped through its doors during the experiment, one in every five experienced only these changes in the registration process. That is, they received the revamped registration form handed out in advance, making the journey towards becoming an organ donor easier and more efficient. But would making the process easier be enough? Or could we increase registrations even further if we gave consumers a little nudge?

For the remaining four out of five customers that visited the centre during our experiment, in addition to the process improvements, a persuasive twist was added to the mix in the form of either an information brochure or a ‘nudge’ statement. This idea of a nudge, introduced by the Nobel Memorial Prize winning economist Richard Thaler and legal scholar Cass Sunstein, involves making small changes in the choice environment to influence peoples’ behaviour in a predictable way, all without taking away choices or altering their economic incentives. In this case, it meant adding a behaviourally informed persuasive message to the registration form.

One group received the province’s standard organ donation brochure (which provided information and FAQs about organ donation), along with the simplified form. If low registration rates were the result of not having enough information at the time of making the decision to register, then this brochure should’ve outperformed our other conditions. However, what if it was not just information that was needed, but the motivation to act today to help others?

The rest of the visitors were presented with the redesigned registration form, along with one of three persuasive statements printed within the colourful attention-grabbing banner at the top of the form. These statements were designed to nudge readers into considering organ donation. Relying on a concept called ‘reciprocal altruism’, which is similar to the ‘tit for tat’ strategy, the first statement had individuals consider what they would do if they themselves needed an organ transplant. Building on the concept of perspective-taking, the second statement prompted individuals to think about how they would feel if they or a loved one needed a transplant but couldn’t get one. And the third statement prompted individuals to think about how others would feel if they or a loved one needed a transplant. Each statement concluded with a heartfelt plea: ‘Please help save lives and register today.’

Unlocking the Power of Persuasion: A Game-Changer for Organ Donor Registrations

As behavioural scientists, we always hope our work will have real-world impact, that we can help improve the behaviours and well-being of others. We were excited about this experiment and our partnership, having the ability to test process and persuasive improvements in the field, in an area that matters. We were even more excited when we saw the results.

By developing an improved registration process and adding persuasive elements to this streamlined process, we were able to significantly increase real, in-person organ donor registrations. The results of the study revealed that adding any of the persuasive elements in addition to the process improved registration rates in comparison to the centre’s standard approach.

Among the persuasive elements tested, one stood out as the winner: the ‘reciprocal altruism’ nudge: ‘If you needed a transplant, would you have one?’ Adding this seemingly simple question to the top of the simplified form had a massive impact, increasing actual registration rates from 4.1 per cent to an impressive 7.4 per cent. That is an astonishing 80 per cent increase when compared to the process changes alone and a staggering 135 per cent improvement over the centre’s standard approach.

If all of our recommended changes were adopted province-wide, we estimate that they would have the potential to increase organ donor registrations by about 225,000 individuals annually.

While the ‘reciprocal altruism’ prompt performed best, all three perspective-taking prompts demonstrated positive effects, emphasizing the viability of various approaches to inspire donors. Moreover, they highlight the importance of altruistic motives for motivating action and overcoming the intention-action gap, particularly in an area where acting involves helping others.

Here is the other game-changer: our streamlined organ donation process made sure that every single customer had a chance to take in the materials and decide at their own pace. No rush, no hassle. The shorter forms took less time to complete, and a survey indicated that they significantly reduced the time it took a service agent to complete a new donor registration. Faster registrations also have the potential to reduce overall wait times, making the entire process more consistent and efficient.

But it doesn’t stop there. This new approach also lightened the load for the service agents, who no longer needed to push registrations, making the experience smoother for both customers and staff. It was a win-win for everyone involved!

What made this win even sweeter was its affordability. The only costs associated with this experiment were a one-time cost for redesigning the form and the cost of printing the coloured form on cardstock. In essence, we developed a low-cost scalable solution with high-impact potential.

The study’s findings also underscore the significance of considering the process and timing of delivering promotional materials. Intercepting customers at the right moment and providing them with information and prompts can be instrumental in encouraging organ donor registrations. This was particularly evident with the brochure. The brochure is always available at ServiceOntario centres for people to browse, and it is mailed home with driver’s license renewals. Even though it was always available through those other channels, having it at the time of decision making significantly increased registrations.

Taken together, if all of our recommended changes were adopted province-wide, we estimate that they would have the potential to increase organ donor registrations by about 225,000 individuals annually. That is a staggering number of potential lifesavers stepping up to make a difference.

Our findings also align with related work conducted in the United Kingdom, where a ‘reciprocal altruism’ nudge performed best and increased individuals’ organ donor sign-ups from 2.3 per cent to 3.1 per cent. This study was the first to illustrate the potential for low-cost, scalable interventions and persuasive messages, specifically, to improve actual organ donor registrations. However, it was unable to distinguish between registrations from new and existing donors. In addition, it was conducted online at a time when most transactions were done in person.

Thus, we not only replicated the success of the ‘reciprocal altruism’ nudge but also extended its impact by showcasing its effectiveness in encouraging new registrations within a typical in-person registration system. It is clear that a little nudge can go a long way when it comes to saving lives through organ donation. We also replicated its effectiveness in a different national culture. Together our findings increase the confidence that ‘reciprocal altruism’ is a useful nudge in the behavioural scientist’s toolbox.

Changing the World With Research

The research documented in our paper was recently awarded the 2023 FT Responsible Business Education Award, which recognizes academic research that addresses societal challenges, with evidence of positive impact on policy or practice. Only four projects from researchers around the world received this honour. It has also received several other academic and government honours, including the Ontario Government’s Amethyst Award for Outstanding Achievement in the Ontario Public Service.

However, what has been most impressive about this work is the significant and real-world impact that it has had in several areas. In 2016, ServiceOntario partially adopted our recommendation by introducing a simpler organ donor registration form, with our best performing reciprocal nudge printed on top, which has been attributed to increasing actual organ donor registrations in the province. In addition, Ontario’s Trillium Gift of Life Network developed communications inspired by our research, as illustrated in the social media post and paid ad below.

Social Media Post and Paid Ad

In closing Marketers expend an enormous amount of effort to gauge purchase intentions (analogous here to intentions to donate). But as researchers who study consumer behaviour have shown, the gap between intentions and follow-through can be wide. By leveraging behavioural science, we have shown that the gap between what people say they will do and what they actually do can be reduced.

With thousands currently on the transplant waiting list, the need for organ donors is urgent, and as the population ages, the demand is only predicted to increase further. By intercepting potential donors with well-timed information, perspectivetaking prompts, and a streamlined registration process, we were able to bridge the gap between intention and action. The study’s success in motivating individuals to become registered organ

donors offers hope for addressing the growing demand for lifesaving transplants and improving public policy. It also highlights the potential to enhance societal welfare and save countless lives through innovative approaches to organ donation.

Nicole Robitaille (Rotman PhD 2014) is an Assistant Professor of Marketing at the Smith School of Business, Queen’s University. Nina Mažar is a Professor of Marketing at the Questrom School of Business, Boston University.

This is a great example of how the government is using evidence-based decision making to better serve its citizens, and how research into consumer behaviour can benefit society. Behavioural interventions, when supported by streamlined processes, can be cost-effective and efficient ways to improve societal well-being without limiting the freedom of individuals, raising ethical concerns, or passing new legislation.

Taking Charge of Financial Information

DO YOU KNOW the standard fee for making a late payment on your credit card bill? More importantly, would you like to know? According to nationally representative survey research done by Harvard legal scholar Cass Sunstein and economists Lucia Reisch and Micha Kaiser, about 60 per cent of people across 11 democratic nations say they would not like to know. About half also indicated that they would rather not know how much it costs them annually to run their household appliances.

These are a couple of the facts from a growing body of research that shows that people regularly turn down potentially useful financial information, even when it is free to get and easy to access. However, it is not as if all types of information are rejected. It turns out that people are keen on receiving prospectively good financial news — and will expend considerable time, effort and money to get it, even when they cannot do anything with the information — whereas they go out of their way to avoid information that attests to potentially negative outcomes.

One way of thinking about it is as follows: If I gave you an unscratched lottery ticket and only scanned the barcode to show you that you can cash it in for $1000 without playing the lottery, would you bother to scratch the rest of the ticket? What if I revealed the ticket was a loser? If you are thinking that you might bother to find out more about the winning ticket — maybe in the thrill of victory — and recoil from the losing one — maybe in the agony of defeat — then you would be exhibiting some form of the ostrich effect. It is called this because people, like the proverbial ostrich, are said to bury their heads in the sand when threatened with unfortunate financial news, whereas they lap up feel-good, and sometimes otherwise useless, information.

It makes intuitive sense to believe that decisions (and especially consequential financial decisions) made with incomplete information are going to be inferior. Therefore, the ostrich effect continues to attract special interest in the social sciences because people who neglect negative financial information cannot make decisions that are in their or

The ostrich effect has been implicated in everything from global financial crises to poverty traps.

anybody else’s best interest. Up to now, the ostrich effect has been implicated in everything from global financial crises, where market movers and makers are accused of ignoring the writing on the wall about imminent economic downturn, to poverty traps, where ordinary citizens forego monitoring and therefore managing their balances when they are on the wrong side of zero.

Illustrative examples from the field of behavioural economics provide a sense of the diverse ways the ostrich effect can materialize in practice. For example, shareholders monitor their investment portfolios less often in falling markets and do less research on stocks when notified that they have gone down in value. They will monitor their portfolios diligently during the weekend if they have increased in value during the week, even though the market is closed, their returns have not changed, and they cannot make any trades. Chequing account holders monitor their balances least often when they are most strapped for cash, which is ironic when you consider that these are precisely the individuals who would be most adversely impacted by costly overdrafts and unexpected fees.

After taking stock of these kinds of findings and ones like them in the domains of education, healthcare, social media, and politics, I and Dilip Soman of Behavioural Economics in Action at Rotman (BEAR) were motivated to develop a research program to investigate why this phenomenon happens and what can be done about it. After all, we figured that it would be beneficial if we could come up with a way to unbury people’s heads from the sand when confronted with useful, albeit painful, information. After digging a bit more deeply into it, we were pleased to learn that the ostrich effect is not exactly a universal phenomenon — pleased because it means that under at least some conditions people are willing to take in misfortunate news.

Three core findings ended up shaping the trajectory of the research program, and we believe they are central to any investigation on the ostrich effect: (1) Contrary to the predictions of the ostrich effect, people pay as much, if not more, attention to how much they can lose as compared to how much they can win when confronted with a choice between two risky lotteries, (2) people who score higher on self-efficacy — the belief that they are in control of their destinies — are more open to receiving negative outcome information; and (3) in healthcare, interventions focussed on boosting ego through self-affirmation and prompting patients to brainstorm reasons to pick up prospectively unpleasant medical test results for treatable diseases increases patients’ willingness to receive them. Together, the evidence suggests that people at least sometimes have a healthy appetite for bad news, but this appetite isn’t really whet until people believe that the information will allow them to do something about the situation.

We tested this proposition, partly because if true, it would reconcile contradictory findings in previous research, but also because there are some obvious practical implications. Our idea for testing it involved developing a computer-based, incentive-compatible gambling experiment. The experiment was basically a game in which players were shown a series of paired lotteries with a 50 per cent chance of winning or losing. Each lottery in the pair offered even odds of winning or losing small, randomly selected amounts of money. Note that there were four possible dollar numbers in a given choice — the amount they could win or lose for each of the two lotteries. These numbers were hidden behind tiles that are marked as containing either positive (‘win’) or negative (‘lose’) financial information. Before playing, participants needed to uncover three of the four amounts by clicking on respective tiles. The key was that each participant

Efforts to get people to pay more attention to negative financial information should focus on boosting their beliefs that they can do something about it.

could only ever find out about three of the four amounts, which forced them to decide whether they wanted full information about how much they could win or lose before a lottery was chosen, and allowed us figure out whether players were partial to learning more about how much they could win, lose, or neither.

Importantly, there are two versions of the game, and they are identical in every way except for one: In one version, players get to choose which gamble to take; in the other, this choice is made for them at random by the computer. We figured that if the ostrich effect is most robust when people (believe that they) are unable to act on negative financial information, then they should be more inclined toward uncovering negative information when they (as opposed to the computer) have the ability to make the choice.

The evidence from this experiment suggests that people are much more interested in taking on prospectively unpleasant financial news when they can do something about it (i.e., in the second version); otherwise, they generally seem to ignore it (i.e., in the first version). For those who are unable to control their financial situation, out of sight, out of mind, seems to be a fitting mantra. Participants were about ten percentage points more likely to prioritize loss information when they, as opposed to an algorithm, chose which gamble to play. In both versions, we observed an ostrich effect but, when participants were not in control, the effect got bigger over time. When we zoomed in on individual differences, we discovered that very few participants — about 1 in 20 — consistently wanted to find out about how much they stood to lose when they were not in control, whereas about 1 in 5 consistently wanted to know when they were in control.

Zooming back out, this set of findings helps us not only understand why some studies find ostrich effects and others do not, but they also provide deeper insight into why

the phenomenon occurs in the first place. Our current view is that the findings suggest that it is not only plausible, but likely, that investors, accountholders and others at risk of financial hardship do not believe they have viable options for dealing with looming misfortune. In line with the information avoidance literature in healthcare, we believe that efforts to get people to pay more attention to negative financial information should focus on boosting their beliefs that they can do something about it.

When the situation is beyond control, it is unclear whether any form of intervention aimed at having people take charge of their finances is warranted. In these cases, many may fritter away their time processing information that makes them happy and ignore information that does not. This is not necessarily a bad thing, so long as their beliefs in the amount of control that they have are aligned with reality. To abuse another familiar saying, there is no use in crying over spilled milk, but it may be worth shedding a tear or two if it can be cleaned up.

Matthew Hilchey is the Senior Manager of Behavioural Science at Scotiabank. He was a senior researcher at the Behavioural Economics in Action at Rotman (BEAR) research centre at the University of Toronto while this research was conducted.

BEHAVIOURAL SCIENCE IN THE WILD: THE DIGITAL NUDGE

Consumer choice, healthy behaviour and workplace productivity are just three areas where behavioural science and technology can combine to powerfully influence behaviour.

by Michael Sobolev

AS EACH DAY PASSES, more and more of our choices and behaviours are occurring online. We work, shop and communicate with the help of platforms like Amazon, Twitter and Zoom. Even daily behaviours such as sleeping, eating and exercising are performed with our constant companions by our sides — smartphones, wearables and smart-home devices. Most of these devices are powered by algorithms and recommendation systems that are designed to automate and augment our choices. And as a result, the digital world is rapidly transforming human behaviour as we know it.

Technologies that simplify consumer choice — for example, providing the ability to quickly search for products and recommendations — have bumped the concept of ‘choice architecture’ to the top of the agenda for many organizations. Choice architecture can be defined as ‘the design of different ways that a choice can be presented to decision-makers, and the impact of

that presentation on their decision-making’. A simple example is donor cards: Research shows that if the user is required to opt-in to become a donor, take-up rates are low; but if people are automatically signed on unless they opt-out (i.e., the default changes), take-up rates increase.

The current landscape of information technologies has significantly changed the myriad of choice architecture we face daily. The result: Choice architecture is both more digital and more intelligent than ever. And wherever there is choice architecture, there will be nudging.

The Realm of the Digital Nudge

A nudge can be defined as ‘any aspect of the choice architecture that alters people’s behaviour in a predictable way without removing access to other options or significantly changing economic incentives’. To count as a nudge, the intervention must

Online nudging practices are constantly being evaluated and adjusted, making them increasingly effective over time.

be easy and cheap to avoid. For example, social media platforms such as Facebook design choice architecture to encourage user engagement by curating information based on observed user preferences and past behaviour; online retailers such as Amazon nudge users to buy more by scaling up traditional choice architecture with the help of personalized recommendations; and smartphone devices and applications themselves — on which people spend hours a day — are designed precisely to reinforce mindless consumption and habitual checking. These and other online nudging practices are being rapidly evaluated and adjusted with the help of online experiments in the form of thousands of A/B field tests — making them increasingly effective over time.

As indicated, digital nudging applies the idea of choice architecture to a digital environment by passively changing user interfaces to guide choices. And it augments proven-to-beeffective behavioural nudges (e.g., feedback, reminders, planning prompts) by designing and implementing them in novel technologies such as apps and wearable devices. Digital nudging ultimately allows designers to select the ‘right’ nudge at the right time, for the right individual, by learning to adapt to that individual’s preferences and context. Two key principles of digital nudges are:

DESIGN IS IMPORTANT. Research in human-computer interaction (HCI) has extensively explored how to design technologies to be persuasive and support behaviour change, allowing for the emergence of common principles for good design, including selfmonitoring, simplification, personalization and rewards.

DATA IS ALSO IMPORTANT. The recommendation systems on Google, Amazon and Netflix are prime examples of AI technologies that simplify and guide choice by learning about user preferences from data on their past choices. Current and future advancements in machine learning and AI will allow digital nudging applications to be even more effective and precise.

The Tools of Choice Architecture

Defaults are one of the most powerful tools of choice architec-

ture — and digital technology is full of them, ranging from security and privacy settings on your devices to tipping defaults on Uber and product rankings on Amazon. Defaults are most useful to individuals when they fit their preferences. For example, after an Uber ride, nearly 60 per cent of people never tip, and only one per cent always tip. This observation warrants personalizing the default option so riders don’t need to make an active choice after every ride. Beyond personalization at the individual level, the characteristics of a specific ride could also be used to tailor the tipping suggestion for every ride (e.g., ‘late arrival equals less tip’). Almost every active nudge can be augmented with technology and data. Reminders can be personalized to an individual and optimized using timing; and feedback can be increasingly smart. A famous example is the fixed criterion of 10,000 steps a day, which serves as an imperfect reference point for optimal behaviour. Digital nudging allows for the design of personalized and adaptive goals based on an individual’s baseline, their ability to achieve the goal and their previous response to feedback.

Following are three domains where digital nudging can be the most powerful:

CONSUMER CHOICES. Digital technologies have transformed the choice architecture of everyday consumer decisions. Organizations use digital tools to personalize user interfaces and deploy recommendation systems to enhance engagement, influence choice and increase revenue. Digital nudging tools can also be just as effectively targeted to help consumers choose products, services and activities that maximize their own preference and utility (i.e., products that are healthy, sustainable and costeffective.)

Changing the underlying choice architecture is the standard approach to influencing consumer choice. For example, a search ranking of products can focus on nutrition for Consumer A, and cost for Consumer B. Another approach is to provide the consumer with tools to change the choice architecture in support of their preferences. For example, recommendations can

Examples of Digital Nudges

CONSUMER BEHAVIOUR

Smart Feedback

Smart Reminders

Technology

Defaults

Personalized spending limits based on budgeting

Reminders to shop healthy when opening a grocery shopping app

Personalized default tipping on Uber

HEALTH BEHAVIOUR

Personalized and adaptive feedback on step counts

Reminders to engage in physical activity based on calendar events

Walking as default in Google Maps and not public transport

PRODUCTIVITY AND WORK

Setting time limits on distracting mobile applications

Reminders about important unanswered e-mails

Opening productive software (e.g., writing) as opposed to e-mail as default

be designed to be more user-centric by introducing ‘preferenceelicitation mechanisms’ for users, which entail a short survey completed by the user. A third approach is to design third-party digital nudging applications that augment consumer choices across platforms, such as automatically providing price comparisons between online retail stores. The opportunities and design choices in digital consumer environments are unlimited.

HEALTHIER BEHAVIOUR. Changing one’s behaviour to sustain health is considered one of the most important challenges in the world today. Mobile technologies are often proposed as being part of the solution, as evidenced by the proliferation of mobile apps designed to help people become more physically active, eat healthier, stop smoking and drinking, and reduce stress. While health apps and devices are full of digital nudges, most have not been systematically tested in controlled settings. For instance, who says 10,000 steps per day should be everyone’s goal? To drive change, it is necessary to leverage emerging health technologies while accounting for basic elements of human behaviour and ever-developing findings in the science of health-related behaviour.

To date, research has targeted the nudging of daily health behaviours such as medication adherence, physical activity and smoking. Reminders can be tailored and optimized with information about a user’s past behaviour and current context. Nudging can also be designed to drive engagement and reduce attrition with evidence-based clinical products — a major challenge of current health technology. Moreover, the role of data and machine learning is rapidly growing to realize the vision of personalized precision-health technology.

PRODUCTIVITY AT WORK. Few would argue that the future of work

will rely on digital technology. Work technologies like Zoom for videoconferencing and Slack for communication are now familiar to most of us, but they could be redesigned to support goals such as increased productivity and reduced stress. And digital nudging tools can support these goals in the flow of daily work. Simple productivity advice can also be supported with digital nudging. For example, research indicates that individuals should start working on their most important projects at the beginning of the workday, rather than checking their e-mails first. Here, the gap between intention and behaviour can be mitigated with reminders and planning prompts. Digital nudging tools could support this behaviour even further by:

• Delivering reminders at the start of the workday based on location data;

• Defaulting operating systems to open the software with the most important project first; and

• Providing rewarding feedback when an individual follows through, and reminders when they do not.

One of the more novel and interesting applications of digital nudging involves the use of technology itself. Armed with the tools of behavioural science, choice architects are often responsible for creating engaging, addictive and distracting digital technologies ranging from smartphones to social media websites. Digital nudging can help individuals self-regulate and tackle their own habitual and mindless use of technology. Examples of such digital nudges include changing the colour of the screen to be less engaging and creating personalized and adaptive goals for distracting social media apps.

Designing Digital Nudges

Based on a synthesis of current research and the affordances of

technology, I propose three guidelines for the design of digital nudging solutions. All three can apply to the digitization of existing nudging practices as well as to the development of novel applications.

