BASIC Toolkit Web

THE
TOOLS AND ETHICS FOR

APPLIED BEHAVIOURAL BASIC
TOOLKIT
INSIGHTS
What are the ingredients of good Foreword
behavioural public policies? When and
Behavioural Insights (BI) help understand how context and
how should BI be applied to policymaking? other influences can impact decision-making and inform
Can we use our knowledge of how people the actual behaviour of people and organisations. The BI
think and make decisions to create approach focuses on understanding what actually drives the
decisions and behaviours of citizens, rather than relying on
more effective interventions? The BASIC
an assumption of how they should act. By doing so, it helps
Guidebook addresses these fundamental ensure that policies reflect actual needs and behaviours for
questions and provides a starting point to greater impact and effectiveness.
learn about the process of incorporating BI

Behavioural practitioners and policymakers around the
into policymaking. world have called for guidance on how to apply BI more
systematically and responsibly. This report responds to this
demand and provides practitioners and policy makers with a
set of tools that can be applied to policymaking along with
other existing behavioural frameworks.
The toolkit presented here guides the policy maker through a

methodology that looks at Behaviours, Analysis, Strategies,
Interventions, and Change (abbreviated “BASIC”). It starts
with a BASIC guide that serves as an indispensable and
practical introduction to the BASIC manual.
FOREWORD:1
The BASIC methodology includes a set of ethical guidelines This brochure was prepared by Jun Nakagawa and
to help policymakers ensure they are applying BI responsibly. Francesca Papa, with inputs from Filippo Cavassini, James
BI can raise ethical concerns related to collecting data on Drummond and Faisal Naru, and assembled by Andrew Esson
individual or group behaviours, as well as using experimental of Baseline Arts Ltd. The OECD developed the BASIC Toolkit
methods to test theories at small scales before implementing in partnership with Dr Pelle Guldborg Hansen of Roskilde
them more broadly. Issues around privacy, consent and University, drawing partly on tools developed by his work
ethics of applying certain solutions to only some groups with iNudgeyou – The Applied Behavioural Sciences Group.
can arise. This toolkit presents both general principles for Early versions of the toolkit were tested with participants
the ethical application of BI, as well as a set of dedicated to the Western Cape Government-OECD Behavioural
guidelines to follow during each stage of the BASIC process. Insights Conference in Cape Town, South Africa, from 27-28
September 2018 as well as through public consultation with
NOTE TO READERS the behavioural community.
This toolkit is composed of two pieces that helps walk you
through the why and what of behavioural insights:
l The BASIC Brief: gives a quick overview for senior

leaders and policy makers of why behavioural insights
matter to address the problems faced by governments
and citizens.
l The BASIC Guidebook: provides a practical instrument

for policy officials working in ministries, departments and
public agencies on what is the process through which
the behavioural aspects of a problem can be identified, Identify Analyse Design Test Scale
Behaviours Biases Strategies Interventions for Change
scoped and addressed.
A more detailed how-to manual of the BASIC Toolkit is

available on the OECD iLibrary: http://oe.cd/BASIC
2:THE BASIC BRIEF
A
People’s attention
is limited and
easily distracted.
Attention
The “ABCD” of behavioural insights
in public policy People rely
on mental
Be
D B
People’s
termination
shortcuts
Have you ever missed an important appointment
lief f
willpower is
because you had too much to do and forgot? limited and The ABCD and often
over/under
ormatio
Given up on properly filling out a public form
subject to framework estimate
psychological
because it was too cumbersome and hard to biases. outcomes
De
and probab
n
understand? Driven a little above the speed limit
ilities.
because all the other drivers were going fast
as well? C h oice
C
These are everyday examples of how context and
People are influenced by the
behavioural biases can influence decision-making. framing and the social as well as
situational context of choices.
A better understanding of human behaviour can lead to
better policies. If you are looking for a more data-driven
and nuanced approach to policymaking, then you should Table 1.1 provides examples of successful behaviourally
consider what actually drives the decisions and behaviours informed strategies and their impact. These are interpreted
of citizens rather than relying on assumptions of how they through the lens of the ABCD framework, which focuses
should act. on four key drivers of behavioural problems: Attention,
Belief Formation, Choice and Determination.
This is exactly what behavioural insights (BI) provides.
Drawing from rigorous research from behavioural economics You can start applying BI to policy now. No matter where you
and the behavioural sciences, BI can help public bodies are in the policy cycle, policies can be improved with BI through
understand why citizens behave as they do and pre- a process that looks at Behaviour, Analysis, Strategies,
test which policy solutions are the most effective before Interventions, Change (BASIC). This allows you to get to the
implementing them at large scale. By integrating BI into root of the policy problem, gather evidence on what works,
policymaking, you can better anticipate the behavioural show your support for government innovation, and ultimately
consequences of your policy and ultimately design and improve policy outcomes. This toolkit guides policy officials
deliver more effective policies that can improve the welfare through these BASIC stages to start using an inductive and
of citizens. experimental approach for more effective policymaking.
THE BASIC BRIEF:3
TABLE 1.1. Examples of behaviourally informed strategies using the “ABCD” framework
The “ABCD” of
Problem behavioural drivers Behavioural strategy Impact
Patients fail Attention: People have Send SMS reminders 25% reduction in missed
to attend limited attention and that include the cost of appointments.
their medical recall, but tend to respond a missed appointment to
appointments. to environmental cues. the health system.
Residents speed Belief formation: Paint series of white lines 36% fewer crashes in
up at sharp turns People tend to to create the illusion of 6 months.
resulting in more underestimate speed speeding up so people will
car crashes. and be overconfident slow down.
when performing tasks
like driving.
Households do not Choice: People tend to Send letters to residential 2.0%, reduction in
make sufficient align with the behaviour utility customers electricity consumption
efforts for energy of others and what others comparing their electricity (estimated reduction of
efficiency. think is appropriate. use to that of their 450 000 tonnes of CO2
neighbours. and USD 75 million in
energy savings)
Job seekers are Determination: People Create a “commitment 23% more job seekers
struggling to find have difficulties staying pack” that includes found work.
work. motivated to long‑term meeting with an
goals if left to their own employment advisor to
devices without any plans create an actionable job-
and feedback. hunting plan.
Sources: Hallsworth (2015); Nudgeblog (2010); BIT (2014); (Allcott, 2011); BIT (2018).
4:THE BASIC GUIDEBOOK
THE INTRODUCTION
BASIC
GUIDEBOOK
Have you ever grabbed a chocolate bar at the check-
out line only to regret it later? Filled up your entire bowl
with pasta even though you intended to only take a small
portion? Found yourself sticking with the side of chips or
fries, instead of substituting for a salad?
It is easy to make unhealthy choices even when you choose

to be healthy. These everyday choices add up. Today, more
than one in two adults are overweight or obese in OECD
countries (OECD, 2017b). Around the world, obesity-related
illness is estimated to cost USD 1.2 trillion by 2025 (WOF,
2015). Why do people make such choices? Before making
assumptions, it is important to consider what drives decision-
making given the specific context.