PRINCIPLE 1: EMBRACE TECHNOLOGY

Designers of digital nudges should choose a technology for their nudge based on its proximity to the targeted behaviour and the unique features of the technology. For example,

• Online shopping websites allow designers to easily track and nudge behaviour at the point of purchase using choice architecture tools;

• Smartphones are both personal and ubiquitous and allow one to design and deliver a large set of digital nudges via mobile apps; and

• Wearables allow for the collection of novel data (e.g., heart rate) and nudge individuals with novel feedback (i.e., haptic feedback).

A Laboratory of Behavioural Interventions by

PRINCIPLE 2: PUT DATA TO WORK

Designers of digital nudges should identify the necessary data to track behaviour and other contextual variables to optimize nudging techniques. For example,

• Reminders can be delivered to anyone via text message, but tailoring reminders with data about location requires mobile devices and applications;

• Planning prompts will be more effective when tailored with data from calendar events about other activities of daily life; and

• Adaptive feedback on physical activity requires sensors equipped to capture these activities, for example, in mobile phones or wearables.

PRINCIPLE 3: EXPERIMENT

Designers of digital nudges should leverage digital experimentation to rapidly optimize nudges and choice architecture. Digital

Nicola Lacetera and Mario Macis

Blood transfusions allow medical professionals to address a wide variety of emergencies, from trauma-induced blood loss to blood replacement during surgeries. Population aging and the increase in surgical procedures such as organ transplants are further increasing the demand for blood. The question is, What will influence human behaviour more: appealing to altruistic motives to the request or providing an economic incentive? We decided to find out.

Interventions targeting altruistic motives. According to the guidelines of the World Health Organization , the objective of national blood procurement and allocation systems is “to obtain all blood supplies from voluntary non-remunerated donors.” There are several ethical and behavioural foundations at the origin of this stance. The first is a strong reliance on the power of altruism as a motivator: People might want to help others because of a genuine desire to increase the other’s wellbeing, even at the expense of their own. In his 1759 book The Theory of Moral Sentiments , Adam Smith claims that this “virtue of beneficence” derives from natural human sympathy. Behavioural scientists call this tendency ‘warm glow’ or ‘impure altruism’. Another strong underlying source of altruistic behaviour is the desire to signal to others — or to oneself — that one is ‘a good person,’ and to improve one’s reputation in a community. Yet another source of altruism derives from reciprocal feelings, whereby a person behaves altruistically with the expectation that the recipient will reciprocate in the future, should the need arise.

Interventions applying economic incentives. The second key behavioural claim at the root of promoting purely altruistic donations is that material rewards or compensation might actually have perverse effects that do not correspond with the ‘rational agent’ view. According to this view, the higher the reward for an activity, the more an individual will make an effort in that task. In his 1970 book The Gift Relationship , British scholar Richard Titmuss claimed that material rewards might change a donor’s perception of the nature of their act from being altruistic to resembling a market exchange. This could reduce altruistic drive without a counterbalancing increase in material incentives — especially if the amount of the reward is small. Titmuss also argued that monetary incentives would attract donors with a higher risk of carrying infectious diseases. According to Titmuss’s hypothesis, rewards are especially appealing to low-income individuals who are, in turn, more likely to be affected by blood-transmissible diseases due to unhealthy lifestyles and risky behaviours.

The strong objections to providing any form of compensation to blood donors derive from multiple concerns. At the same time, both in the Western world and even more so in low- and middle-income countries, guaranteeing the availability of blood for transfusions is critical and has proven difficult. The seasonality of the blood supply compared to the constant need, moreover, adds to the difficulty of keeping an adequate inventory level. Even in contexts where the organization of procurement and allocation is advanced, relying on altruism is often insufficient to cover the demand for blood.

experimentation should allow for the following:

• The evaluation of a large set of possible nudges;

• Personalization of nudges based on individual preferences and adaptation to the changing contexts of daily life; and

• Deploying and evaluating the effects of digital nudges continuously.

In closing

Nudges in the realm of digital technology are not a panacea for influencing and improving human behaviour. For some choice settings, choice architecture and nudges can remain offline as they would not benefit from personalization. For others, technology might add unnecessary complexity and limit scalability. The science of digital nudging is still nascent and warrants some caution before it is applied ubiquitously to every choice environment. And importantly, the digital divide remains very real. Not everyone can benefit from digital technology equally, and

designers must be aware of this as consumer choices continue to shift to the digital realm.

This conundrum has stimulated efforts to devise incentive programs that would minimize potential ‘crowding out’ effects as well as addressing ethical concerns. In Italy, for example, employees have the right to a paid day-off work on the day they donate. In our own research we investigated the effectiveness of this policy and found that those who benefit from it do indeed donate blood more frequently. This implicit economic incentive, however, is expensive: We estimated that each additional unit of blood collected due to the ‘day-off incentive’ would cost about 400 Euros.

In the U.S., American Red Cross ( ARC ) blood drives sometimes offer rewards such as T-shirts and gift cards to people who donate. In research we did in collaboration with the ARC, we found that these rewards did increase donations, with rewards of higher economic value resulting in more additional donations than rewards of smaller value. However, we also found that part of the increase in donations at blood drives offering rewards came from a reduction in turnout at neighbouring blood drives that were not offering incentives. On the one hand, our findings highlight the importance of carefully distributing rewards across geographic areas to avoid competition between blood drives, and on the other, they imply that the use of rewards in periods with lower seasonal supply can be effective.

Our research indicates that even small economic rewards can increase donations in a cost-effective way. For example, the cost of one extra unit of blood collected using a $5 gift card as an incentive was $22. In recent years, the ARC has offered gift cards during periods of low supply — for example, in the winter. In an

Michael Sobolev is a Project Scientist at the Cancer Research Center for Health Equity at Cedars-Sinai Medical Center in Los Angeles. He holds a PhD in Behavioural Science from the Technion Israeli Institute of Technology and an MSc in Business and Organizational Behaviour from Tel Aviv University. This article has been adapted from his chapter in Behavioural Science in the Wild (Rotman-UTP Publishing, 2022), co-edited by the Rotman School’s Canada Research Chair in Behavioural Science and Economics, Dilip Soman

intervention conducted in collaboration with an Argentine blood bank, we found again that gift cards motivated additional donations, whereas ‘social-image stimuli’ (e.g., T-shirts with designs indicating the wearer was a blood donor) were not as effective.

Overall, the evidence demonstrates that certain economic incentives have positive consequences in terms of the quantity of blood collected. People respond in a rather standard way: higher-value rewards have bigger effects than lower-value rewards — even in the case of an archetypically altruistic activity like blood donation. The fact that several organizations allow and make broad use of (appropriately framed) economic incentive programs suggests that there are further opportunities out there to leverage motives beyond altruism — so long as it is done in a way that is socially and ethically acceptable.

Nicola Lacetera is Chief Scientist at the Behavioural Economics in Action Research institute (BEAR) at the Rotman School of Management and a Professor of Strategic Management in the Department of Management, University of Toronto Mississauga with a cross-appointment to the Rotman School. Mario Macis is a Professor of Economics at the Johns Hopkins Carey Business School and Affiliate Faculty at the Hopkins Berman Institute of Bioethics. This is an excerpt from their chapter in Behavioural Science in the Wild (Rotman-UTP Publishing, 2022.)

The Behavioural Science of Online Manipulation

Organizations have two choices: exploit human biases or ‘nudge for good’. The latter will shape the evolution of the digital world — and build fairer, better markets.

by Elisabeth Costa and David Halpern

NO ONE WOULD ARGUE that the Internet has transformed the way we live, work and play. Today you can make ‘friends’ around the world without ever saying hello; compare products from dozens of shops without leaving the house; or plan a date with a stranger without breaking a sweat. But while the benefits of life online are significant, so too are the economic and social costs.

By now, most readers are familiar with the term ‘nudge’; but there is a darker side to the evolution of online and digital markets: ‘Sludge’ and other techniques seek to harness our behavioural biases against us rather than for us.

Nobel Memorial Prize winner in Economics Richard Thaler, a long-standing adviser to the Behavioural Insights Team, often signs his books with the maxim ‘Nudge for good’. In this article we will explore some of the ways in which organizations might seek to enact this sentiment in the online world to shape its evolution —and build fairer, better online markets.

Exploiting Consumer Biases

In the same way that our behaviour in the offline world is sensitive to subtle cues from the environment, our behaviour online is shaped by the design and characteristics of the websites, platforms and apps we interact with. Nobel Memorial Prize winners Robert Shiller, George Akerlof and Thaler have written about how companies seek to manipulate and exploit our cognitive biases and psychological weaknesses. Two types of manipula-

tion identified by them are particularly common: The deliberate exploitation of information deficits and behavioural biases (‘phishing’); and the deliberate addition of frictions and hassles to make it harder to make good choices (‘sludge’). Let’s take a closer look at each.

PHISHING. Companies have long used techniques to target and exploit our psychological weaknesses. Shiller and Akerlof call this phishing. Offering a cash-back deal is a classic example. The vast majority of us erroneously predict that we will go to the effort of actually claiming the money. Our purchase decisions are more sensitive to the promise of the cash than to the effort required in actually claiming it, while redeeming the cash-back offer is driven more by the effort involved than the cash available. Sellers can experiment to find the optimum level of cash-back that will tempt you, but not be so large that you will actually complete the claim.

Today, a shopper may subscribe to Amazon Prime, believing that they are saving money via free shipping — when in reality they are just being tempted to buy more. In fact, the longer someone is a Prime member, the more they spend on the platform. In the past, we had no way of knowing whether we were fast enough to be in the ‘first 50 responders’ to claim a free gift or discount. But today we are likely to trust real-time updates, for example, from hotel-booking sites urging that there is ‘only

The intersection of data, imbalances of information and intelligent marketing opens up new opportunities to exploit our biases.

one room left at this price’ or ‘five other people are looking at this room right now’, especially if they have already shown you all the sold-out hotels in your initial search results. This is designed to harness our sense of scarcity and our desire to see ‘social proof’ that other people are making choices similar to ours. What they don’t tell you is that those five people aren’t necessarily looking at the same room as you.

The intersection of data, imbalances of information and intelligent marketing also opens up new opportunities to exploit our biases. Conspiracy theorists might claim that our devices are ‘listening’ to our conversations, when in fact the data we willingly share is more than enough to predict what we might buy, when, how much we are willing to pay and the flawed decisions we might make along the way. Further, the data we share online also allows companies to move away from the model of active, consumer-led searches and towards prompting us with targeted advertising and information at times when we are most vulnerable to being manipulated. For example, the rise of ‘femtech’ apps to track menstrual cycles and fertility has allowed businesses to collect large and robust data sets to produce tailored and timely marketing. More disturbingly, social media platforms’ ability to identify users’ emotional states based on their activity means that — in the alleged words of Facebook staff — advertisers now have the capacity to identify when teenagers feel insecure or need a ‘confidence boost’.

SLUDGE. The conscientious nudger seeks to design systems and processes to help people make better choices for themselves — for instance, by using defaults that make it easier to save, rank deals based on best value and tell us what other people like us are doing. Thaler coined the term ‘sludge’ to describe the deliberate use of these same tactics to thwart us acting in our best interests.

There is plenty of sludge offline — notably in administrative paperwork burdens — but online environments allow sludge to thrive. You can set up an online news subscription with one click, yet when you go to cancel it, you’ll find yourself in a complicated loop of online and offline hurdles: Calling a phone number during restricted business hours, filling out a detailed form and posting it to the company, or even being required to visit a physical store. These muddy frictions are deliberate. Companies know that they matter.

Online airline bookings are also fertile ground for sludge. An enticingly-low flight cost and large ‘book now’ button is often

followed by a series of pages slowly adding on additional choices, information and costs. When you get to reserving your seat, you discover that you can pay online or if you want to be randomly allocated a free seat, you’ll need to queue at the check-in desk on the day of the flight. All designed to sell you more and discourage you from choosing free or cheaper options.

Of course, friction can also be dialled-down or removed from processes to encourage behaviour that is self-defeating. You can be approved for a high-cost payday loan with just a few clicks, and shopping or gambling online at 3 a.m. is easy to start and difficult to stop. The practice on Netflix and other streaming platforms to automatically start the next episode in a series is a clear example of the power of removing friction, or flipping the default: People fail to press stop, and thereby watch another episode they might not have watched if they had to press play.

Morals, Ethics and Social Networks

It is undeniable that digital markets are changing the nature of the economy. Yet as Adam Smith exposed, there is more to a market than money. Markets are entwined with ‘moral sentiments’, social networks and a myriad of habits and customs that shape our lives and societies.

‘Social capital’ refers to our connections to others and the level of trust, informal rules and common understandings that facilitate communication and exchange within those networks. It helps information flow, lowers transaction costs and drives fairer exchange. Online markets — and social media in particular — have the potential to significantly affect the character and form of our social capital — for example, by creating ‘echo chambers’ that fool us into thinking our social and political ‘bubbles’ are representative.

Nudge co-author Cass Sunstein has written about a phenomenon that exacerbates these echo chambers: ‘asymmetrical updating’. This is a strong tendency to favour evidence that confirms our beliefs and to ignore or misread new evidence that does not. This tendency has been observed in beliefs about topics such as climate change, the death penalty, affirmative action, and sexism in male-dominated subjects and industries.

Social media helps asymmetrical updating thrive, since many now use it as a primary news source without noticing how their choices about who to follow have fundamentally altered the balance of information they receive. Website algorithms curate what we see and prompt us where to go next based on our past

usage. The powerful logic of ‘people who liked this also liked that’ takes us ever deeper into the bubble.

Algorithms can also create or destroy social capital more directly by drawing on personal data to exclude people from opportunities. For example, a babysitting app that uses Facebook news feeds to rate the suitability of babysitters on the basis of whether they’ve discussed drugs or seem to have a ‘bad attitude’ online.

Studies have already highlighted examples where online transactions and interactions are leading to new forms of discrimination and disadvantage. While platforms like eBay have persisted with pseudonyms, others have encouraged users to provide information about themselves to build trust and, over time, reputation. For example, Airbnb requires real names and it has been shown that guests with distinctively African American names are 16 per cent less likely to be accepted compared to identical guests with distinctively white names; and male guests in implied same-sex relationships were up to 30 per cent less likely than other couples to have their bookings accepted by Airbnb hosts. The discrimination was not in the form of outright rejections but more insidious, with hosts failing to respond at all.

But the impacts on social capital and cohesion are not all negative. Digital markets have also been able to supplement and enhance society’s stock of social capital. Platforms such as eBay have made it possible to trust relative strangers; LinkedIn can extend the ‘weak ties’ that play a key role in getting a job; Facebook has made it easy to stay in touch with old friends; and sites like TripAdvisor create a much wider network than word-ofmouth recommendations.

A related concern is how the evolution of online environments is affecting the character of our social relationships and communication with others. Most obviously, anonymized communication reduces the reputational costs usually associated with negative behaviour and exchanges. This can lead to more aggression (such as anonymized subjects administering more pain on others) and dishonesty, though it can also leave more room for self-expression and the ability (for some) to resist problematic social norms. There is also some evidence that the characteristics that people adopt in online avatars ‘leak’ into their offline behaviour, for good or bad.

A number of social media sites and online marketplaces are wrestling with this issue, and hate-speech laws pick up on some of it. But the majority of the time online incivility occupies a grey area between the law and public acceptability of hurtful com-

ments, inaccurate reviews, bullying, and shaming. A survey commissioned by Amnesty International in Australia found that nearly half of women aged 18 to 24 had experienced online harassment, including sexist abuse, ‘trolling’ (posting deliberately offensive or provocative content), threats of sexual violence, the posting of intimate pictures online without consent and ‘doxxing’ (the sharing of identifiable details). It is clear that a selfregulatory dynamic is not enough.

One of the knock-on consequences of all of this is widely thought — though not yet proven — to be on mental health. New evidence suggests two potential factors at play: negative feelings due to hurtful interactions or negative content, and substituting time away from activities that enhance well-being, such as offline socializing.

In one large-scale field experiment, researchers found that Facebook users who deactivated their accounts for four weeks spent less time online, reduced political polarization (although at the expense of factual news knowledge) and, most crucially, increased subjective well-being. This increase was ‘small but significant’, in particular on self-reported happiness, life satisfaction, depression and anxiety, and the size of the effect on overall subjective wellbeing was equivalent to about 25 to 40 per cent of the effect of interventions such as therapy.

Of particular concern is the impact of social media on the mental health of younger people. A 2017 survey of people aged 14 to 25 conducted by the Royal Society of Public Health comparing the five most popular social media platforms on a range of positive and negative mental health outcomes, found that YouTube was the most positive, for example, by raising awareness and self-expression, while Instagram was the most negative — via its impact on body image and fear of missing out.

These findings are not trivial. Young adults are currently estimated to be spending an average of four hours online every day, and this shift has coincided remarkably with a rise in mental health conditions in teenage girls, including depression, anxiety disorders and self-harm. While each individual case of mental illness is complex, there is increasing evidence that social media is at the very least associated with these issues. One study analyzed survey data from 10,904 14-year-olds in the U.K. to explore the relationship between social media use and depressive symptoms. It found, in particular for girls, that greater social media use was associated with online harass-

More focus needs to be placed on nudging and prodding companies to change their behaviour.

ment, poor sleep, low self-esteem, and poor body image, and that these were in turn linked to depressive symptoms.

What Can, And Should, We Do?

An effective response to these challenges must make good use of the entire regulatory toolkit and also build on it to introduce new ways of shaping markets. And not just government and regulators: industry has a key role to play. Following are a few principles to consider.

SMARTER DISCLOSURES. There are many opportunities to improve the information being provided to consumers online — from how their data is being collected, to whether algorithms are being used to make decisions, to a website’s terms and conditions (T&Cs) — and to vary when and how this information is presented. Where there is no clear regulatory power to compel companies to provide information in a particular format, these smarter disclosures can be encouraged through best practice guides, codes of conduct and reporting requirements.

Our colleagues concluded a series of online experiments testing ways of applying behavioural science to improve consumer comprehension of (and engagement with) online T&Cs and privacy policies. They found that telling customers how long a privacy policy normally takes to read increased open-rates by 105 per cent; and using a question-and-answer format to present key terms and illustrating them with explanatory icons increased understanding of T&Cs by more than 30 per cent.

EXHORTATION. A lot of policy, and certainly political activity, urges citizens or firms to do something differently. The U.K.’s Chief Medical Officer, for example, urged parents to ‘leave phones outside the bedroom at bedtime’ and enjoy ‘screen-free meal times’ with their families.

Exhortation is not a bad place to start when there are genuine concerns but also doubts about direct interventions and a desire to let people to make up their own minds. Yet in some areas we should be ready to move beyond general exhortation, and be more targeted in two ways: First, to focus on urging companies rather than consumers to change their behaviour, and second, to look to change the behaviour of specific companies as an example to industry.

Consumers have often been exhorted to be more ‘engaged’ in a range of markets and, in particular, to switch more often. In energy, insurance and banking markets, economists and some

regulators have bemoaned the passivity of consumers. ‘If only more would be rational and switch, then they would get better deals and the market would work’. The implication is that ‘there is nothing wrong with our model’ and that market failures are the ‘fault’ of consumers. We now know enough to be wary of this interpretation. It’s not just that consumers may have better things to do with their time than endlessly shop around. It’s also that markets are evolving to add subtle frictions, distractions and confusions that make it hard for consumers to switch — and very easy to stick.

More of our focus needs to be placed on nudging and prodding companies to change their behaviour to make it easier for consumers to identify better deals and switch. More fundamentally, government and regulators need to be guiding the evolution of markets — and sometimes directly intervening — to make sure that good companies and practices are the ones that are winning market share, and poor companies and practices are squeezed out.

CHOICE ARCHITECTURE. How overt should a choice be? When should it be made? To what extent should choices be different, or presented differently, for different groups? These are the types of questions involved in building powerful choice architecture.

One thing regulators and innovative companies can seek to do is to put as much control as possible back into the hands of the individual. This is not the same as just encouraging people to switch a product or service. Rather, it is about allowing consumers to be able to express their preferences and to modify their online experiences in line with those preferences.

Adding-in specific and obvious user controls, particularly on quasi-monopolistic platforms, could be a tool for regulators — and a market advantage for challengers. Just as now a user can easily adjust the font size on a screen to suit their eyesight, they might also be able to — or prompted to — adjust settings that:

• control the level, type or source of advertising (or other) material they are seeing (noting the potential commercial implications of this);

• control how their data is collected and shared; or

• choose or weight the criteria by which things are ranked and presented — for example, by independent retailer, female commentator, lowest total price, most reliable news source, or healthiest option.

Currently, it is hard to change privacy settings and other controls buried way down in settings menus and a key challenge is how to prompt users to engage with and adjust these settings and controls. Google Chrome content settings, for example, are hidden deep down in the ‘advanced’ settings, where many users are unlikely to find them. Sometimes this is done to add deliberate friction. But sometimes it’s done for good reason — to avoid filling screens with rarely used choices or because a choice that might seem important to one user may not be for another.

There may be a role for regulators here, in encouraging or requiring companies to make user settings more prominent, either during ongoing use or in the set-up phase. However, this activity is more likely to be driven by fostering the development of new intermediaries. For example, the MIT Media Lab’s Center for Civic Media was building a tool called Gobo from 2017 to 2020 to aggregate content from large platforms and then enable users to customize that content. It gives users a series of ‘sliders’ to curate what news they see and what is hidden. For example, you can express a preference to see wider news sources or more female commentators. Gobo and other intermediaries could offer more control and customization to users without regulators requiring large platforms to change their own user controls.

Another way of putting control back into the hands of users is through the use of prompts and reminders. These have been used to great effect in the offline world, to encourage people to switch their energy provider; attend, cancel or rebook their medical appointments; attend career appointments; pay their court fines; and save money or repay their credit cards. Online, they can be a powerful mechanism to elicit consumer preferences. This is especially the case if they contain a meaningful and ‘active’ choice — requirements to respond ‘yes’ or ‘no’ to a question before proceeding — offered at a salient moment.

An illustration of the power of an active choice is that researchers found that presenting consumers with an active choice on whether they would pick up their prescription from the pharmacy or have it home-delivered moved the take-up rate for home delivery from six per cent to 42 per cent. It was not that they couldn’t have chosen home-delivery before: it just was not the default option.

In closing

The time has come for industry and government to address the real and appropriate public concerns that exist around issues ranging from algorithmic bias, to disinformation, to the men-

• Where you are, accessed through your GPS location (depending on your device settings and permissions)

• Your phone number, even if you haven’t provided it in your profile

• Access to what you see through your camera when you’re using the app’s camera feature, regardless of whether you are recording

• Your mobile signal strength or internet connection speed

• Your contacts, who you’ve called and your SMS history (if you choose to sync, upload or import your phone contacts with the Instagram app, for example)

• How much battery you have left on your phone or laptop

• How much storage is available on your phone or laptop

• The other apps you have installed on your phone or laptop

• Information about other devices nearby or on your network

tal health of children and young people online. A sophisticated understanding of human behaviour — including active and constructive dialogue with the public — should be at the heart of designing successful policy solutions.

Our governments and regulators stand at the vibrant intersection between civil society and market functioning. How we respond to, and shape, the evolving character of the digital landscape is critical not just because it is pivotal to our economies, but because it is society and the human character itself that we are shaping.

Elisabeth Costa is a Senior Director at the U.K.-based Behavioural Insights Team, where she leads a portfolio of teams working on economic policy. David Halpern is Chief Executive of the Behavioural Insights Team, which he has led since its inception in 2010. He is the author of Inside the Nudge Unit: How Small Changes Can Make a Big Difference (Ebury Press, 2016).

Surprising Data Being Shared by Facebook and Instagram Users

Behaviourally Informed Cash Transfers to Address Homelessness

The average wait for someone in homelessness to get into housing was six months in Vancouver, but most cash recipients moved into housing within one month of receiving the cash transfer.

by Kenneth Ong, Daniel Daly-Grafstein and Jiaying Zhao

HOMELESSNESS IS A RAGING GLOBAL CRISIS. According to the World Economic Forum, about 150 million people worldwide experienced homelessness in 2021. Traditional solutions to homelessness typically involve providing emergency housing, social and health services, all of which are enormously costly. In particular, these services can cost up to $132,000 in Canada and $83,000 in the U.S. per person per year. What’s worse is that these solutions are not working very well, as evidenced by the surging homeless populations. Take Metro Vancouver for example; the homeless population has increased by 67 per cent from 2005 to 2020, while the entire population of Vancouver only increased by 23 per cent in the same period. We felt that there was a better approach to address homelessness.

Claire Williams, the co-founder of Foundations for Social Change, and one of us (Jiaying Zhao, who goes by JZ), teamed up in 2016 to create an innovative solution that gives cash, specifically $7,500, directly and unconditionally to individuals experiencing homelessness. This new solution was based on Claire and JZ’s insight that the root cause of homelessness is not just a lack of housing, but a lack of money. Indeed, the most common

reason for someone to become homeless is the inability to pay rent, which leads to eviction. So giving people cash seems like an obvious solution that allows them to get back into housing.

From a behavioural perspective, another key insight that Claire and JZ had was that the cash transfer is not just money, but a signal to the recipients that society believes in them. Unconditional cash transfer is a form of agentic help, unlike in-kind goods and services, which are paternalistic forms of help. In other words, rather than constraining the recipient, it empowers them. The cash transfer is unconditional, meaning that recipients can spend it on whatever they want, however they need to meet their needs. This gives them freedom, agency and choice to move beyond homelessness. This insight is based on research that JZ had conducted over the years with Princeton Professor Eldar Shafir and University of Chicago Professors Sendhil Mullainathan and Anuj Shah that found that poverty is as much a scarcity of cognitive resources as of money. The insight has led to the motto ‘believe in someone’ for Foundations for Social Change.

But giving people money is not as simple as handing out envelopes of cash to people on the street. Claire and JZ designed

and launched the New Leaf Project, the world’s first randomized controlled trial that distributes a one-time unconditional cash transfer of $7,500 to each of 50 individuals experiencing homelessness, with another 65 individuals in a ‘control condition’ in Vancouver, BC. The $7,500 equaled the annual income assistance in BC in 2016 and was about 60 per cent of the annual personal income of the individuals participating in the project.

To run the New Leaf Project, Claire and JZ galvanized a village of partner organizations and a team of interviewers, students and staff. To make sure that the cash transfer would not negatively impact participants’ existing or future benefits, they established an agreement with the BC Ministry of Social Development and Poverty Reduction to allow participants to keep the cash transfer while still being eligible for social benefits. They partnered with Vancity, a local credit union, to open free chequing accounts for participants to receive the cash transfer. They worked closely with several shelter organizations to deploy a team of interviewers to visit 22 shelters across the Lower Mainland to recruit participants on a weekly basis. It took the team two years to screen 733 people, 229 of whom passed the screening criteria, and 115 of whom were officially enrolled into the study (50 per cent of those who passed screening). The reason that only about half were enrolled was the practical difficulty of finding these people again after the initial screening, given they didn’t have a phone, an email or an address. Sometimes the interview-

ers had to drive around the city for weeks to find one participant. But once enrolled, the interviewers got all the contact information from the participants (including where they would like to hang out) and closely followed the participants for one year to track the impact of the cash transfer on their lives.

Surprises on What Worked

Before the trial was launched, Claire, JZ and the research team made a number of predictions about subjective well-being and cognitive outcomes at one month after the cash transfer and preregistered these predictions on the Open Science Framework. Much to their surprise, none of these predictions panned out. In contrast, some of the best-case scenarios — outcomes that were beyond their dreams — turned out to be true. One surprise was how fast people moved into stable housing after getting the cash transfer. The average wait for someone in homelessness to get into housing was six months in Vancouver, but most cash recipients moved into housing within one month of receiving the cash transfer. This goes to show how resourceful and prepared they were to get back to stability. In particular, many participants told us that they knew exactly where to find housing; all they needed was money to get into housing.

Over one year, cash recipients spent 99 fewer days in homelessness on average compared to an otherwise similar group that did not get the cash transfer. Less time in homelessness also meant less time on the street, in the shelter system, in the hospital, or in jail. This led to a cost savings of $8,277 per person per year. The cash transfer was $7,500, so there was a net savings of $777 per person a year, which was a big surprise to the team. The net savings also meant that the cash transfer could save the government and taxpayers money on homelessness, when compared to status quo approaches.

Another surprise was how well the cash recipients managed their money. Over one year, they kept $1,160 more in savings, challenging the assumption that people in homelessness will spend any money given to them all at once. Cash recipients also increased monthly spending by $429 more than control participants. Specifically, they spent more on rent, food, transit, and durable goods (e.g., furniture, car). Mothers who received the cash transfer spent more on food and clothing for their children. Importantly, spending on temptation goods (i.e., alcohol, drugs, cigarettes) was not different between cash recipients and control participants. This finding challenges the (unfortunately)

A cash recipient from the New Leaf Project.

Our finding is consistent with dozens of other studies on cash transfers to low-income individuals around the world that consistently show no change in spending on temptation goods after receiving the cash transfer.

pervasive assumption that people experiencing homelessness squander money on alcohol and drugs. Our finding is consistent with dozens of other studies on cash transfers to low-income individuals around the world that consistently show no change in spending on temptation goods after receiving the cash transfer.

Other surprises included increased food security and positive affect at one month after the cash transfer and improved cognitive function and reduced substance use severity at three months after the transfer, but these effects were transient.

Surprises on What Did Not Work

While the team was delighted at these outcomes, there was a running list of things that did not work as well as expected. As mentioned earlier, subjective well-being and cognitive outcomes did not improve at one month after the cash transfer. One reason was that it was very difficult to reach cash recipients one month after the transfer. This is because in the month after the cash transfer, most cash recipients were busy finding housing, getting furniture, and basically re-organizing their lives, so understandably the last thing they wanted to do was meet our interviewers and do a survey. Some cash recipients moved away to other provinces to reunite with their families. Social connectedness also decreased at nine months for cash recipients, presumably because they were in a new community with fewer friends. Two cash recipients voluntarily gave the entire cash transfer away to their family or friends.

One requirement from the UBC Behavioural Research Ethics Board was to provide non-cash support to participants to mitigate the potential risks of the cash transfer, so the team provided a workshop and coaching to a subset of cash recipients and control participants. The workshop was designed based on the best evidence from behavioural science, consisting of a onehour session every three months to each participant for one year, where the participant completed self-affirmation, goal setting and plan making exercises to help them brainstorm strategies to regain stability in their lives. Coaching consisted of three 45-minute phone calls per month for six months with a certified coach trained to help participants learn from their own experiences to increase self-efficacy in developing life skills and strategies to achieve their goals.

Despite our best attempt at providing these non-cash supports, neither the workshop nor coaching had a meaningful impact on the participants. One reason was compliance. Most par-

ticipants completed only a few workshops over one year, and one or two calls with their coach. Another reason was a mismatch between the supports and the participants’ needs. The support provided by the workshop and coaching was aspirational, designed to clarify people’s goals and plans and boost their self-efficacy. But what the participants needed was instrumental support, like getting their ID replaced, completing their resume to apply for jobs or school, and finding affordable housing. These instrumental needs could not be easily met by completing the workshop or coaching.

This finding suggests that these well-established behavioural interventions which are typically done with middle- or higher-income individuals cannot readily benefit individuals experiencing homelessness. In fact, this is consistent with other studies that show limited success of interventions like reminders in improving behavioural outcomes of low-income populations. The barriers that low-income individuals face cannot be easily fixed by light-touch behavioural interventions. Sometimes the best intervention to fight poverty is just cash.

Behavioural Insights For Success

Although the primary intervention in the New Leaf Project was the cash transfer, the team used a number of behavioural insights to improve the chances of success of the cash transfer intervention. The team implemented a two-week delay period between the cash news and the cash transfer. This delay period was necessary because participants needed time (two weeks on average) to open a chequing account at Vancity, and it was intentional because the delay gave participants time to think about how to use the money, which may have helped reduce impulsive spending.

Another insight was mental accounting, specifically the psychology of how recipients were encouraged to think about the cash that they were going to receive. When the cash participants were informed about the cash transfer after enrolling in the study, the interviewers encouraged participants to have two mental accounts for the cash transfer, one growth account and one flexible account. The growth account was for spending on essentials, personal development and investment. The flexible account was for spending on emergencies or things that were discretionary. Both accounts were in the participants’ minds and not real bank accounts. Mental accounting gave participants an opportunity to anticipate unexpected issues in advance and prepare strategies to cope.

Well-established behavioural interventions which are typically done with middle- or higher-income individuals cannot readily benefit individuals experiencing homelessness.

The participants in the project had already faced significant traumas and marginalization. When conducting a study with this population, the research team took extra care to avoid further traumatization or marginalization. During recruitment, the screening process involved participants answering questions about their experiences of homelessness. However, some participants did not think that they were homeless despite not having stable housing (e.g., living in their car, or couch surfing). Also, the label ‘homeless’ could trigger harmful stigma and stereotypes in even the participants themselves. Thus, we found that a better way to identify people without stable housing was to ask about their current living situation, rather than mentioning the term ‘homeless’.

Since the control participants did not receive the cash transfer, extra care was taken during debriefing when they were informed about the cash condition and how they ended up in the control condition. While a majority of the control participants expressed gratitude and understanding of the project, a few control participants initially expressed frustration upon learning that they were not in the cash condition. Due to previous trauma and abuse, they might have automatically thought that they were not worthy or deserving of the cash, an assumption that the interviewers needed to address immediately. Specifically, the interviewers explained that the assignment to the cash or control condition was completely random, based on a coin flip, and no personal information was taken into account at the random assignment. Importantly, at the end of the project, we re-screened all the control participants we could reach and gave them the same cash transfer if they screened in. This was to ensure that all participants had the same opportunity to receive the cash transfer by the end of the project.

Public Mistrust Toward People Experiencing Homelessness

Demonstrating the positive impact of the cash transfer was only half the battle. The other half was to convince the public of the value of the cash transfer so they could support such a policy down the road. Over the last several years, Claire and JZ have been in close communication with a dozen elected officials at the municipal, provincial or state, and federal governments in Canada and U.S. about the New Leaf Project. When the two presented the findings of the cash transfer experiment to policymakers, a common response from the policymakers is that they like the results of the study and see the need to increase cash support

to people in homelessness, but they are uncertain about how the public would perceive and respond to a potential cash transfer policy. In science-based policy, obtaining scientific evidence is perhaps the easier part of the puzzle. The harder part is convincing the public and policymakers of the value of policy change.

This prompted the team to conduct two additional studies to (1) understand the public perception of individuals experiencing homelessness, in particular, the perception of their ability to manage money and (2) identify strategies that would improve public support for a cash transfer policy as a constructive way to deal with the public perception.

When Claire and JZ discuss the cash transfer experiment with a public audience (e.g., community members), an overwhelming number of people say that they do not trust cash recipients to manage their money well. They also often say that the cash recipients will spend the transfer on alcohol and drugs. Individuals experiencing homelessness tend to be dehumanized and perceived as less competent. This bias can favor the provision of paternalistic forms of help over more agentic forms, presenting a barrier to the cash transfer policy.

To quantify this mistrust, we conducted another study that asked the public to make predictions about how participants in the New Leaf Project would spend the $7,500 cash transfer. People predicted that the cash recipients would spend on average $329 per month on temptation goods (alcohol, drugs, cigarettes) when the recipients were described as being homeless, whereas they would spend $182 per month if they were not homeless. People’s mistrust in homeless individuals led them to predict that these individuals would spend 82 per cent more on temptation goods. However, in reality, cash recipients in the New Leaf Project only spent around $120 per month on temptation goods, only 36 per cent of what the public had predicted them to spend.

Effective Messaging to Increase Policy Support

To combat the public’s mistrust, we conducted a third study to test different messaging that would increase public support for a future cash transfer policy. To make the cash transfer more palatable to the public, the team tried two types of messaging: a counter-stereotype message to describe that the cash recipients in the New Leaf Project did not spend more on temptation goods and instead increased their spending on rent, food and clothing, and a utilitarian message to describe that the cash recipients in the New Leaf Project reduced their reliance on social services

and saved more money than the cash transfer itself, producing net savings for society. Compared to a no-messaging condition, either the counter-stereotype message or the utilitarian message increased public support for the cash transfer policy by about 10 per cent. This suggests that public support can be strengthened by counteracting the stereotype of homeless individuals, or by providing utilitarian evidence that the cash transfer generated a positive impact on the recipients and net savings for society. These two messages can be used in future campaigns for cashbased homelessness reduction policies.

Overall, the New Leaf Project is a success. But it’s only a start. To replicate and expand on this project, Foundations for Social Change and JZ are in the middle of an expansion project designed to provide more individuals experiencing homelessness with a larger amount of cash. They have also consulted for a dozen organizations around the world to pilot cash transfer

trials to reduce homelessness. With the high inflation rates and massive displacement from climate change (e.g., wildfires, flooding), the need to raise the income floor of the most marginalized has never been greater.

Kenneth Ong is a medical student at the University of British Columbia and a project manager at Foundations for Social Change. Daniel Daly-Grafstein is a PhD candidate in Statistics at the University of British Columbia. Jiaying Zhao is an Associate Professor in Psychology and Sustainability at the University of British Columbia.

LEANING IN OR NOT LEANING OUT?

Narrowing the Gender Promotion Gap with Choice Architecture

Q & A with JOYCE HE

Interview by Oceana Ding

One of many employment-related gender gaps is the ‘gender promotion gap’. Can you describe what this is?

The gender promotion gap is the term used to describe the phenomenon whereby women are less likely to be promoted than men, even after accounting for any possible differences in ability and performance. Put differently, the gender promotion gap happens when — all else held constant — the likelihood of successful promotion for women is lower than that for men.

What does the literature say about why there is a gender promotion gap?

There are many contributing factors to why this gap happens. One possible factor that we investigate in our study is driven by two facts. First, many organizations usually require individuals to nominate themselves and apply to be considered for promotion. Second, women are less likely to opt-in to competition. As a consequence, the existing process for promotion at many organizations is biased against women simply by reducing the pool of women who are being considered for promotion. In our study, we wanted to investigate whether removing the self-nomination requirement could nudge women to compete for promotions.

What are the consequences for businesses when women don’t opt-in to the competition?

When women don’t opt-in to being promoted, businesses are unable to consider the full pool of eligible and qualified talent when filling open roles. Over time, both within an organization and across organizations, there can be more occupational segregation, with greater numbers of men employed in senior management roles and more competitive fields than women. Furthermore, the gender promotion gap can also factor into gender differences in salary progression, leading to other employmentrelated gender gaps like the gender pay gap.

In the first part of your study, you investigated how likely it would be for women to be promoted compared to men under an opt-in and opt-out competition framing. Can you explain the difference between these framings and what your main findings were?

In my research with Rotman Professors Sonia Kang and Nico Lacetera, we first thought about how the process of promotion actually occurs, starting even before the time that the formal assessment of the candidate is done. Reflecting how promotion

Thoughtful, behaviourally-informed structural changes can help an organization make great strides in becoming more diverse and inclusive.

traditionally happens in the real world, an opt-in system is one in which employees must opt-in, or make an active choice to be considered for promotion. To put it another way, the default is to not be considered. In contrast, an opt-out system is one in which employees, as long as they fulfil certain performance criteria, are automatically considered for promotion but can choose to optout. The default here is to be considered for promotion.

Under the opt-in promotion scheme, we found that women were less likely to opt into competition, which is a finding in line with the existing literature. Under the opt-out promotion scheme, we found that approximately the same number of men and women participated in competition for promotion. So when the default was to be included in the up-for-promotion talent pool, women were just as likely to participate in the competition for promotion as men.

Did nudging women to compete have any consequences on their performance or well-being?

We wanted to investigate whether the opt-out framing would lead women to experience any unintended negative consequences, like poorer performance or increased anxiety, but found that there were none.

In one of the study’s conditions, we actually found a benefit — women were more likely to maximize their performance with the competition framing they chose.

You also looked at whether an opt-out promotion scheme would make employers less likely to promote a woman compared to in an opt-in promotion scheme. How does the optout framing impact promotion decisions made by employers? Under an opt-out promotion scheme, where all eligible employees are automatically considered for promotion, there is the possibility that employers might think women desire the promotion less and therefore be less likely to promote them. In ongoing research, we find that changing the system of the promotion scheme from opt-in to opt-out does not affect how likely evaluators are to promote women. They are just as likely to promote women in the opt-out promotion scheme as in the opt-in promotion scheme.

Are women less likely to be competitive?

Our research suggests that it depends on the context. Under the opt-out frame (when competition was the default), women were more likely to choose to compete than under the opt-in frame (when non-competition was the default). We conducted a follow-up study to investigate the question of why this was. Compared to the opt-in framing, participants in the opt-out framing expected more participants of their own gender to choose to

compete as well. Employees under this framing might be encouraged to compete because the opt-out framing suggests that competing is the norm. So women respond to contextual or situational factors and can be just as likely to compete as men.

Where else has the opt-out framing been employed?

Opt-out framing has been used in a number of other domains, notably in organ donation registration and retirement savings programs. This means that in these programs, organ donation registrations will occur automatically, and employees will be enrolled by default in their company’s retirement savings plan unless they opt-out.

What are some popular interventions that have been used to reduce the gender gap, and how effective have these interventions been?

Some of the more common interventions are diversity training and coaching women to ‘lean in’ by being more active and visible at work. Businesses have invested millions of dollars into unconscious bias training and coaching programs, but it is unclear whether these types of programs are effective. In particular, unconscious bias training has shown to have mixed effects on attitudes and little to no effect on behaviour. Women can also face backlash from others when they adopt more stereotypically masculine behaviours recommended to them in courses, like acting more assertively. In addition, coaching women to ‘lean in’ and behave in different ways places both the cause and the responsibility of closing of the gender gap on individual women.

What type of intervention do you suggest, and what would you like to see happen next in organizations?

I would like to see us reorient toward fixing the system rather than fixing individuals. Opt-in promotion schemes favour those employees who are more comfortable with competition and are more likely to opt into it — more often men than women. There could be a business case for making structural changes as well — instead of investing large amounts of money into diversity and ‘lean in’ training, organizations can consider making small changes to the choice architecture of their practices. Thoughtful, behaviourally-informed structural changes can help an organization make great strides in becoming more diverse and inclusive.

Joyce He (BSc 2016, Rotman PhD 2021) is an Assistant Professor of Management and Organization and Behavioral Decision Making at the Anderson School of Management, UCLA.

Designing Incentives to Improve Outcomes

INCENTIVES ARE COMMONLY used to motivate behaviour. The basic idea is simple — if a deal is not sweet enough to mobilize people, make it sweeter! When children are not studying hard enough despite the long-term benefits of academic achievements, parents promise them gifts for good grades. Likewise, when some organizations struggle to mobilize employees’ engagement in ESG initiatives (e.g., getting them to reduce energy use at the workplace), they may provide employees short-term incentives for meeting ESG goals. However, applied behavioural scientists are cautious about using short-term incentives to motivate behaviour, worrying that such incentives may crowd out intrinsic motivation for the behaviour. For example, while a cash reward may enthuse employees to look for ways to save energy, it may have the undesirable effect of directing the employees’ attention to the reward itself rather than the sustainability benefits of energy saving. A ‘crowd-out effect’ may result — the employees become less keen on conservation once the reward is removed. Concerns about crowd-out have ignited a debate about whether extrinsic rewards should be used to motivate behaviours, limiting their use in behavioural change.

We take a different perspective in our research. Rather than discussing when and why incentives work (and don’t work) to change behaviours, we study how to design incentives to produce more sustainable behavioural changes. We studied this question in the context of non-mandatory education. Non-mandatory education is an important part of adult continuing education that upgrades the competency of the workforce for organizations and society. However, procrastination and present bias may impede the inception of such self-improvement behaviour.

In a typical study of our research, we offered learner participants an incentive to take a course (that usually at-

tracts a fee) for free and examined whether and how the design of this incentive changed learning behaviours. We assigned participants to receive the incentive in one of two ways. First, we offered a complete waiver of the fee. Second, we charged the full price upfront but offered a complete refund upon completion of or withdrawal from the course. From the perspective of course uptake, waivers would seem to be more effective because they are a more attractive option for learners. However, we found that a refund is superior to a waiver in instilling a more lasting motivation to learn. This effect is due to a more conscious consideration of the inherent benefits of the learning activity by the learner. In particular, the learners under the refund arrangement thought more about the benefits of learning to a greater extent than those who received a waiver arrangement. Consequently, the former showed greater persistence in learning and a higher likelihood of learning in the longer term. To understand why, let us take a deeper dive into the motivations that drive discretionary learning.

What Gives People Motivation to Do the Hard Work?

Discretionary learning, like many other positive behaviours, is extra hard work that an individual needs to engage in. They would not engage in expending extra effort unless they evaluate the pros and cons of the behaviours and decide that the benefits outweigh the costs. The provision of short-term incentives works on lowering the costs side of this comparison. However, research shows that incentives can reduce a consumer’s motivation to consider important product-related information. Free offers can drive a consumer’s attention to the promotion itself and away from other factors such as information about the product quality. As such, while free offers can attract initial product adoption, they may reduce

Catherine

When incentives are used to nudge positive behaviours, would they direct attention away from the inherent values of such behaviours?

customer repurchase intention in the long run. Put differently, when consumers are offered an enticing incentive like a free offer, they become more likely to think ‘I’m getting this product because it is free!’ instead of ‘I’m getting the product because it brings me benefits’.

In our research, we asked: When incentives are used to nudge positive behaviours, would they direct attention away from the inherent values of such behaviours? If they do, are there ways to boost attention towards the benefits of the incentivized behaviours? Further, would shifting people’s attention back from the incentive to the benefits of the behaviour encourage commitment?

Designing Incentives to Preserve Motivations

We picked two closely comparable forms of incentives — fee waiver and refund — and studied their different effects on motivations. As we will discuss later, the comparison provided meaningful insights into how people react to different designs of incentives. If people simply consider the nominal value of an incentive, whether it takes the form of a refund or a waiver should not matter. However, we expected waivers and refunds to differentially influence the level of thinking and elaboration that learners would have towards the benefits of learning.

Our work is built around the principle of mental accounting, originally developed by Nobel Memorial Prize winner Richard Thaler and extended by MIT Professor Drazen Prelec and Carnegie Mellon University Professor George Loewenstein. According to their formulation, when individuals pay for something, they might create a mental account that remains negative until they receive benefits from the purchase. Individuals are reluctant to close the account with a loss, so they actively track costs and seek benefits so as to close the account with a net positive balance. This theory applies to our context, where learners who have made an upfront (refundable) payment for a course will be driven by a stronger need to account for the upfront payment and deliberate more on the personal benefits of the incentivized course (e.g., personal interest, career benefits), compared to those who have no specific payment to account for. These benefit-centric thoughts will draw more attention to the inherent value of the incentivized learning activity, thereby

laying a foundation for greater persistence towards the learning activity.

In one of our studies, a number of undergraduate students from a university in Hong Kong were asked to consider using online English learning materials for self-improvement. These materials were drawn from a book entitled The Elements of Style. One group of students were told that the fee for these materials was HK$10 (US$1.3), but since the University had already made a payment, the fee was waived. A second group paid HK$10 upfront, knowing that the fee would be reimbursed at the end of the experiment. A control group was given no price information. In the experiment, the learning materials for each English word were displayed on a separate webpage; the students’ reading duration for each page as well as the number of pages read were tracked. The students were also given the option to return to the website and reread the materials after the experiment on their own time.

We found that those who paid upfront (and later got a refund) read more pages and spent more time reading the pages than the control group, but those who received a fee waiver showed no difference. Moreover, more upfront payers returned to the site post-experiment than those in the control or waiver group.

In two subsequent studies, we gained further insights into how incentive design changes what individuals consider when they think about taking up incentivized learning opportunities. Those who received a waiver thought more about the incentive and less about the pros and cons of learning. In contrast, those who were offered a refund arrangement thought more about learning. Of course, simply thinking more does not directly imply a higher intention to take the course; the final decision is influenced by the content of the thoughts. Nevertheless, our findings show that waivers and refunds shape thought and deliberation differently, with refunds promoting thoughts on the inherent nature of (incentivized) learning and waivers limiting such thoughts.

Adding Sludge to the Incentive

While our proposed solution to boost commitment in incentivized behaviours via a simple tweak to the design of

The act of expending a cost, even when it is eventually refunded, increases commitment and improves outcomes.

the same extrinsic incentives seems straightforward, in practice there might be situations that render it impractical to implement. For courses with high fees, the idea of making an initial payment, even if refundable, may deter learners. Refunds worked because the upfront monetary costs prompt a benefits assessment. Can this vital role of monetary cost be performed by other forms of costs, such as time and effort required for filling out a lengthy application? Although often viewed as undesirable, such sludge (in the right dose) might stimulate accounting for time spent and motivate attention to learning benefits.

In our final study, online participants from the U.K. contemplated a £490 communication skills training program. As in prior studies, certain participants were offered a tuition fee refund, others a waiver. The signup process for this incentive varied. Some were merely asked to provide an email (the short process), while others were asked to fill a lengthy form (the long process). We found that among participants who went through the short process, those offered a refund were more likely to base their decision on course benefits than those given a waiver. This was similar to our earlier findings. However, among those who went through the long process, decisions based on course merits were similarly likely regardless of receiving a refund or a waiver. Perhaps the effort cost associated with the long process promoted people to think more about how these costs, in addition to the monetary costs, could be recovered.

Implications

Taken together, our findings paint a consistent picture of how financial incentives can be used to promote long-term learning motivation: The key is to implement it in a way that strengthens learners’ focus on the benefits of learning, even when enticed by an incentive. We found that a refund outperformed a waiver in this regard: Learners offered a refund were less likely to drop out of the learning activity, spent more time reading the learning materials and were more likely to return to the activity. A small sludge together with a waiver could be an option to replace a refund arrangement to instill deliberation that could potentially sustain commitment in learning.

The implications of our research can potentially extend beyond the domain of learning and be applied to a broader category of positive behaviours. Undoubtedly, incentives can be a powerful motivator and if we begin from this premise, then the question is not whether incentives should be used, but rather how to design incentives that also help develop intrinsic motivations. In this respect, the key design consideration is to preserve thoughts on the inherent benefits of the incentivized behaviour to promote lasting behavioural change. The findings presented here pertain specifically to learning; it is vital to remember that results may vary when applying these same theories to other behaviours. We recommend conducting in-situ tests to readers considering implementing these incentives in different behaviour change contexts. In instances where you are employing sludge — obstacles like effort and time costs — it is crucial to find the right balance. Too much sludge can discourage the positive behaviours you are aiming to promote. Finding that perfect middle ground — a spot where the lure of the incentive offsets any reluctance caused by the sludge — might be challenging, which is why we recommend on-site testing.

We specifically studied the impact of sludge (time and effort costs) because it is a type of cost. In practice, there might be other ways to encourage commitment. For example, co-creating value of the encouraged behaviour and thereby highlighting its benefits could be an alternative approach. Our key point is simple — the act of expending a (monetary or non-monetary) cost, even when it is eventually refunded, increases commitment and improves outcomes!

Catherine Yeung is an Associate Professor of Marketing at the Chinese University of Hong Kong. Yih Hwai Lee is an Associate Professor of Marketing at the National University of Singapore (NUS) Business School.

The Elements of Context

WHILE THE FIELD of applied behavioural science has had much success in helping solve business and societal problems, there has been a recent wave of books and papers reflecting on the state of the field and advocating for a more nuanced approach to practicing it. Recent books like Behavioral Science in the Wild and The Voltage Effect identify a ‘portability problem’ — the idea that findings do not always translate or scale well across domains, or from smaller well-controlled laboratories and pilots into regional or national rollouts. A recent book What Works, What Doesn’t (and When) is replete with cases showing that sometimes, rolling out an intervention that has been successful in one setting produces no effects whatsoever in a different setting, while in other instances there is a fair degree of heterogeneity. In particular, the intervention works for a subset of the population but not for another subgroup, resulting in weak overall effects.

These translation and scaling challenges occur for a variety of reasons. Practitioners of applied behavioural science have their own hypotheses regarding why the results are not perfectly portable from one context to another. It has long been documented that human decision making is a function of context. Things in the environment that should not impact decision making often do. In his book Misbehaving, Nobel Memorial Prize winner Richard Thaler referred to these as ‘Supposedly Irrelevant Factors’ (SIFs). When we take a finding that has worked well in some contexts and apply it in others, it is very likely that these ‘contextual factors’ (CFs) will be different across applications. Hence, the effectiveness of

the intervention will change. As a general principle, when an intervention works in one setting but not in another, the difference can be attributed to a difference in context. Moreover, it is also fair to say that interventions might translate and scale well if the context remains the same across multiple applications. But what exactly is context?

In Search of the Elements of Context

While a number of researchers have attempted to document the changes in the effectiveness of interventions as a function of a specific contextual variable (for example, the medium in which an intervention is delivered or the exact design of communication), there has been no systematic overview of what exactly the elements of context are. When we make statements such as similarity of context or differences in contexts, could we think about developing a checklist of CFs that an applied behavioural scientist could use in assessing whether an intervention might be portable across two settings? This would also allow the practitioner to determine when to test extensively before transplanting an intervention and when a test might be less critical. Our team set out to do exactly this in the context of seven popular interventions (Box 1).

Focusing on these interventions, we used a threephased approach to generate a scorecard identifying the elements of context. We built on work by John A. List, the author of The Voltage Effect, suggesting that translation challenges happen because there are differences in two kinds of

FACULTY FOCUS Selina Yang, Meghan Yeung, Nathaniel Barr, Chang-Yuan Lee, Wardah Malik, Nina Mažar, Dilip Soman and David Thomson

BOX 1: Seven Popular Interventions

DEFAULTS

SIMPLIFICATION

SOCIAL NORMS

REMINDERS

IMPLEMENTATION INTENTION PROMPTS

FRESH STARTS

MATCHING

Changing the system so that outcomes change if people choose not to make any active choices (e.g., opt-in vs. opt-out processes)

Simplify complex information by any one (or more) of: a) reducing volume, b) chunking, c) presenting visuals or presenting the same information visually (e.g., in a flowchart)

Presenting information about any of a) the majority of people, b) the typical person, or c) ‘people like you’

Sending reminders to undertake an action via letters, emails, or text messaging

Getting people to make plans about how, when, and where they would get things done

Positioning an event at the beginning of a time period (e.g., new week / month, after a major life event) to increase motivation by generating a ‘fresh start’

Encouraging financial contributions (e.g., to pensions, savings, or charity) by offering to add a percentage from another source

variables — differences in the situation in which the intervention was delivered and received and differences in the recipient population of the intervention. We first combed through past research — published and unpublished — to identify variables that had previously been shown to weaken or strengthen the effect of interventions. We next surveyed several behavioural scientists from academia and practice and asked them to generate hypotheses about what elements of context matter. Finally, we conducted a workshop to analyze the qualitative and quantitative data and generate a list of the elements of context. We also sorted the comprehensive list of elements into seven dimensions of the situation (Box 2) and four dimensions of the target population (Box 3). Furthermore, we categorized the dimensions of situation into two subgroups — elements arising from the design of the content of the intervention and dimensions arising from the implementation (where the designer might have little control).

Collectively, Boxes 2 and 3 represent our framework that codifies differences in the situation and the target population. Our goal was modest — we wanted to sensitize practitioners to the portability problem and to caution them against simply transplanting an intervention that had been successful elsewhere into their own setting. Furthermore, our goal was to add nuance to the concept of contextual differences by decomposing a vaguely defined construct into specific dimensions. Given the lack of prior research on the relative role that each of these dimensions plays on the ef-

ficacy of interventions, our framework does not allow us to identify the relative importance of each of these dimensions, nor indeed whether different dimensions matter for different interventions.

Instead, our goal is to be as comprehensive and identify as many elements of context as we could in order to offer the practitioner a laundry list to make judgments about.

Beware of Nudgestore Shopping

It is normal for any practitioner to want to use an intervention that has been successfully deployed elsewhere. Imitation is indeed the best form of flattery. However, we caution the practitioner from blindly imitating a previously successful intervention. As one of us (Soman) along with Tanjim Hossain at the University of Toronto wrote elsewhere, the act of simply mimicking a successful intervention from a proverbial ‘Nudgestore’ can be problematic without a thorough adaptation to suit any differences in context.

We would therefore encourage the practitioner to ask two questions: First, is the situation in which they will use the intervention significantly different from the setting of the original intervention? Box 2 will allow the practitioner to go through the list of seven dimensions in order to make this assessment. In particular, it will also allow the practitioner to assess the dimension(s) on which the setting might be different, and whether these differences arise from a dimension that is in their control or something that they have little control over. Second, is the nature of their target population

BOX 2: Dimensions to Assess Differences in Situations

Dimension

Description and Possible Elements (CFs)

A: ELEMENTS OF CONTEXT – DESIGN FEATURES

1 FEATURES OF INTERVENTION

2 COMMUNICATION AND MEDIA CHOICES

Specific ways in which the intervention is implemented.

Possible CFs: Prompts to elaborate, Frequency of exposure to intervention, Opportunities to rehearse information, Transparency of the intervention, Intervention information formatting, Assortment features, Distinctiveness of the intervention, Individual vs. group intervention, Community involvement in intervention design, Minimum level of enforcement, Training, Rehearsal of intervention

Specific manner in which communication is structured.

Possible CFs: Media chosen (digital, print, oral) and its effect on delivery, In-person or online engagement, Level of engagement, Complexity or simplicity of communication, Additional communication elements (visuals, audio)

3 INCENTIVES

The incentive for the target user to change behaviour.

Possible CFs: Economic or non-financial benefits to target, Immediacy, visibility, and salience of benefit

B: ELEMENTS OF CONTEXT – EXTERNAL FACTORS

4 SOCIAL ENVIRONMENT

5 PHYSICAL ENVIRONMENT

6 TIME

7 COMPETITION FOR ATTENTION

The effects social elements may have on user performance.

Possible CFs: Presence of others (crowds, kids), Presence of the entity implementing the intervention, Personal connection with the intervention designer, Identity and nature of the intervention designer (e.g., government, for-profit enterprise, academic institute)

Details of the physical environment in which the intervention is delivered.

Possible CFs: Ambient temperature, Lighting in the environment, Noise level, Colours in the environment, Indoor or outdoor setting, Seasonal ambiance, Hours of sunshine

The duration of and temporal location at which the intervention is delivered.

Possible CFs:: Season, Day of the week, Time of day, Duration of exposure to stimulus, Timing of reminders, Time pressure, Time horizon, Level of busyness at the time of intervention, Weekday vs. weekend

The ability of the target to devote attention to the intervention stimulus.

Possible CFs:: Presence of competing sensory experiences and/or stimuli, Information formatting, Time pressure, Salience of message, Stress and time-constraints

significantly different from the population in the original intervention? Box 3 will allow them to make that assessment.

The greater the judged difference between the target and the original contexts, the more important it is for the practitioner not only to a) potentially adapt the intervention to suit the target context, but also to b) test the intervention in-situ (i.e., in the new setting) before deploying it at scale to

ensure that elements of context will not dampen its efficacy. Accounting for context will require reflection, creativity and rigour but stands as a necessity for maximizing the impact of behavioural interventions and advancing the field.

BOX 3: Dimensions to Assess Differences in Target Population

Dimension

1 DEMOGRAPHICS AND SOCIO-ECONOMIC STATUS

2 COMMUNICATION STYLES

3 PSYCHOLOGICAL FACTORS

a) Traits

b) States

c) Skills/Ability

Description and Sources of Variation

Traditional demographics can result in heterogenous responses.

Sources of Variation: Age, Gender, Culture, Subculture, Ethnicity, Income, Education, Religion

Differences in the manner and nuances of communication of a target population or an individual.

Sources of Variation: Native language, Nuance of language, Language norms, Language consistency and congruence

Differences in the psychology of a target population or on an individual level.

Differences in people’s characteristic patterns of thoughts, feelings, and behaviours

Sources of Variation: Baseline motivation, Cognitive bandwidth, Cognitive load, Curiosity, Fatigue, Level of mental stimulation, Need for cognition, Growth mindset

Influences on user performance that are short-term or momentary and may arise, fluctuate, and conclude during or before the intervention.

Sources of Variation: Anxiety levels, Emotions, Fatigue, Motivation, Time pressure, Busyness

Differences in people’s knowledge about a particular subject or topic, or their ability to perform specific tasks or activities

Sources of Variation: Financial literacy, Metacognition of marketplace, Physical activeness, Strength of habits, Intelligence and problem-solving

d) Self-Regulation

e) Attitudes and Beliefs

4 COMPOSITION OF GROUP

Differences in people’s ability to manage and control their emotions and behaviours to achieve desired outcomes.

Sources of Variation: Motivation, Anxiety levels, Attitude to change, Growth mindset, Consideration of future consequences, Discipline, Accountability, Feeling of responsibility, Goal orientation, Impulsivity vs. forward-thinking, Time management

Differences in people’s evaluation about an object, event, or experience.

Sources of Variation: Belief of past performance, Perceived task complexity, Prior beliefs, Pluralistic ignorance

Whether the targets of behaviour change are primarily naïve intenders (who want to change behaviour but are unable to) or opponents (who need to be convinced of the need to change behaviour).

Selina Yang (BCom 2024) is a research assistant at the Behavioural Economics in Action at Rotman (BEAR) research centre at the University of Toronto. Meghan Yeung (BCom 2026) is a research assistant at BEAR. Chang-Yuan Lee is an Assistant Professor of Marketing at the Rotman School of Management, University of Toronto. Nathaniel Barr is a Senior Innovation Advisor at BEworks and a Professor of Creativity and Creative Thinking, Sheridan College.Wardah Malik (BSc 2009) is the CEO of BEworks. Nina Mažar is a Professor of Marketing at the

Questrom School of Business, Boston University. Dilip Soman is the Canada Research Chair in Behavioural Science and Economics at the Rotman School of Management, University of Toronto. David Thomson is the Vice President of Strategy and Head of Practice Innovation at BEworks.

Are Choice Architecture Interventions Considered Ethical?

SUPPOSE A CITY wants to encourage its residents to get their annual flu shot. They might put out advertisements informing residents of the benefits of getting their flu shot. Or, perhaps, they might ask local clinics to automatically sign their patients up for a flu shot appointment, with an option to cancel if the patient chooses not to get vaccinated. This type of strategy, often known as changing a default, is an example of a choice architecture intervention. In fact, in a field experiment with a large medical practice, psychologist Gretchen Chapman and her co-authors found that automatically scheduling flu shot appointments for the practice’s patients increased the total vaccination rate by 10 percentage points.

So, if interventions such as defaults work so well, why might there be resistance to using them for all sorts of behaviour change challenges? While choice architecture interventions have been shown to oftentimes be highly effective at changing behaviour, they have also been criticized for being possibly manipulative and unethical. Specifically, advocates against choice architecture interventions argue that these strategies might impede upon people’s freedom of choice and personal autonomy. Cass Sunstein, Nudge co-author and former Administrator of the White House Office of Information and Regulatory Affairs, was even

referred to as ‘the most dangerous man in America’ because ‘[his] job is to figure out how to make people do what the government wants them to do, but in a stealth sort of way’. Since choice architecture involves changes in the context in which choices are made, and all choices must be made in a context, choice architecture is inherently unavoidable and discussions about its ethics are therefore complicated.

While there is much debate and discussion about ethics, there is precious little experimental work that asks what people actually think about the ethics and acceptability of interventions. In one experiment, I worked with Renante Rondina, Jordan Hutchings, Bing Feng, and Dilip Soman to answer this question. Participants read through different scenarios that presented an example of an intervention and the behaviour that it was meant to change. We also varied how the intervention was explained to the participant to see whether different rationales might lead to different opinions. For example, a participant might have read that a government wants to encourage citizens to become organ donors (the behaviour change), so they provide citizens with information about the number of people who have already signed up to become organ donors (the choice architecture intervention). Finally, they were told that the government

The perceived ethicality of choice architecture interventions is highly context dependent.

is doing this to prevent the loss of lives (the rationale for the intervention). Participants rated how acceptable and autonomy-threatening they found the interventions to be. We generated scenarios using combinations of five different behaviour changes, five interventions, and five types of rationale.

Our results showed that the perceived ethicality of choice architecture interventions is highly context dependent. Changing defaults were rated as less acceptable and more autonomy-threatening than other interventions in our study (like sending reminders or informing the participant about the number of people who have already signed up). Participants’ opinions about the acceptability and perceived threat to autonomy of the different scenarios in our study varied depending on the type of intervention used and the intended behaviour change. The type of rationale that influenced ethicality also varied across different combinations of these factors.

We observed other interesting patterns. For example, people may intuitively believe that emphasizing the effectiveness of an intervention is a universally good strategy to try to encourage perceptions of acceptability. However, we found that this may not always be the case. In some scenarios, highlighting effectiveness made the scenario seem more acceptable to participants. But the reverse was true for other scenarios. For example, when trying to encourage people to get their flu shot using incentives, highlighting the effectiveness of financial incentives improved perceptions of acceptability. But, when used to encourage people to save towards retirement, that same incentives intervention was rated as less acceptable when using the effectiveness rationale. All these patterns lead us to the same conclusion — talking about the ethics of choice architecture isn’t particularly helpful unless you’re discussing a specific behaviour change using a specific type of intervention.

So, are choice architecture interventions considered to

be ethical? Well, our somewhat unsatisfying yet appropriate answer is that it depends. But the encouraging news is that, among the participants in our study, the various scenarios were overall rated as generally acceptable and not particularly threatening to one’s autonomy. Perhaps a city wanting to try to incorporate choice architecture interventions into their policies shouldn’t necessarily worry about being referred to as dangerous.

In closing, we make two points. First, we would discourage blanket statements about the ethicality of choice architecture interventions. Second, we strongly encourage that interventions be tested experimentally for people’s perceptions of ethicality in their unique contexts prior to largescale implementation. What works and is considered ethical in one setting and for one behaviour change problem is not at all guaranteed to translate perfectly into an entirely different situation.

Daniella Turetski is a PhD candidate in Marketing at the Rotman School of Management, and a researcher at the Behavioural Economics in Action at Rotman (BEAR) research centre at the University of Toronto.

EARLY CAREER RESEARCH Stacey Choy

Data is Everywhere!

THE DYNAMICS OF INTERACTIONS among various players in the capital market play a crucial role in shaping its functioning and outcomes. There is ample anecdotal evidence suggesting that firm managers have private discussions with analysts. These private communications have emerged as a crucial source of information for analysts for making decisions. Understanding these dynamics has therefore become important for anyone seeking to unravel the mechanisms that drive information production and transmission in the capital market.

However, private interactions are, well, private! Without any data on these private interactions, it has been impossible to ascertain exactly what role they play.

In research with Professor Ole-Kristian Hope, we overcome this problem by leveraging publicly available data that New York City (NYC) collects and publishes in the interest of transparency. The NYC Taxi and Limousine Commission (TLC) collects pick-up and drop-off dates and times of all daily taxi trips in NYC, along with their precise pick-up and drop-off locations. While researchers in the field of transportation and economics have extensively used the data to study and evaluate the quality of taxi services and competition within the taxi industry, limited consideration has been given to its potential as a tool for uncovering interactions between firm managers and analysts in NYC. We were able to map these detailed, large-volume taxi trip records to the GPS coordinates of companies and brokerages to identify the presence and timing of private meetings between firm managers and analysts.

Our analysis led to several conclusions. Ride volumes increase significantly around earnings announcement dates, suggesting a substantial upswing in interactions between managers and analysts. This observation underscores the importance of earnings announcements as the most valuerelevant information events in the capital market, triggering analysts’ need to interact privately with managers.

Our investigation shows that analysts may learn different information about earnings depending on the timing of private communications because of managers’ control over what information to convey depending on their incentives. Specifically, the increase in ride volumes following earnings announcements is negatively associated with analysts’ earnings forecast errors. The association is further stronger when earnings announcements are bundled with more public disclosures. Our findings suggest that post-earnings announcement private communications can furnish analysts with supplementary intricacies and contextual insights into corporate developments and forthcoming prospects. This, in turn, enables them to better understand the implications of current earnings signals for future earnings. In contrast, the increase in ride volumes before the earnings announcement explains analysts’ downward adjustment of nearterm earnings forecasts. In other words, pre-announcement private communications seem to operate as a direct mechanism, steering the earnings consensus downward to generate marginal positive earnings surprises. Furthermore, when it comes to who managers interact with, we find that managers are more likely to have in-person interactions

with analysts with unfavorable views of their companies, consistent with management’s desire to re-educate them and control information flow in their favor.

We hope that this paper provides a novel way to partially open the black box of private communications that have been difficult to examine due to the lack of data availability. The development and availability of traffic or big data allow researchers to draw more direct inferences about the transmission of information among different entities or agents. By leveraging taxi data, this study advances our understanding of how private communications can serve as an important conduit for conveying firms’ messages to the investment community through the voice of sell-side analysts.

The accessibility and granularity of taxi trip records also lays the foundation for future explorations. By employing a methodology akin to this paper’s approach, forthcoming research could expansively extend its focus to scrutinize contexts wherein private dialogues between pivotal stakeholders — managers, analysts, institutional investors, auditors, lenders, legal experts, rating agencies, regulatory bodies, standard setters, and other relevant participants — play a decisive role in unraveling diverse business outcomes.

From the moment we wake up and look at our smartphone, to the various interactions we have with devices, stores, kiosks, and payment terminals, we generate and contribute to a vast sea of data. These data can become a catalyst for scientific progress by shedding light on interactions and events that previously remained in the shadows.

Stacey Choy is a PhD candidate in Accounting at the Rotman School of Management, University of Toronto.

Banking on Behavioural Science: Commonwealth Bank of Australia

The unit has much to look forward to, as it continues to grow and reinvent the model of what a modern full-service corporate behavioural science unit could look like.

by Jingqi Yu, Sanskriti Merchant, Bing Feng and Dilip Soman

DESPITE THE FACT that every organization is fundamentally in the business of behaviour change of its many stakeholders, not many organizations have successfully embedded behavioural science into its DNA. One exception is the Commonwealth Bank of Australia (CBA), or CommBank

CBA’s Behavioural Science unit has driven hundreds of millions of dollars in customer, business and societal outcomes, won a range of international awards, and worked to develop cutting edge research outputs and customer-facing industry innovations. Today, the team is far from alone; in banking, there are impressive counterpart teams at banks such as Toronto-Dominion (TD) Wealth, Lloyds Bank, JPMorgan, and National Westminster (NatWest) Bank. Like many, the CBA story is neither simple nor linear. Starting out as an experiment of its own, at a time when applied behavioural science was new, and through various hype cycles and false starts, the team has reinvented, reshaped and carved out a strong international reputation. Today, the team comprises around 25 staff and around 35 research partners in three different continents. The unit has much to look forward to, as it continues to grow and reinvent the model of what a modern full-service corporate behavioural science unit could look like.

Back to the Beginning: Before CBA

The year 2011 was a different time for applied behavioural science. Concepts like Nudge were just starting to gain traction in business fields, and the book Thinking Fast & Slow by Nobel Memorial Prize winner Daniel Kahneman had not yet been released. Behavioural units were scarce in any sector, and even the U.K. Government, which is a now a pioneer in the field, had just launched their own Behavioural Insights Team (BIT) as an experiment with a short sunset clause to self-fund the unit. Opportunities in behavioural science roles were limited, and the Sydney market was no better.

William Mailer, having recently completed his M.Sc. program at the Centre for Decision Research and Experimental Economics (CeDEx) in Nottingham, returned to Sydney and noticed the lack of opportunities in the field. He initially joined PricewaterhouseCoopers (PwC), a global accounting and consulting firm, where he quickly connected with Jason Collins, a PwC manager. Their shared passion for behavioural science sparked an immediate partnership, leading to the establishment of the PwC Behavioural Economics practice. They also formed the Sydney Behavioural Science Network, which eventually grew to include 4,000 members. The launch of the BIT

The team showcased the differentiated value unlocked by a behavioural lens, demonstrated tangible impact, and garnered widespread support and recognition.

Sydney office (the first behavioural unit on the ground in Australia) helped generate further interest in the country. The team at PwC went on to win PwC’s global innovation award and expand their reach with the inception of satellite teams in Toronto and London.

The CBA Experiment

In 2016, Mailer was speaking at a local event, and was approached by a senior representative from CBA’s newly formed Financial Wellbeing team. Around the same time, Collins was coincidentally approached to establish a similar team at a financial services regulator.

Mailer joined CBA on a six-month contract, serving as a trial period to demonstrate the value of behavioural science. Amidst the rising support for behavioural science, some approached its capabilities carefully. The CBA’s behavioural science team was set up as a type of experiment in itself, at a time when the Bank was looking for new and more effective ways to accelerate their financial well-being mission. The team’s differentiating factor lay in their unwavering commitment to an evidence-based approach, encompassing meaningful measures, scientific rigour, and a laser focus on driving positive financial behaviours, surpassing mere literacy or information provision.

The team’s early efforts came from a lot of ‘knocking on doors’ to find groups in the bank that had interesting problems, the aptitude for experimentation, and the data. Mailer initiated tactical trials nationwide to showcase the power of behavioural science. Trials included reducing ‘did not attend’ (DNA) rates for financial health check appointments, redirecting staff bonuses towards rainy day savings, and supporting customers to repay debt.

The team got promising results through these trials, gained an internal following, received early press coverage, and formed valuable partnerships. A community started to grow around CBA, as well as connections with other banks, particularly in the U.K. This collaborative approach led to major financial wellbeing innovation projects with government entities, regulators, community groups, and banking experts.

In addition to existing collaborations with the BIT, the team achieved two notable partnerships: one with Behavioural Economics Team of the Australian Government (BETA) and the other with the Sustainability Transparency Accountability Research (STAR) Lab at Harvard, led by Professor Michael Hiscox.

The Growth and Acceleration

Through a collective effort, the team showcased the differentiated value unlocked by a behavioural lens, demonstrated tangible impact, and garnered widespread support and recognition both internally and externally.

Consequently, Mailer was offered a permanent contract and received approval to start hiring, indicating the unit’s value to CBA. At a conference in Boston, Mailer connected with Rafael Batista from the Busara Center for Behavioral Economics. Recognizing Batista’s expertise, CBA brought him to Sydney to help in growing the team. Aligning with their commitment to applying behavioural science, the team established a range of behaviourally informed practices to make sure they were ‘taking their own medicine’. This included de-biased recruitment processes, ‘leader speaks last’ meeting formats, and regular premortem exercises for major initiatives. In their first recruitment wave, the team recruited six staff members from five continents, including some incredible talent that may not have been as easily identified through traditional HR processes. Much of the team’s early success can be attributed to the fantastic range of diversity in backgrounds, experiences and skillsets that the team was able to bring together at this time.

Over the course of the next three years, the team experienced growth and maturity. They further developed partnerships, produced a range of research outputs and began creating proof of concepts for new types of products. During this time, they were able to demonstrate tangible outcomes for customers and for various business lines, attracting attention internally and externally.

A pivotal moment in the team’s history occurred when they launched an ambitious financial well-being event. The idea was to move the bank away from just being a consumer of behavioural science, but to also being a co-producer. The event brought together senior executives from CommBank, eight Harvard professors, and partnering researchers. It comprised 35 behavioural innovation workshops and roundtables, business teams, researchers, community groups, regulators, and advocates, aiming to prioritize and design breakthrough interventions and innovations in financial services that were informed by behavioural science. The team addressed important issues such as debt reduction, building savings, product choice and utilization, retirement planning, benefit claiming, measuring well-being, and implementing experimentation models and internal interventions like diversity, equity and inclusion (DEI) ap -

proaches. The outcome of these workshops was an extensive list of ideas, from which a number of impactful projects were prioritized. This includes a collaboration with Ryan Buell and MoonSoo Choi on operational transparency, the world’s first Financial Well-being Scales, several projects with John Beshears, and the prototype of an innovative offering called the Benefits Finder, a tool that allowed customers to find and apply for benefits and rebates that they might be entitled to. These partnerships not only generated ground-breaking ideas but also paved the way for new projects in the domains of financial well-being, climate action, and internal employee behaviours. These innovations, the partnership, and the ramp up in activity caught the attention of national media outlets and international peer banks.

Over time, the unit gained more media attention, group recognition and industry experience, which led to approvals for further team expansion. They successfully recruited exceptional talent from various parts of the world, including the U.K., U.S., Australia, Mexico, Netherlands, and Kenya. As the team matured, they refined their ways of working. Engagement models, project prioritization scoring, recruitment processes, and team positioning and profile evolved and were codified.

The unit transitioned from a charge model based on timesheets to a centrally funded entity, benefiting from stability and new flexibility to identify priority behavioural opportunities. The team formalized their partnership with the Harvard STAR Lab for research collaborations. A significant achievement was the development of a world-first shared research platform, enabling researchers from renowned institutions around the world to work within secure CBA data environments. Currently, the team collaborates with over 35 researchers from leading research institutions globally through this innovative research collaboration platform. This platform could potentially be scaled up to conduct mega-experiments to address questions like financial capability — questions that go beyond any one financial institution.

The team operates across the entire spectrum of behavioural solutions, incorporating a wide range of approaches. They employ behaviourally informed incentives, boosts, and sludge removal strategies. Meanwhile, recognizing the importance of domain expertise in areas such as hardship, financial abuse and problem gambling, the team avoids overstating the potential of behavioural science alone. They view behavioural science as a valuable lens and tool, working alongside traditional approaches such as incentives, education, pricing, and mandates.

In pursuing tailored recommendations, the team implements a range of tactics, from subtle changes in copy, websites and call scripts to impactful initiatives that reshape choice environments and address systemic factors (S-level changes). These strategic interventions encompass aligning bill dates with wage deposits, facilitating more frequent pay cheques, reducing or eliminating fees, and leveraging AI to prioritize opportunities for customers.

Twists and Turns

As with any new unit, the team experienced numerous changes and growing challenges due to organizational and industry shifts. Despite this, they have consistently adapted and refocused to maximize their impact on customers and collaborate effectively with partners and stakeholders. Being relatively small within a broader workforce of 49,000 employees, the team has been careful to prioritize the right opportunities to have the most positive impact possible. The team looks for opportunities that have a clear behavioural element at their core, which are scalable to large populations, feasible to implement, and help improve customers’ financial lives in important ways.

Carlos Vazquez (Lead, Behavioural Research and Innovation) articulated the process used to prioritize what the team works on:

“The decision is made collegially; there’s a forum where everyone can express an opinion. The projects that get prioritized have three characteristics: First, there must be a clear business problem that needs to be solved. Second, we have the capability to do it. Finally, the project should make a contribution to applied behavioural science more generally.”

The Modern Team

Around 2021, the team rebranded from CommBank Behavioural Economics to CommBank Behavioural Science to reflect its diverse network, team composition, and expanded scope. The team currently has 25 members and works with 35 research partners in eight universities across three continents.

Apart from its growth in size, the team has experienced expansion along two key dimensions:

1) BACKGROUND AND EXPERTISE: It has been particularly fortunate in attracting exceptional talent, moving beyond the days of hiring economists and social psychologists to now include field

The team’s observation calls for attention to contextual factors when developing and evaluating interventions, cautioning against assumptions that what works in one setting will automatically be effective in another.

experimentation experts, lab managers, statisticians, data scientists, AI experts, economists, psychologists, neuroscientists, business / consulting / strategy experts, and commercial leads. It has evolved into a hybrid team comprising Ph.D. students and postdocs (in-house), practitioner partnership panels (allowing the team to flex with demand while working under the team’s model or bringing approved suppliers into the team), and global research platforms (Harvard, Chicago, Sydney, Melbourne, UTS, Macquarie, and Toronto). The team is now a member of the Behaviourally Informed Organizations network (at BEAR, University of Toronto) and the Global Association of Applied Behavioural Scientists (GAABS).

2) PROJECT FOCUS: With support from senior leadership and changes in the group mission, the team currently focuses on a) climate action; b) financial health and digital banking; c) scams and fraud prevention; d) next generation customers’ needs from financial services; e) new innovations in experimentation and partnerships; and f) support for critical national challenges (e.g., COVID, cost of living).

Despite rapid expansion, the team maintains scientific rigour by following strong ethical guidelines based on literature and global counterpart models. It employs rigorous methods such as replication hackathons pre-registration and multi-context trials. Its resources prioritize developing robust methodologies and scaling promising findings within the business. The team’s commitment to research integrity and evidence-based strategies is reflected in publications and co-publications in leading international journals. Andreas Ludwig (Principal Behavioural Science) emphasized the importance of a focus on rigour;

“A very successful initiative is bringing academics in to give presentations on latest research in behavioural science to a wider audience. The exposure to actual research from the researcher (rather than reading a press report or summary) is highly valued by executives across the bank. Our pipelines to create evidence in large scale lab and field experiments allow the team to test the scientific insights directly in an Australian or specific bank context.”

Ludwig is not the only one to put the spotlight on the scientific approach.

“The behavioural economics unit at CBA is different to the many others that have been established by other banks around the world simply because it puts an emphasis on the science.

Its major projects have to be good enough to be published in peer-reviewed scientific journals” (Australian Financial Review, 2019)

Today, the team actively collaborates with global counterpart teams, local community groups, governments, regulators, advocates, etc. It takes a leadership role and cooperative approach in the industry, supporting other behavioural units to establish and grow. They keep pushing the boundaries of applied behavioural science, propelling innovation to new heights.

Next Era – Broader Focus, New Tools

Today, the world has several hundred behavioural units across a range of sectors, looking at a range of behavioural challenges in commercial and societal contexts. The structure of these units is contingent upon various factors, including operating models and strategic focuses. Some units are situated within a particular vertical (e.g., financial services), while others are integrated into broader capabilities (e.g., data analytics or customer experience). Moreover, the home base of a unit may evolve over time as it progresses. For the behavioural science unit at CBA, it started its life in the digital business line of the retail bank and now resides now within a ‘Decision Science’ business domain within the CBA, which also comprises AI, data science, and analytics capabilities. The team’s proximity to the AI and data science teams in CommBank presents unprecedented opportunities, enabling and accelerating the development and testing of new AI tools and interventions to help customers with more tailored services, enhance safety and build trust. As Dan Jermyn, (Chief Decision Scientist) emphasized,

“At CBA, our singular focus with AI is to build a brighter future for our customers and communities. That’s why we see our behavioural science and data science capabilities as two sides of the same coin: our innovation in AI has to be measured through the lens of what it means to the humans who interact with it, driving better outcomes when theory is put into practice.”

Several exciting and impactful projects have been completed and published, while others are underway. These include:

• BENEFITS FINDER: An online tool that simplifies complexity and guides customers to valuable benefits and rebates, connecting customers to more than AUD$1 billion in rebates, benefits and concessions since inception.

• OPERATIONAL TRANSPARENCY: Highlighting the benefits and costs of financial product choices, enabling customers to make better informed choices.

• DEBT REDUCTION STRATEGIES: Utilizing categories, gamified interfaces, windfalls, and pre-commitments to accelerate debt repayment.

• ENCOURAGING HEALTHIER FINANCIAL BEHAVIOURS: Leveraging the CBA banking app to help customers align major expenses with payday, anticipate key upcoming expenses to better plan, and receive immediate feedback on spending.

• RESHAPING STAFF EXPERIENCE AT CBA: Designing and implementing internal strategies to promote savings, increase benefits uptake, drive culture change, and enhance DEI.

In March 2022, the CBA Group broadened its stated purpose to include focus on supporting the national economic recovery, reimagining financial services for customers, and turning more attention to major global challenges like climate action and financial scams (on top of an ongoing commitment to financial well-being). The team adapted its strategic priorities to align with these broader ambitions and developed new and specialist research relationships in support. They also put a heightened emphasis on the scalability and generalizability of their behaviourally informed solutions, aiming to reach the most representative population.

This focus became more critical when the team noticed significant discrepancies in the financial literacy test (Lusardi 3) results across audience groups in their own study: Internal bank audience scored 100 per cent, while customers scored 30 per cent (national survey) to 50 per cent (participant cohort). This suggested that many organizations may not have an accurate view of how people from different communities are managing money and developing financial strategies, which can lead to misidentifying the highest priority financial interventions or needs. The team’s observation calls for attention to contextual factors when developing and evaluating interventions, cautioning against assumptions that what works in one setting will automatically be effective in another.

To be a driving force in promoting sustainable financial behaviour, the team leads a national longitudinal study in partnership with Macquarie University, focusing on Australian young adults. They are also developing a research portfolio in collaboration with the Harvard STAR Lab, assisting customers in reducing carbon footprints, offsetting carbon activities, and investing in new climate-friendly technologies.

Towards the Full Stack Behavioural Science Unit

The team’s success can be attributed to a number of factors, including: a) an exceptional talent pool united by a shared vision (concerted and genuine focus on financial well-being); b) topdown support from early on; c) diverse talents (geographies, training, etc.); d) dynamic and experimental culture at CBA; e) solid infrastructure (time, space, technology); f) effective partnerships (Harvard, BEAR); g) an organization with scale and reach; h) early successes with major national flagship innovations (Benefits Finder AUD$1 billion milestone).

With its reset, broader remit and continued results, the team is growing again and building out skillsets in new spaces: AI, Data Science & Analytics, Ethnography, Design, Experimentation & Statistics, and Behavioural Science. The team is also in the process of developing a multi-speed experimentation model to be able to match speed and rigour with business challenges and stages of programs, including a new experimental lab partnership (University of Technology Sydney), incentive-compatible online surveys, exploratory analyses (ethnographic and data), natural quasi-random designs, randomized controlled trials, labto-field designs, and megastudies.

As with the field of applied behavioural science, the team had started with solving last mile problems, but over the years had developed the expertise and gravitas to address critical first mile problems. As Mailer reflected on the successes and the evolution, he could not help but wonder what additional capabilities the team would need to build, what metrics the team should develop to measure the impact of behavioural science in their work, and what larger and more strategic challenges his team should attempt to address next.

Jingqi Yu is a postdoctoral research fellow at the Behavioural Economics in Action at Rotman (BEAR) research centre at the University of Toronto.

Sanskriti Merchant (Rotman MBA 2024) is a consulting lead at BEAR.

Bing Feng (Rotman MBA 2019) is the Manager of Behavioural Finance at TD Wealth.

Dilip Soman is the Canada Research Chair in Behavioural Science and Economics at the Rotman School of Management, University of Toronto.

P w C CONSULTING: STRATEGIC VALUE FROM BEHAVIOURAL SCIENCE

The

Behavioural Insights Practice at PricewaterhouseCoopers

(PwC) Canada stands out as one of the larger and longerstanding teams dedicated to applied behavioural science.

by Sanskriti Merchant, Ankita Suresh and Jingqi Yu

OVER THE PAST DECADE, applied behavioural science has emerged as an increasingly important tool in strategic management. While dedicated behavioural insights units are becoming more common in industries like financial services, technology and the public sector, relatively few consulting firms have developed specialized units dedicated to this field.

The Behavioural Insights Practice at PricewaterhouseCoopers (PwC) Canada stands out as one of the larger and longerstanding teams dedicated to applied behavioural science. PwC Canada was early to spot the potential of behavioural science to enrich the many solutions they offer clients , and they have a dedicated behavioural science practice that traces its origins back to 2016. Alex Henderson, the current practice lead, describes the team as aligned with PwC Canada’s goals of building trust in society and solving important problems by helping clients craft solutions that go with the grain of human psychology.

The World of Consulting

PwC is one of the four largest professional services networks in the world measured by revenue (i.e., the so-called ‘Big 4’). The PwC network consists of firms which are separate legal entities, but which operate under a shared PwC brand and collaborate to deliver high-quality services to clients globally. PwC currently operates in 152 countries, and PwC Canada is a member firm in the network with over 8,000 employees.

PwC Canada offers a range of professional services, including consulting services that span business management, technology and strategy consulting. The goal is to solve clients’ most important problems by bringing together teams with diverse capabilities and expertise.

While every industry is unique, there are certain elements that are characteristic of consulting. First, the nature of the work is dynamic and client-driven. Each project’s direction hinges on

Solving more strategic and transformative problems for clients requires a different approach to team structure and project delivery.

the business problem the client needs to solve. Staying relevant in the business requires crafting differentiated offerings that meet clients’ evolving needs. Doing this effectively requires seizing the opportunities presented by new developments in technology, analytics and related disciplines and making them relevant to the core business problems that clients are trying solve.

Second, a core measure of a practice’s performance is its sales, reflecting the clients’ willingness to pay for the services it offers. To be successful, a consulting practice needs to develop a robust sales pipeline that keeps the team billable on client projects. Practice leaders must also ensure that they are selling projects that align with the interest and expertise of their team, otherwise they risk losing their most capable talent. Doing this effectively requires collaborating closely across different parts of the firm to develop integrated offerings that deliver tangible business outcomes for clients and deploy the firm’s human capital on projects where they can be most impactful.

Third, consultancies are matrix organizations and deploy team members (or ‘resources’) on a project-by-project basis, which means individuals are not necessarily fixed to a team. For each project — which can range from a few weeks to over a year — there will often be a different engagement manager and team assembled to deliver the work. That means it is not enough to be a pure subject matter expert (although this is important) — every consultant must be able to take on the role expected by other consultants at their level. For example, a Manager in the Behavioural Insights Practice can be expected to bring a behavioural lens to projects, but must also be able to manage the delivery of a variety of consulting engagements that may require them to develop capability in other subject areas (e.g., strategy, operations).

Each of these elements appear to have implications for building a new practice within the firm.

Grassroots Beginnings and the Market Shifts that Enabled It

During the mid-2010s, the field of behavioural science, which originally took root in academia, began to garner curiosity within Canada’s business community, fueled by popular books like

Nudge by Thaler and Sunstein. When Melaina Vinski, a cognitive scientist by training, joined PwC in 2016, she saw an opportunity to bring a behavioural lens to problems that clients face.

The insurance industry served as an ideal testing ground for this approach. According to Matthew Lawrence (Partner, Insurance Consulting), the industry was going through a shift from product-centricity, an approach that structures the client experience around how the business is organized internally, to client-centricity that focuses on improving client interactions and adapting business operating models accordingly. Understanding the application to insurance, Lawrence and Vinski jointly led the sales and delivery of successful engagements using the ‘behavioural toolkit’.

As they saw success, Vinski continued to spread the word about the applicability of behavioural insights in other sectors, including government, retail and telecom. Lawrence recalls it being a ‘one-person show’ in the early days, with Vinski educating the firm on what behavioural science was and how it could help clients achieve tangible business outcomes. Eventually, other influential individuals in the firm saw its value and helped connect the capability with a wider set of business problems.

“The collaboration wasn’t ‘let’s do some behavioural work’,” Lawrence pointed out. “It was rather, ‘let’s focus on the business problem that might exist in a certain industry vertical, and let’s work together to go to market jointly’.”

The behavioural science capability also found synergistic intersections with other areas of the firm. For example, the Customer Experience practice and the Workforce of the Future practice were interested in using a deeper understanding of how people make decisions to help clients shape meaningful interactions with their customers and employees. Kim Vander Aerschot (Partner, Workforce of the Future in Financial Services) noted that the scientific and data-driven approach to tackling ‘people challenges’ appealed to her clientele in banking.

At the same time, the general awareness of behavioural insights was rising in the business community. As the market was catching up, PwC had already built a large library of case

studies from previous client work, including a number of cases where behavioural insights had a significant and measurable impact. There was also a realization within the firm that PwC could continuously use this capability to augment its existing work and get better results.

By 2018, there was enough evidence to answer the question: Can behavioural insights function as a standalone capability in a large consultancy? That is when the firm decided to create a formal Behavioural Insights Practice with dedicated staff to help Vinski meet and grow the market demand. As Lawrence reflected, “We saw the market forces happening in combination with the demand, that led us to having a formal structure around creating a team for this [capability].”

As the use of behavioural insights continued to gain traction around the world, other parts of the PwC network, including the U.K. and Australian firms, have also built similar capabilities. While each firm’s behavioural team has been built organically at the country level, the teams collaborate on knowledge sharing as a global community of practice.

PwC Canada’s Behavioural Insights Practice currently sits under the ‘Financial Services’ homeroom, as PwC is structured by industry verticals. This is partly a function of the practice’s origins in insurance and banking, which continue to be areas of focus for the team. While they have an industry focus, the practice also supports a broader range of industries.

Henderson stepped in as the leader of PwC’s Behavioural Insights Practice in December 2021, after Vinski moved on from the firm. Henderson’s vision for the team at PwC focuses on scaling the capability by expanding the range of problems that behavioural science can contribute to solving and broadening the techniques and technologies the team uses in their work.

Evolving From a Novel Capability into a Sustainable Business

The goal of any consulting practice at PwC is to solve important business problems for clients. A central measure of whether a capability is effective in achieving this goal is whether, and to what degree, clients are willing to pay for it. The transition from a nov-

el capability towards a self-sustaining business is an important chapter in the growth arc of any new capability in consulting.

In the early days of the practice, there was a greater focus on the ‘quick wins’; the projects primarily revolved around solving behavioural problems using nudge techniques, such as communication optimization work that demonstrated high impact while implementing relatively simple solutions. While these types of projects are important for building awareness internally and with clients, solely relying on this approach can be self-limiting. Henderson noted that, “For the most part, clients don’t have behavioural problems; they have business problems that behavioural science can help them solve.”

As the practice has grown and worked to more firmly institutionalize itself within the organization, it has made a conscious effort to move beyond ‘nudge’ projects and towards a deeper integration of behavioural science into PwC’s strategy, workforce and transformation offerings. This evolution will enable the team to tackle larger and more complex problems that behavioural scientists wouldn’t typically be invited to work on.

Solving more strategic and transformative problems for clients requires a different approach to team structure and project delivery. Rather than just deploying as a behavioural insights team, it typically requires bringing together an integrated team that blends multiple capabilities, including behavioural science, with domain or industry expertise. Henderson is focused on aligning the team’s capabilities with well-established and relevant client challenges and finding points of integration with other offerings throughout PwC.

During this phase of trial and error, Vander Aerschot believes that it is important to have strong advocates and sponsors for the team. Historically, it has been mostly Partners in the Financial Services vertical (where the team is currently housed) who have seen the value of applying the behavioural toolkit to tackle problems they were solving with clients.

Andrew Dooner (Partner, Strategy in Financial Services), for example, regularly engages the behavioural team to shape and solve client problems that involve human behaviour or

The team envisions a future where behavioural science becomes an integral tool in the toolkit that PwC uses to solve these core sets of client problems.

interaction. The collaboration has yielded positive outcomes across projects, spanning customer segmentation to piloting tools that enhance frontline banking interactions.

For the practice to continue to grow, Dooner believes there needs to be a larger group of Partners who will actively pull the team into client conversations and find ways to meaningfully embed them in engagements. “It can’t just be ‘Hey, we’re PwC. We do this work, and by the way, here is my behavioural science widget,’” Dooner pointed out. “That’s just a recipe for doing oneoff projects that don’t work or don’t necessarily scale the impact.”

In an effort to connect the dots between the team’s capabilities and business problems, Jennifer Xue (Senior Associate, Behavioural Insights Practice) has been leading the development of the team’s go-to-market playbook. The goal of the playbook is to articulate the team’s offerings in a way that resonates with Partners and helps them connect the capability to business problems commonly encountered by clients.

Any behavioural practice goes on a journey from its inception towards its full institutionalization within the organization. The team at PwC Canada is midway on this journey, having moved past the novelty phase and successfully navigated the experimental capability phase.

It is now at the stage where the team is working to build a sustainable business by establishing credibility as a collaborator across the firm’s core set of offerings, broadly focusing on strategy, workforce and transformation problems. The team envisions a future where behavioural science becomes an integral tool in the toolkit that PwC uses to solve these core sets of client problems.

Rebuilding the Team

Henderson joined as the new lead for PwC Canada’s Behavioural Insights Practice in December 2021. He had launched a similar behavioural insights offering with colleagues in his previous role at a different ‘Big 4’ consultancy, and was convinced of the value behavioural insights could bring to improve business outcomes. PwC Canada had advanced this capability beyond the ‘commu-

nity of practice’ stage (where most other large consulting firms were at), and Henderson was attracted by the opportunity to join a firm that had invested in this capability by creating a dedicated space for it in the operating model.

By the time Henderson joined, the practice’s founder Vinski had been gone for a number of years, and most of the team she built had also exited during the years of the COVID pandemic. This created a unique opportunity for Henderson to rebuild a mature practice more or less from scratch. He had the chance to hire an almost entirely new team.

Finding the right talent is often a challenge, especially in a nascent field like applied behavioural science.

“When hiring in this space, candidates need a combination of deep expertise in the relevant research literature as well as the ability to translate that into applicable business recommendations,” Henderson noted. “It takes time to develop meaningful expertise in the behavioural sciences, and that’s time not spent learning how businesses and organizations function. It can be very challenging to find people with both of those elements.”

While deep technical skills around behavioural science are important to maintain scientific rigour and credibility, it must be paired with business acumen to own and deliver on a client problem. As Dooner emphasized, “In our operating context, we need people who can play multiple roles on the field and have a specialization in an area versus being just a 100 per cent deep functional expert on behavioural science.” That means team members must cultivate the capability to deliver on the business problem, even if only 25 per cent of the answer lies with behavioural science.

Henderson underscores the importance of a structured approach to mentorship and coaching to cultivate both rigour in research methods and the ability to connect with clients. Part of that includes cross-training with teams from other disciplines, such as strategy and workforce, to round out the team’s skillset. For example, Alexandra Bulat (Manager, Insurance Consulting) has been working closely with the Behavioural Insights team since its inception, supporting client engagements as well as market initiatives. She highlighted that working across teams

provides an opportunity to “share perspectives, learn from each other, and bring new ideas to market that consider both the client and industry context.” This ensures the ability to deliver against the core problem and not just the behavioural component.

Melanie Kim (Manager, Behavioural Science), who joined the team after a variety of roles in behavioural science across international and public sector organizations, believes there is a unique opportunity at PwC. “The consulting environment is entrepreneurial and supportive of bringing new ideas to life,” Kim described. “If you have a good idea that resonates with the types of problems clients want to solve, then there isn’t much bureaucracy to make it happen. There’s a lot of room to chart our own path as a team, and that’s exciting.”

With a strong team that has cultivated a positive reputation within the firm, Henderson is excited to see what the future holds for the practice.

PwC Behavioural Insights: Looking Ahead

From the early days of 2016, PwC Canada has come a long way in building and growing a Behavioural Insights Practice. Lawrence, who has been along the entire journey, believes the practice has achieved its initial ambition.

“It has been successful from the standpoint that we’ve proven the case that we can have a standalone team dedicated to behavioural insights,” Lawrence reflected. “The original intent was to show that we could have not just one person, but a full team dedicated as a practice area, and we’ve achieved that objective.”

As a natural progression, ambitions for greater scaling are top-of-mind for the team. As the newly built team gains more traction in the market, Henderson has plans for growth in a few directions.

First is to continue embedding behavioural science more robustly into other big problems that the firm is solving. For example, the team has been exploring how behavioural science can add value to post-merger integration, specifically in facilitating a more evidence-based approach to cultural integration. Mergers

often fail due to cultural friction, so this is a significant problem the team can contribute to solving.

Second is to expand the team’s behavioural toolkit by finding ways to bring new technologies to clients. The team has been developing partnerships with innovative companies that have created technologies that can help clients expand their ability to generate and apply behavioural insights across their businesses.

Third is to further explore the intersection between behavioural science and artificial intelligence. With one team member specializing in behavioural analytics and the team collaborating closely with PwC’s cloud and data analytics practice, there is an opportunity to flesh out compelling offerings in this space.

The journey of PwC Canada’s Behavioural Insights Practice continues, with a clear vision of bringing the behavioural toolkit to a wider range of problems to support clients.

Sanskriti Merchant (Rotman MBA 2024) is a consulting lead at the Behavioural Economics in Action at Rotman (BEAR) research centre at the University of Toronto. Ankita Suresh (Rotman MBA 2024) is a consulting lead at BEAR. Jingqi Yu is a postdoctoral research fellow at BEAR.

DEFAULTS ARE NOT THE SAME — BY DEFAULT

When introducing defaults into real-world environments, choice architects need to be mindful that defaults can have varied responses.

by Jon Jachimowicz, Shannon Duncan, Elke U. Weber and Eric J. Johnson

WHEN YOU RESEARCH DECISION-MAKING for a living, it’s hard not to observe choice architecture everywhere you look — even on vacation. Disney World is the land of magic and fairytales, but even there you cannot escape science. When ordering something to eat, one of us (Jon) noticed that the default choices for kids’ meals were all geared towards healthier options. The menu swapped soda for juice and french fries for fruits and vegetables. Indeed, a recent study shows that this change in Disney World’s policy led to the consumption of 21 per cent less calories, 44 per cent less fat and 43 per cent less sodium. These defaults are helping ‘the happiest place on Earth’ become a healthier one.

Defaults are one of applied behavioural science’s biggest success stories. There are two reasons underlying their widespread adoption. First, defaults can be very simple — even consisting of just the one-word difference between ‘If you want to be an organ donor, please check here’ (opt-in) and ‘If you don’t want to be an organ donor, please check here’ (opt-out). Second, defaults are

surprisingly effective in a wide variety of contexts, from retirement planning decisions to health decisions to consumer decisions.

Despite — or perhaps because of — the widespread use and success of defaults, a couple of important questions have remained in the background until now: How are defaults being implemented? And does it matter how they are implemented? Knowing when and why defaults work highlights the importance of actively rather than passively considering and applying choice architecture tools.

There are many ways researchers, policymakers and other practitioners have attempted to use defaults. As alluded to above, defaults can be implemented in a variety of domains, such as in consumer settings (CFL versus incandescent lightbulbs) or health domains (organ donations). Defaults can also vary in how easy it is for the user to opt out, ranging from a simple click on a website to requesting several forms under Austria’s organ donation law.

In decisions where there are two possible options, the option that is preselected is chosen 27 per cent more often.

The second question revolves around how to see the effectiveness and widespread adoption of defaults in context. Defaults are only one of many tools available in a choice architect’s toolbox. For example, while citizens could be defaulted into health insurance plans, they could also be asked to select their health insurance plan from a smaller, curated set. Similarly, employees could be defaulted into retirement savings plans when joining a company, or alternatively, they could be given a limited time window in which to sign up. Policymakers thus have an array of options to choose from, beyond defaults, when determining how to use choice architecture to attain desired outcomes.

What matters, then, is understanding how effective defaults are as a choice architecture tool, as well as how different

Consumers Are Becoming Wise to Your Nudge

I know exactly how the conversation will go. I’m interviewing Chris, a 52-year-old man living a small coastal town, for the second time. We’ve been exploring the new checkout process for a client’s redesigned website. The new site isn’t performing as well as the company thought it would, so I’m exploring why and seeing what we can learn from competitors.

“Only two rooms left? They don’t expect me to believe that, do they? You see that everywhere.”

I leave with a wry smile. The client won’t be happy, but at least the project findings are becoming clear. Companies in certain sectors use the same behavioural interventions repeatedly. Hotel booking websites are one example. Their sustained, repetitive use of scarcity (e.g., ‘Only two rooms left!’) and social proof (‘16 other people viewed this room’) messaging is apparent to even a casual browser.  For Chris the implication was clear: this ‘scarcity’ was just a sales ploy, not to be taken seriously.

My colleagues and I wondered, was Chris’s reaction exceptional, or would the general public spot a pattern in the way that marketers are using behavioural interventions to influence their

kinds of implementations alter a default’s effectiveness. This was the aim of our recent meta-analysis of all prior default studies, which was published in Behavioural Public Policy

In total, we analyzed 58 default studies with a sample size of 73,675 participants. The studies came from a wide variety of contexts, topics, fields and countries. One thing quickly became apparent: on average, defaults are a strong choice architecture tool, shifting decisions by 0.63 to 0.68 standard deviations. What this means is that in decisions where there are two possible options, the option that is preselected is on average chosen 27 per cent more often than the option that is not preselected. That means that the average default study was about two times more effective in changing behaviours as other

by Simon Shaw

behaviour? Are scarcity and social proof messages so overused in travel websites that the average person does not believe them? Worse yet, do they undermine brand trust?

The broader question—one that is essential to both academics and practitioners—is how a world saturated with behavioural interventions might no longer resemble the one in which those interventions were first studied. Are we aiming at a moving target?

This was the basis for a research project we completed in 2019 examining reactions of the British public to a range of behavioural interventions. We took a nationally-representative sample of 2,102 British adults and undertook an experimental evaluation of some of marketers’ most commonly-used tactics.

We started by asking participants to consider a hypothetical scenario: Using a hotel booking website to find a room to stay in the following week. We then showed a series of nine real-world scarcity and social proof claims made by an unnamed hotel booking website. The result: Two thirds of the British public (65 per cent) interpreted examples of scarcity and social proof

Designing Defaults: The 3E’s

Factor

Endorsement

Ease

Endowment

Intensity and Distribution of Preferences

Question

Does the decision-maker believe that choice architect has their best interest in mind?

How much autonomy do decision makers believe they have?

How (physically and psychologically) easy is it for the decision-maker to change away from the default?

How important is the decision for the decision-maker?

How much prior experience do decision-makers have within the decision domain?

How much do decision-makers believe they are endowed with the default?

What are the decision-makers’ preferences in the defaulted decision?

What is the variability in individuals’ preferences?

strong behavioural interventions that shift decisions by 0.2 to 0.3 standard deviations — one of them being, for example, Opower’s social norm intervention on energy savings, another widely popular choice architecture tool. So, on the one hand, defaults work!

On the other hand, there were also substantial differences in the effectiveness of defaults. In some studies, a default was far more effective than in other studies; and in others yet, defaults did not alter participants’ decisions. This is an important caveat, which highlights that choice architects should not blindly apply defaults to all situations, but instead be more careful in when and how they implement them.

We wondered what factors make defaults particularly

claims used by hotel booking websites as sales pressure; and half said they were likely to distrust the company as a result of seeing them (49 per cent). Just one in six (16 per cent) said they believed the claims.

The results surprised us. We had expected there to be cynicism among a subgroup—perhaps people who booked hotels regularly, for example. The verbatim commentary from participants showed people see scarcity and social proof claims frequently online, most commonly in the travel, retail, and fashion sectors. They questioned truth of these ads, but were resigned to their use:

“It’s what I’ve seen often on hotel websites—it’s what they do to tempt you.”

“Have seen many websites do this kind of thing so don’t really feel differently when I do see it.”

In a follow up question, one third of participants (34 per cent)

more likely to be effective. To do so, we drew on a theoretical framework which highlights that defaults operate through three channels:

1. Defaults work because they reflect an implicit endorsement from the choice architect — your company’s HR department, your city’s policy office, your credit card company or your child’s school.

2. Defaults work because staying with the defaulted choice is easier than switching away from it.

3. Defaults work because they endow decision-makers with an option, meaning they’re less likely to want to give it up, now that it’s theirs.

expressed a negative emotional reaction to these messages, choosing words like ‘contempt’ and ‘disgust’ from a pre-coded list. Crucially, this was because they ascribed bad intentions to the website. The messages were, in their view, designed to induce anxiety:

“… almost certainly fake to try and panic you into buying without thinking.”

“I think this type of thing is to pressure you into booking for fear of losing out and not necessarily true.”

For these people, not only are these behavioural interventions not working, but they are actually having the reverse of the desired effect. We hypothesize that a phenomenon called ‘psychological reactance’ is at play: People kick back when they feel they are being coerced.

Several measures in our study support this. A large minority (40 per cent) of the British public agreed that that ‘when someone

FIGURE

As a result, we hypothesized that default designs that trigger more of these ‘three Es’ (endorsement, ease and endowment) would be more effective. In our analysis, we found partial support for this idea. That is, we found that studies that were designed to trigger endorsement (defaults that are seen as conveying what the choice architect thinks the decision-maker should do) or endowment (defaults that are seen as reflecting the status quo) were more likely to be effective.

In addition, we found that defaults in consumer domains tend to be more effective, and that defaults in pro-environmental domains (such as green energy defaults) tend to be less effective. What this highlights is that the intensity and the distribution of decision-makers’ underlying preferences — what it is that they care about and want — plays an important role in how effective defaults are. When decision-makers care less about a particular choice, a default may be more persuasive in swaying their decision. Likewise, when preferences within a population are more varied — such that some people may have preferences that align with the default, but many people may not — then a default may be less effective.

One domain that people tend to care about deeply — and which tends to be divisive — is their environmental attitudes. As a result, someone who holds more pro-environmental attitudes may be more likely to stick with a default that offsets the carbon emissions arising from their flight, while someone who holds anti-environmentalism attitudes may be more likely to switch away from the default. In addition, environmental attitudes tend to vary broadly throughout the population, as research on the acknowledgement of human-caused climate change, or lack thereof, shows. As a result, both the strong intensity with which people hold environmental attitudes and their broad distribution in the population make it less likely that defaults will be effective.

In contrast, a domain that people tend to care less deeply about — and which tends to be less divisive — is which search engine they use. While there are many search engines available, including lesser-known ones like DuckDuckGo or Qwant, more than 75 per cent of searchers currently go through Google. This metric is accounted for in part because Google is the default search engine on a number of browsers, including

Continued from page 107

forces me to do something, I feel like doing the opposite’. This is even more pronounced in the commercial domain: Seven in ten agreed that ‘when I see a big company dominating a market I want to use a competitor’. Perhaps Brits are a cynical bunch, but any behavioural intervention can backfire if people think it is a cynical ploy.

Stepping back from hotel booking websites, this is a reminder that heuristics are not fixed and unchanging. The context for any behavioural intervention is dynamic, operating in a co-adapting loop between mind and world. Over time, repeated exposure to any tactic educates you about its likely veracity in that context. And certain tactics (e.g., scarcity claims) in certain situations (e.g., in hotel booking websites) are being overused. Our evidence suggests their power is now diminished in these contexts.

In our study, we focused on a narrow commercial domain. It would be unwise to make blanket generalizations about the efficacy of all behavioural interventions based on this evidence alone. And yet nagging doubts remain.

QUESTION 1: Like antibiotic resistance, could over-use in one domain undermine the effectiveness of interventions for everyone?

If so, the toolkit of interventions could conceivably shrink over time as commercial practitioners overuse interventions to meet their short-term goals. Most would agree that interventions used to boost prosocial behaviour in sectors such as healthcare have much more consequential outcomes. In time, prosocial practitioners may be less able to rely on the most heavily used tactics from the commercial domains such as social proof and scarcity messaging.

QUESTION 2: How will the growing backlash against big tech and ‘surveillance capitalism’ affect behavioural science?

Much of the feedback from the public relates to behavioural interventions they have seen online, not offline. Many of the strategies

the company-owned Chrome, but also Firefox and Safari — a default setting that prompted Google to pay Mozilla and Apple billions of dollars. Because people don’t care very deeply about which search engine they use, a default setting is likely to be more effective.

To help understand how to best design defaults, using the three Es and taking into account intensity and distribution of preferences, we put together a checklist of questions that policymakers and other practitioners could ask themselves during the next choice-architecture design meeting (see Figure 1). We note that these questions are not exhaustive but highlight specific aspects to pay attention to when designing defaults.

In closing

Our research exploring when and why defaults work highlights the importance of actively, rather than passively, considering and applying choice architecture tools. It also shows the benefits of understanding how they work. Such ideas may help us predict how well a default could operate in a given setting and figure out how to design defaults that work better.

In addition, defaults may not always be the most effective solution. They represent just one of many tools in the choice architect’s toolbox. To better explore when defaults should be used over other tools, choice architects should also evaluate the effectiveness of defaults versus other possible interventions.

Jon Jachimowicz is an Assistant Professor of Organizational Behaviour at Harvard Business School. Shannon Duncan is Associate Director for the Center for Decision Sciences at Columbia Business School and is a PhD candidate in Marketing at the Wharton School. Elke U. Weber is the Gerhard R. Andlinger Professor in Energy and the Environment, Professor of Psychology and Public Affairs at Princeton University. Eric J. Johnson is a faculty member at Columbia Business School where he is the inaugural holder of the Norman Eig Chair of Business and Director of the Center for Decision Sciences. This article appeared in The Behavioural Scientist, a non-profit digital magazine. For more: www.behavioralscientist.org.

for which big tech companies are critiqued centre on the undermining of a user’s self-determination. The public may conflate the activities of these seemingly ubiquitous companies (gathering customer data in order to predict and control behaviour) with those of the behavioural science community. If so, practitioners might find themselves under much greater scrutiny.

There probably was never an era when simple behavioural interventions gave easy rewards. Human behaviour—contextdependent, and driven by a multitude of interacting influences— will remain gloriously unpredictable.

The lesson to be taken away from our study? Feedback loops affect the efficacy of behavioural interventions more than we realize . Just because an intervention was successful five years ago does not mean it will be successful today. Practitioners should pay as much attention to the ecosystem in which their interventions operate as their customers do. There is no better place to start than spending time with them—talking, observing, and empathizing.

We should also consider our responsibilities as we use

behavioural interventions. Marketers should design nudges with more than the transaction in mind, not only because it is ethical or because they will be more effective over time, but also because they bear responsibility toward the practitioner community as a whole. We owe an allegiance to the public, but also to each other.

Simon Shaw is a co-founder of Trinity McQueen, an insight consultancy in the U.K., and vice-chair of the Association for Qualitative Research. He is a regular contributor to The Behavioral Scientist. For more: www.behavioralscientist.org

The Between Times of Applied Behavioural Science

Every organization is fundamentally in the business of changing the behaviour of its stakeholders.

by Dilip Soman, Jingqi Yu and Bing Feng

IT IS PERHAPS UNCONTROVERSIAL to say that this claim, first made by one of us eight years ago, is now accepted as universal truth. Governments, for-profit organizations, not for profits, startups, consumer protection agencies, financial market regulators, and welfare organizations all believe that a good understanding of behavioural science is critical to behaviour change — and therefore to organizational success.

While the academic field of behavioural science is old, the publication of the 2008 book Nudge by Nobel Memorial Prize winner Richard Thaler and Professor Cass Sunstein thrust it into the spotlight and made it an applied discipline. Nudge not only made the science accessible and understandable to the general public, but it also showed that behavioural science has a fundamental role to play in the success of organizations. Changing the situation in which decisions are made can change the choices that people make. The book made a case for reverse engineering this phenomenon and asked the following: Can we create suitable situations that will steer people towards appropriate choices (a process they called ‘choice architecture’), without significantly changing economic incentives, using persuasion attempts, or using legislations and restrictions? Utilizing numerous examples from a wide array of application areas, the book showed that organizations can use choice architecture to harness the science and help stakeholders make better choices.

Nudge was followed by the birth of the first government behavioural science unit (then referred to as the ‘Nudge Unit’ and later as the Behavioural Insights Team) by the U.K. government. Subsequently, many other governments followed suit. At

the time of writing in May 2023, the OECD estimates that there are at least 200 different units across 40+ countries working in the areas of public policy and citizen welfare, whilst other surveys suggest there are about 600 behavioural science units operating in governments and businesses. It is tempting to conclude that applied behavioural science has been a success in organizations and to rest on our laurels.

However, when one considers the fact that there are an estimated 333 million companies and 195 countries worldwide, 600 units seems like a miniscule drop in the bucket. Note that we do not claim that an organization must have a dedicated behavioural science unit to say that it uses the science. Even if one in ten organizations that use the science have a unit, we are still left with just a proverbial drop! If behavioural science is as influential and as central to the success of an organization, why has it barely scratched the surface when it comes to organizations globally?

The Adoption Challenge

To answer this question, we turn to what we know more generally about the adoption of innovations. We believe that the hesitation shown by organizations in embracing behavioural science isn’t unique to behavioural science. Indeed, we see some of the same challenges in the successful adoption of electric vehicles by consumers, artificial intelligence by firms, and even electricity by society. We ask the question: Why might organizations (or customers) not adopt innovations (products) that might have clearly demonstrated benefits? What might the barriers to adoption be, and how might they be overcome?

Four Approaches to Adopting Behavioural Science

SITUATION BENEFIT COST EXAMPLE

EASY SELLS SMALL SMALL Internal behavioural science consulting

SURE FAILURES SMALL LARGE Requiring an RCT to test for a product that is already a top-seller

LONG HAUL LARGE LARGE Using behavioural science to overhaul design, segmentation, recruitment, compliance or other major processes

SMASH HITS LARGE SMALL Behaviourally informed products or programs (e.g., Save More Tomorrow, Self-Control Products). Using quick online experiments to help a product that is struggling

Harvard University Professor John Gourville wrote an article called Eager Sellers and Stony Buyers that helps us understand why. According to Gourville, the advocates of any innovation (here, behavioural science) have garnered sufficient evidence to showcase its value. They are well-versed with how the science works, how it can be used within organizations, and the processes needed to make it succeed, and they have personally tasted success. However, they might not be sensitive to the fact that other organizations might operate in different circumstances than theirs — and consequently not value the innovation as much. Advocates often wrongly assume that organizations will leap at the chance to embrace behavioural science.

Innovations will be adopted only if the benefits of the designed change are perceived to be subjectively higher than the costs (e.g., changes in processes, time, and resource-building costs). More so than individuals, organizations find comfort in familiarity, as processes become etched into routines, accounting and finance systems, compliance procedures, and other forms of rituals — and they are slow to change. Organizations also run on timelines that are not kind to science; while good science is slow, iterative and thoughtful, most practitioners have hard and non-negotiable deadlines to work with. Table 1 shows four approaches to using behavioural science as a function of costs and benefits.

From Easy Sells and Quick Wins to Fundamental Innovations

Much like the simple claim at the beginning of this article, three of our University of Toronto colleagues — Ajay Agrawal, Avi Goldfarb and Joshua Gans — begin their most recent book with another uncontroversial claim, namely, ‘Electricity has changed our society. It changed the way we live. It also changed the way we work. They also showcased the fact that while electricity is all

around us and it is hard to imagine a world without it, it took a good amount of time (40 years at the least) for its benefits to be experienced widely. Agrawal and colleagues make a distinction between ‘point solutions’ and ‘system solutions’. They refer to processes such as replacing steam engines with electric engines as ‘point solutions,’ i.e., situations in which an innovation is able to improve a specific part of an existing process, without any need to change the process itself. On the other hand, large-scale transformations, such as redesigning factories to streamline the movement of materials are referred to as ‘system solutions’, i.e., the idea that an innovation is valuable but only if changes to the basic operating procedures are made.

What does this mean for behavioural science? Well, behavioural science can clearly offer point solutions. We can imagine that the use of simple experiments can allow marketers to design better communication campaigns, governments to design better programs and policies, or charities to design campaigns that maximize donations. None of these fundamentally changes the way in which the organization operates, but behavioural science has the potential to be a disruptive force that entirely reshapes an organization’s value creation processes, including design and segmentation. It could also fundamentally change the way in which organizations think about issues and make decisions. These are significant sources of value, but they will take a lot of time and patience to accrue. Agrawal and colleagues poetically refer to this passage of time as The Between Times.

Speaking of electricity, we note two features of the state of that innovation today. First, with a few minor exceptions, it is available everywhere and is not concentrated in a small number of central pockets. Second, it gets no (or very little) credit for the value-creating inventions that it has enabled. Eventually, we suspect that when behavioural science is completely embraced by

TABLE 1

If behavioural science is as influential and as central to the success of an organization, why has it barely scratched the surface when it comes to organizations globally?

organizations at the end of its between times, these two properties will also hold, and it will become a true enabler of value creation — a means to an end beyond the science.

Five Principles for Moving Behavioural Science Further

Along the Adoption Curve

Agrawal and colleagues use the description ‘plenty of enthusiasm, but not much to show for it’ to describe the between times in electricity and artificial intelligence. In our world, we believe that while there is a fair bit to show, the untapped potential of applied behavioural science is very significant by comparison. Is there anything other than waiting out the between times, and can this period be shortened? Indeed, if something like electricity (without which we cannot imagine life today) took at least 40 years to be accepted, what chance does applied behavioural science have in a mere 15 years since the publication of Nudge? The following five guiding principles help accomplish fundamental innovations and ease adoption and application. Depending on organizations’ characteristics, they can determine specific steps that run consistent with these principles.

1. START WITH EASY SELLS

Change does not happen overnight. One strategy to work towards the adoption of behavioural science is to start with small and easy wins and roll it out gradually. Organizations could begin by auditing their current business processes and identifying potential areas that need a point solution and others that need a system solution. For example, in the sales and marketing function, behavioural scientists could advise and change the language of their website, their marketing campaign message, and the design of their loyalty program to increase the customer acquisition and retention rate. Behavioural interventions and point solutions can be easily implemented to drive short-term success. Other system solutions, such as changing the faceto-face sales process to online sales, rebuilding the delivery of training, and redesigning an operating model, will take longer to trial by behavioural science. Additionally, organizations could focus on point solutions to help demonstrate proof of concept before jumping into system solutions. Early wins also serve as good use cases and as advertising for the value that behavioural science can bring to the organization.

2. DON’T FORGET THE SMASH HITS

Smash hits are usually high-impact solutions that do not require the adopting organization or the end-user to do things dramati-

cally differently from what they are used to doing. They often take the form of behaviourally informed products, and while they may sound too good to be true, examples do exist, such as the Save More Tomorrow (SMarT) program developed by Richard Thaler and Shlomo Benartzi, which encourages people to commit in advance a portion of their future salary increases towards retirement savings. This behavioural intervention includes three key components to circumvent human biases: 1) a commitment to save in the future which combats present bias, 2) a linkage between planned increases in saving rates and future pay rises, to minimize the influence of loss aversion, and 3) enrolment in the program unless the participant actively opts out, making good use of inertia. Given the relative ease of implementing this program, SmarT became part of the Pension Protection Act of 2006 and has helped over 15 million Americans effortlessly increase their retirement savings. Another recent example comes from the Commonwealth Bank of Australia, which introduced a feature called Benefits Finder on its webpages and app. The initiative removed sludge and made it easy for individual and business customers to find government benefits, rebates and concessions that they might not have otherwise thought about. A third smash hit is a website and an app called stickK (www.stickk. com), which is designed to use the principles of pre-commitment coupled with monetary or psychological contracts to get people to achieve goals they have been putting off. As of May 2023, users had put US$62 million on the line to get things done, qualifying stickK as a smash hit!

3. FIND ALLIES WHO CAN BE ADVOCATES FOR BEHAVIOURAL SCIENCE

Not all stakeholders will effortlessly embrace the adoption of behavioural science, so it is important to spot allies who can support and champion the idea and help make a case to apply it in the organization. These allies can be individuals and teams who understand the science and therefore can assist with regular tasks of applying behavioural science to point and system solutions. They can also be senior management members who will support the initiative and help sell the idea to other business divisions and units that are willing to partner and work together to make long-term changes. Online tools, such as Mind Tools, can help identify and create allies. To seek internal allies, one can start with employees with a more experimental or evidencebased mindset. They can be found in what we traditionally perceive to be more testing-heavy and evidence-driven roles, such as R&D and data analytics, but they could also be found in other functions where experimentation is not a defining feature (e.g.,

an HR professional with a forward-looking mindset who continuously works on streamlining workflow). Sometimes, external partners can be allies as well. Many behavioural science units in the government and business sectors, for instance, have partnered with academia to build case studies that have been shared internally and publicly to drive the buy-in and adoption of behavioural science, thereby emphasizing the importance of ‘getting out’ to venues where behavioural scientists share their latest insights. Doing so not only exposes organizations to the science, but also provides opportunities to find external allies.

4. BRACE FOR THE LONG HAUL

Easy sells and quick success are necessary for an organization to adopt behavioural science in the early stage, but one should also plan to move towards system solutions and aim for the long haul. Organizations could embed behavioural science in all business aspects in several ways, including designing and developing behaviourally informed products and services, creating a human-centric marketplace, and generating a culture of experimentation. Some behavioural science units have been able to drive organic growth by first working on some point solutions, building a case based on their early success, creating more opportunities internally to work on various point solutions from different business aspects, earning acceptance and support for their work, generating system solutions, and bracing for the long haul. In this process, organizations may need to consider questions such as what systems need to be changed, what parameters can and cannot be changed, and what capabilities we now have but did not have before. These changes take time and need patience, but a gradual increase in ambition over time, coupled with some early wins, will set up the organization for success in the long term.

5. EMBRACE HUMILITY AND USE NUANCE TO COUNTER THE NAYSAYERS

It is inevitable that any process that spans a long period will attract critics and naysayers. A product or service adoption process is particularly likely to attract naysayers because the gain of the new innovation usually comes at a loss to the incumbent. It is therefore important to position applied behavioural science as something that will augment and improve other forms of decision making rather than claiming that it will replace them, especially as the latter positioning will immediately put people on the defensive. Furthermore, the early enthusiasm to promote applied behavioural science as a ‘low-cost and light-touch’ solution to any problem, coupled with publication bias in our journals (which portray success more often than failure), might create the (easily fallible and obviously incorrect) belief that behavioural science always works.

We believe that the question of whether behavioural interventions work or not is an inappropriate enquiry. Responding to critiques with dogged rebuttals is counterproductive and can be

distracting; indeed, this approach could prolong the between times A more pertinent question is when (i.e., under what situation and time) they work, and for whom (i.e., what populations). An explicit acknowledgement that not every intervention will work all the time, along with the humility to accept that the field does not have all the answers, will go a long way to not only recognizing the concerns of critics, but also working towards commonality and a more nuanced science. Furthermore, we call for greater transparency and sharing of results (both positive and negative) to increase trust and to counter the effects of publication bias.

In closing

Adopters are humans, and humans are more sensitive to losses than gains (loss aversion). Furthermore, humans evaluate outcomes with respect to some salient reference point (reference point matters), and they prefer to leave things the way they are over changing them (status quo bias). Moreover, humans value an object that they own more highly than they would value it had they not owned it (the endowment effect), and they make judgments based on how easily an example comes to mind (availability heuristic) or how similar a target is to a prototype (representativeness heuristic). Organizations cannot force humans to be more rational, but they can leverage the knowledge of human behaviour to design better solutions to get the best outcomes for stakeholders. While the economic downturn following the pandemic and ongoing military conflicts may limit an organization’s ability to build a new behavioural science capability, this disruptive force may heighten the importance of incorporating the knowledge of human behaviour to achieve the adaptability, sustainability and scalability of any solutions.

Our point is simple: while the between times are real and patience is called for, we believe they could be shortened by embracing the five principles outlined herein.

A longer version of this article appeared in The Behavioural Economics Guide (2023), edited by Alain Sampson.

Dilip Soman is the Canada Research Chair in Behavioural Science and Economics at the Rotman School of Management, University of Toronto.

Jingqi Yu is a postdoctoral research fellow at the Behavioural Economics in Action at Rotman (BEAR) research centre at the University of Toronto.

Bing Feng (Rotman MBA 2019) is the Manager of Behavioural Finance at TD Wealth. She previously served as the Associate Director at BEAR.

The Behaviourally Informed Organizations Book Series

ROTMAN-UTP PUBLISHING is a partnership that was established between the Rotman School of Management and University of Toronto Press in 2008. Our publishing imprint offers a venue for translating research into actionable insights for practitioners, students and general readers. We have had the privilege of working with Behavioural Economics in Action (BEAR) to publish the five books that comprise the Behaviourally Informed Organizations series. These books are based on work done as part of the Behaviourally Informed Organizations (BI-Org) initiative at BEAR. Together, the books in this series exemplify the best that Rotman-UTP Publishing has to offer: accessible and innovative frameworks for understanding complex topics, multidisciplinary and interdisciplinary contributions from both seasoned researchers and informed practitioners, and features that distill and summarize key learnings for application in variety of contexts.

Prior to BI-Org, there has been a lack of practical advice for organizations based on behavioural research. The Behaviourally Informed Organizations series fills this knowledge gap with a strategic perspective on how governments, businesses and other organizations have embedded behavioural

science into their operations. The book series was conceived as a major dissemination outlet for knowledge generated from the Behaviourally Informed Organizations partnership’s three main research themes: 1) developing a comprehensive framework of behaviour change; 2) extending and applying this framework to complex behaviour change challenges (business and social); and 3) identifying mechanisms (resources, capabilities and operating models) by which organizations can best deliver on the promise of behavioural insights (BI). Books, specifically edited volumes, serve as an ideal format for tackling the complexities and nuances inherent to the challenge of applying BI in real-world contexts. Each volume in the series offers multiple perspectives from a wealth of high-profile contributors, expertly framed by the volume editors to equip readers with varying degrees of fluency in BI with the necessary background to approach each chapter and draw connections among them.

The first, eponymous volume in the series, The Behaviourally Informed Organization, edited by BI-Org Director Dilip Soman and Chinese University of Hong Kong Professor Catherine Yeung and with a foreword by Cass R. Sunstein, offers an introduction to and strategic perspective

The University of Toronto Press looks forward to the continued impact of the Behaviourally Informed Organizations book series.

on how governments, businesses and other organizations have embedded BI into their operations. Every organization fundamentally seeks behaviour change, whether it be a government trying to get a business to comply with environmental regulations, a business persuading its customers to be loyal to its products, or a financial institution encouraging a client to start saving for retirement. Behaviour change is critical to organizational success, but despite its centrality to organizations, we do not have a good understanding of how organizations can successfully employ insights from behavioural science in their operations. To address this gap, this book develops an overarching framework for using behavioural science. It shows how behavioural insights (BI) can be embedded in organizations to achieve better outcomes, improve the efficiency of processes, and maximize stakeholder engagement. Over sixteen chapters in four parts, the book moves from setting the agenda, examining overarching insights and tools, providing examples of behavioural initiatives from business and policy, and offering advice on ‘making it work’.

Behavioral Science in the Wild, edited by Boston University’s Nina Mažar and Dilip Soman, and named the 2022/23 Behavioral Science Book Winner by the Global Association of Applied Behavioural Scientists, examines the possibilities and pitfalls of taking BI research from the lab to the field. Behavioral Science in the Wild takes a step back to address the ‘why’ and ‘how’ behind the origins of behavioural insights, and how best to translate and scale behavioural science from lab-based research findings. Governments, for-profit enterprises and welfare organizations have increasingly started relying on findings from the behavioural sciences to develop more accessible and user-friendly products, processes and experiences for their end-users. While there is a burgeoning science that helps us understand why people act and make the decisions that they do, and how their actions can be influenced, we still lack a precise science and strategic insights

into how some key theoretical findings can be successfully translated, scaled and applied in the field. The book has a four-part structure, with part three examining domainspecific behaviour change challenges such as cultivating diversity and inclusion, counteracting fake news, developing healthy eating nudges, increasing blood and plasma donations, and fostering financial well-being.

The third and fourth volumes in the series offer deeper dives into specific complex behaviour change challenges. Cash Transfers for Inclusive Societies: A Behavioral Lens, edited by Jiaying Zhao from the University of British Columbia, Saugato Datta from ideas42, and Dilip Soman, sheds light on widely prevalent cash transfer programs and shifts the discourse on poverty alleviation from purely economic factors to behavioural ones. The book poses key questions central to these objectives: What is the state of the art in the development of welfare programs? What do we know works in these programs and what does not? How can an understanding of behavioural science better inform the design, delivery and evaluation of welfare programs? Cash Transfers for Inclusive Societies develops a nuanced framework for how governments, development practitioners and civil society organizations should design cash transfer programs to improve inclusivity, reduce poverty and improve equality. It draws on field experiments and case studies to highlight past successes, while also offering prescriptive advice for those working to design behaviourally informed cash transfer programs.

The fourth volume, Precision Retailing: Driving Results with Behavioral Insights and Data Analytics, edited by McGill University’s Laurette Dubé, Maxime C. Cohen, Cornell’s Nathan Yang, and Bassem Monla from IBM, takes a very different focus: how to best optimize and evolve business operations in light of the fact that customer experience has largely shifted into the digital realm. Drawing on a host of expert contributors, Precision Retailing takes a broad per-

spective of precision retailing as the interaction point between individuals, organizations, institutions, systems, and policies that support them in ever-changing contexts. The book assembles precision retailing key concepts, methods and tools that complement existing behavioural research. The decision support tools will help managers better capture in real-time the multiscale drivers of consumer behaviour and successfully integrate these into their retail strategy and tactics. Each chapter includes a short strategic brief for successful human-centred digital transformation that focuses squarely on actionable insights for practitioners. Shedding light on the way we understand and manage this complex customer journey, Precision Retailing examines how retail will evolve in the post-COVID era, shaping how businesses meet their future and the inevitable continuation of the digital transition.

The capstone volume of the series is What Works, What Doesn’t (and When): Case Studies in Applied Behavioral Science, edited by Dilip Soman with a foreword by Michael Hallsworth. This book considers how well behavioural science interventions translate and scale in the real world. Consider a practitioner who is looking to create behaviour change through an intervention — perhaps it involves getting people to conserve energy, increase compliance with a medication regime, reduce misinformation, or improve tax collection. The behavioural science practitioner will typically draw inspiration from a previous study or intervention to translate into their own intervention. What Works, What Doesn’t (and When) presents a collection of studies in applied behavioural research with a behind-the-scenes look at how each project unfolded. Featuring seventeen case studies of translation and scaling projects in diverse domains such as financial decisions, health, energy conservation, development, reducing absenteeism, diversity and inclusion, and reducing fare evasion, the book outlines processes, potential pitfalls and prescriptions for how to enhance the success of behavioural

interventions. Each case illuminates how behavioural science research is done — from getting inspiration, adapting research into context, designing tailored interventions, and comparing and reconciling results.

University of Toronto Press is incredibly proud to publish the Behaviourally Informed Organizations series, and we look forward to its continued impact on further knowledge generation within the academy and on communities and organizations outside the academy. We encourage you to learn more about the series, our Rotman-UTP Publishing imprint, and University of Toronto Press at utorontopress.com.

Jennifer DiDomenico is the Editorial Director of Social Sciences at the University of Toronto Press.

Reframing Change to Maintain Trust in Public Health

HAVE YOU EVER changed your mind about something consequential? And have you ever had to backtrack on what you said? Have you found people saying “Wait, so now you’re telling me…?” or “But I thought you said earlier…?”

Even if your motives were sincere, and even if the change was for good reason — such as gaining valuable new information — you may nonetheless feel guilty. Nobody wants to be a flip-flopper and be perceived as unreliable or untrustworthy.

And yet, change is part of life. It was certainly part of life during the peak of the COVID-19 pandemic. As scientists worldwide learned more about the virus, what it was capable of, and what we might do to combat it, changing advice and public health guidelines were inevitable. So, public health officials were in a bind: Consistent guidance fosters perceptions of reliability and credibility, but the evidence they relied on to provide guidance was rapidly shifting. Would changes in guidance harm the public’s trust? How can public health officials be perceived as trustworthy in situations like these?

Applying a Behavioural Science Lens to Pandemic-Era Health Communication

As we began our research, we expected that reminders of changes in public health guidance would have negative effects: ‘flip-flopping’ could reduce the perceived expertise or trustworthiness of public health officials and scientists. So, that’s the bad news.

This ‘bad news’ hypothesis was informed by prior work which suggested that changes in public health guidance might lead to greater distrust of the messenger. Researchers Charlie Rafkin, Advik Shreekumar and Pierre-Luc Vautrey found that reminders of former U.S. President Donald Trump’s changes in messages regarding COVID-19 risk led people to view the U.S. government as less credible and led them to be to be less influenced by U.S. Centers for Disease Control and Prevention (CDC) projections.

But what is the good news? Is there anything that public health communicators might do to mitigate the potentially trust-busting effects of revisions to health guidance? Revisions which are in part due to the nature of scientific

progress itself?

We hypothesized that changes in guidance would be better received if the change was expected, construed in a positive light, and encouraged perspective-taking.

Our study tested a three-pronged approach to contextualizing changes in public health guidelines. We forewarned people about changes in scientific understanding, framed these changes in a positive light, and encouraged perspective-taking with public health authorities. We hypothesized that this might soften the otherwise negative impacts of changes in guidance.

What research informs our three-pronged approach?

1. MAKE CHANGE EXPECTED

We hypothesized that changes in guidance would be better received if this change was expected, not an unpleasant surprise. This proposal is in line with a variety of streams of research, including work on psychological inoculation. This research, pioneered by psychologist William (‘Bill’) McGuire and more recently examined by University of Cambridge psychology Professor Sander van der Linden and others, shows that beliefs can be ‘inoculated’ (made more resistant to future persuasion attempts) by first being exposed to arguments against those beliefs and then having those arguments refuted. Research on stealing thunder shows that preemptively acknowledging negative information about oneself may be better for preserving one’s reputation compared to having others reveal that information. In many situations, surprises carry more ‘weight’ than expected events.

2. ACCENTUATE THE POSITIVE ASPECTS OF CHANGE

We further anticipated that changes in public health recommendations would be better received if these changes were construed in a positive light: as evolution or discovery rather than ‘flip-flopping’ or ‘backtracking’. We could even frame these changes as a form of consistency: public health officials strive to offer guidance that consistently relies on the best evidence, even when that evidence changes. Work by Paul Han, Senior Scientist at the National Cancer Institute in the U.S., along with collaborators, found that wor-

ry induced by reading about all the uncertainties surrounding COVID-19 can be somewhat mitigated by telling people that uncertainty is normal. However, communicating the normalcy of uncertainty alone was not sufficient to produce a statistically significant reduction in this concern.

3. TAKE THE PERSPECTIVE OF THE PUBLIC HEALTH MESSENGER

The third prong might be thought of as empathy in the form of perspective-taking — imagining what it might be like to be a public health authority conveying (changing) science. Perspective-taking has been shown to encourage people to chalk up others’ actions to situational factors rather than over-emphasizing their personality or dispositions. We hypothesized that encouraging perspective-taking might make people more willing to extend grace to public health officials and offer trust despite flip-flopping guidelines.

So, What Did We Do?

We conducted a series of online research studies to put our hypotheses to the test. In the first study, a sample of around 200 Canadians and 100 Americans were either reminded of consistent or inconsistent (changing) public health guidance. They were asked to rate their familiarity with each issue (e.g., guidance on masking), not because we were particularly interested in how familiar participants were with the issue, but we wanted to provide a reason for them to read the messages.

We had two experimental conditions (with emphasis on inconsistency and consistency in guidance, respectively), but the messages were designed to be otherwise similar. For example, the consistent messages included reminders such as, ‘Months ago, Canadian public health officials advised Canadians to wear non-medical masks. Wearing masks is still recommended, for example, when indoors in public places’. The inconsistent messages included one noting, ‘Months ago, Canadian public health officials advised Canadians against wearing non-medical masks. Wearing masks is now recommended, for example, when indoors in public places’.

Our results hold promise for communicators who find themselves needing to backtrack when the evidence shifts under their feet.

In this first experiment, we found that simply reminding people of inconsistent (versus consistent) guidance negatively affected their impressions of experts in terms of their recommendations related to COVID-19. Participants rated both scientists and public health officials more negatively when they were reminded of changing guidelines — this included lower ratings of expertise and trustworthiness (though trustworthiness was marginally, not significantly, impacted for public health authorities).

Altogether, our first study confirmed that reminders of changes in public health guidance could impair credibility. But could we do anything about it?

Our subsequent studies added our three-pronged forewarning intervention: For around half of the participants, the reminders of (inconsistent or consistent) guidance were preceded by a short message:

The scientific process is one of discovery, in which one research finding might initially be the best evidence available, but even better evidence may be found later and may change scientists’ understanding of a topic.

As a result, the body of knowledge is one that is evolving and improving rather than remaining static.

Imagine what this would mean if you were a public health authority:

In order for you to consistently make decisions and provide guidance based on the best available science, these recommendations will end up changing — because the best available scientific conclusions are changing as well.

A Brief Message Neutralizes Negative Effects of Changes in Guidance

The three-pronged forewarning intervention significantly blunted many of the otherwise negative effects of salient changes in guidance. For example, without forewarning, inconsistency led to public health officials being rated as less

trustworthy and less expert compared to when consistency had been emphasized. However, the forewarning intervention neutralized these negative effects.

These results hold promise for communicators who find themselves needing to backtrack when the evidence shifts under their feet. When possible, we should not only communicate the guidance itself, but also provide context regarding how to understand and provide the guidance. If change is to be expected, and even valued, and if the messengers’ predicament is better understood, then we can communicate honestly and transparently — even if this means some detours and backtracking along the way.

Further opportunities for research remain: Are all three prongs required or is one element, such as positive construal of change, all that is needed? Would an intervention that follows changes in guidance be equally effective?

And of course, our recommendations to public health communicators may change as more evidence accumulates — but perhaps that isn’t such a bad thing after all.

Jeremy Gretton is a Team Lead in the Behavioural Science Office at the Public Health Agency of Canada and worked on this research while at the University of Waterloo. Derek Koehler is a Professor of Psychology at the University of Waterloo. Ethan Meyers is a PhD candidate at the University of Waterloo and intern at Defence Research and Development Canada.

BEAR: A Brief History

IN MANY WAYS, the history and growth of the Behavioural Economics in Action at Rotman (BEAR) research centre is deeply linked to the growth of the applied behavioural science field in Canada and more generally in the rest of the world. There were three separate events that served as the catalyst for the formation of the centre. These events all occurred in 2012-2013, approximately four years after the publication of Richard Thaler and Cass Sunstein’s Nudge. There was a general shared sense of excitement about the possibilities that behavioural science could offer to both governments and for-profit organizations. I recall that my then-colleague Nina Mažar (now a professor at Boston University) and I had several conversations that began with loosely paraphrased questions along the lines of ‘Nudge is interesting and the examples are potentially impactful. However, how exactly does this work and how can we get started? What processes and capabilities are needed to generate behavioural solutions?’

Questions like these form the basis of the first catalytic event, the development of a report that came to be known as A Practitioner’s Guide to Nudging. Mažar and I found allies in Kim Ly (MBA 2008) and our then-colleague Min Zhao (now a professor at Boston College), and the four of us set off to develop a set of guidelines to help practitioners

convert the promise of behavioural science and choice architecture into practical solutions. Work on the guide started in the fall of 2012, and we were helped by then-graduate students Julian House (now at the Ontario Behavioural Insights Unit), Liz Lyons (now a professor at the University of California, San Diego), Poornima Vinoo (now a professor at the Indian School of Business), as well as a number of Toronto area behavioural science practitioners. The guide was published in March 2013 and remains one of our most popular and most downloaded publications. It has been translated into multiple languages and used as course material worldwide.

The second catalytic event was related to the development of new Rotman course material. In response to the emergence of massive open online courses (MOOCs), I received several queries on how MOOCs might affect university education. I figured that one way to answer this question reliably was to go through the experience of developing and teaching a MOOC. With the assistance of several student researchers (notably Joonkyung Kim, PhD 2017; Yue Zhuo, BCom 2013; and Christine Lim, BCom 2014), I developed Behavioural Economics in Action (BE101X) and offered it on the edX platform. Dan Mazzotta and Dave Whittington from Rotman’s audiovisual

BEAR

BEAR’s accomplishments are well beyond the expectations we had when it was first established.

(AV) team helped produce, and Pym Buitenhuis’s marketing team helped add pizzazz to the materials. The course launched in 2013 and was followed by multiple iterations until it closed in 2022. Developing that course allowed me to give a lot of thought to what our then-fledgling science meant for practice and how practitioners could leverage it to solve organizational and business problems.

The third catalytic event was the development of a new elective course, Marketing and Behavioural Economics, at both the MBA and undergraduate levels at the Rotman School. Mažar and I taught this class starting in 2013, and there have been a total of eight additional instructors who have taught this popular elective. This course continues to be offered at Rotman (and has extended to our Global EMBA program). Furthermore, additional behavioural content has been introduced in other courses, including organizational behaviour, marketing and finance.

The year 2013 saw BEAR become a research hub. Mažar and I had begun working with several organizational partners to help them implement behavioural science to solve last-mile challenges. While each of our independent projects took their own shape, we quickly learned that it made a lot of sense to establish ourselves as a group so that we could share experiences, create common assets and procedures, and have this group serve as a hub for applied behavioural work at the University of Toronto. Min Zhao and Claire Tsai were also part of the founding team of this hub. Our first task was to choose a name. After several iterations, we decided on ‘Behavioural Economics in Action at Rotman’ (or ‘BEA@R’) —simply because the phrase ‘Behavioural Economics in Action’ had become relatively

well-known thanks to the MOOC. Over time, we ended up dropping the ‘@’, so ‘BEA@R’ became ‘BEAR’. Eugene Grichko from the Rotman School’s marketing team created a compelling logo that has gained a lot of brand recognition over the years. The Rotman marketing team (in particular Courtney Brownell) spearheaded the development of Nudge: The Animation video (now available in five languages) which played an important role in popularizing applied behavioural science.

The early years of BEAR were characterized by a flurry of activity. We started several research projects with the newly formed Behavioural Insights Unit at the Government of Ontario, the Financial Consumer Agency of Canada and other for-profit enterprises. We also developed a number of white papers and reports, including a popular guide on the behavioural effects of gamification, as well as several white papers on topics that included consumer privacy, consumer credit, mobile payments, financial behaviour online, transportation, and responsible gambling. We also conducted several major events, including the financial literacy and well-being symposium and a symposium on consumer privacy as well as poverty alleviation. Our work was supported by our students including Nicole Robitaille (PhD 2014), Patrick Rooney (PhD 2021) and Liz Kang (BCom 2016). Liz later worked full-time as our knowledge translation manager. Yet, between 2013 and 2016, BEAR remained a research hub — it was not a formal entity, and its work continued to be supported by research grants from two faculty members.

The year 2016 brought with it several major changes, the most important of which was BEAR becoming an

Our students and team members have gone on to become leaders in the applied behavioural sciences.

official centre. I spent a few months as a senior policy advisor and a scholar in residence at the newly minted behavioural team within the Privy Council Office, and Mažar was recruited to help set up another newly minted unit at the World Bank. I worked with Melanie Kim (MBA 2016) to put together a proposal to fund BEAR from a new Rotman School funding program. Our proposal was accepted. Starting in 2016, Mažar and I, as BEAR’s founding codirectors had a five-year window of funding that allowed us to do even more under Kim’s outstanding leadership as associate director. In addition to the MOOC, our courses, academic research and events, we started a new webinar series, enhanced our events portfolio by adding a series of book-related events, conducted several round tables and workshops with practitioners, and instituted a new competition for students called The Market for Self-Control - New Product Challenge. We also instituted an active summer internship program in which a number of masters, undergraduate, PhD, and postdoctoral trainees spent the summer months working on problems that ranged from simple last-mile solutions to larger agenda-setting items. This program not only allowed us to train students and work with a growing list of organizations, but more generally allowed us to establish ourselves in both the student and practitioner communities. One of these summer trainees, Bing Feng (MBA 2019), joined us full-time and went on to serve as associate director for the centre.

Over the years, BEAR has been fortunate in having a distinguished set of advisors that have included David Halpern (BIT), Varun Gauri (World Bank), Kelly Peters (BEworks), Vivek Goel (University of Toronto), Dale

Griffin (UBC), Kyle Murray (UAlberta), Abigail Dalton (World Bank), and Sasha Tregebov (BIT Canada). We were also fortunate to attract four outstanding Rotman Professors, Sonia Kang, Tanjim Hossain, Nicola Lacetera, and Lu Han, to serve as chief scientists; and 18 other faculty members from across campus who served as research fellows. These scholars, in conjunction with our many organizational partners (including TD Wealth, Swiss Re, the IG Group, Manulife, Scotiabank, and MediSystem), did exceptional academic and applied research, advised our students, hosted events, and kept the momentum going. With the expertise of Liz Kang, our outputs took the forms of reports, animations, videos, infographics, interactive online reports, and tip sheets, and Rotman’s media officer Ken McGuffin and magazine editor Karen Christensen played a big role in spreading awareness of the work.

Realizing that there was more to be done that needed even more resources, I worked with Melanie Kim, Bing Feng and Rotman’s Jeanne Li to write a proposal to the Social Sciences and Humanities Research Council of Canada (SSHRC) to fund an initiative that later on came to be known as the Behaviourally Informed Organizations (BI-Org) partnership. This partnership included 20 organizations and 23 academic researchers from across the globe with the mission of developing a science of how best to use behavioural science. Kang designed a powerful brand identity and webpages for BI-Org. The partnership and the funding that came with it allowed us to create and disseminate a lot more in terms of ambitious research projects. We attracted several outstanding post-doctoral researchers (including Matthew Hilchey, Renante Rondina, Colin

West, and Jingqi Yu) who worked with faculty and organizations to generate a number of new insights and streams of knowledge. A partnership of this size and complexity required calm and organized leadership, and our Associate Director Bing Feng provided exactly that! We tackled several ambitious projects — organ donation in Ontario, pension contributions in Mexico, cash transfer and welfare programs, sludge, consumer protection, the effectiveness of disclosures, and helping organizations internalize behavioural science. Our work is enshrined in five books as well as a large number of reports and academic papers that can be accessed at www.biorgpartnership.com.

In 2019, BEAR was selected to host the Behavioural Exchange (BX) conference, the largest and most prestigious gathering of applied behavioural scientists worldwide. After successful runs of the conference in London, Sydney, Singapore, and Boston, we accepted the baton to host the 2020 conference, and Cindy Luo joined us as event manager. Unfortunately, we were hit by a global pandemic just a couple of months before the conference, and it had to be canceled. However, Cindy stayed on with us and continues to work as a project manager for BI-Org.

BEAR has now crossed the seven-year mark as a centre and the 10-year mark as a hub. During this time, we have produced 26 reports and white papers and dozens of academic articles, conducted 55 events and 50 webinars, had over 250,000 registrations to our MOOC, taught over 1000 students in our courses, had 90 students work as researchers and consultants, had an additional 700 students participate in our competitions, hackathons and events, and funded about 25 PhD and postdoctoral research projects. It truly takes a village to make this happen! In addition to all the individuals who shaped our history and the organizations that worked with BEAR, the centre could not have achieved much without the support of various departments (finance, HR, building management, AV, events, research support) at Rotman and the University of Toronto.

The numbers in the preceding paragraph are well beyond the expectations we had when BEAR was first established. But as I reflect on the past ten years, I take more pride in our non-quantifiable achievements. First, I am proud of the fact that rather than simply ‘translating’ knowledge, we truly co-create knowledge with our partners. Second, I am delighted that a number of initiatives that we worked on have now been implemented by organizations and governments all over the world. Third, through our many edited books and webinars, we have been successful in giving a voice to researchers and practitioners who might otherwise not have found a voice. Fourth, and most importantly, our students and team members have gone on to become leaders in the applied behavioural sciences, both in academia and in organizations. I dedicate our success to them, and hope that BEAR and the use of applied behavioural science to solve societal problems continues to live, grow and flourish through their collective efforts.

Dilip Soman served as the co-director, then director for Behavioural Economics in Action at Rotman (BEAR) and the Behaviourally Informed Organizations partnership.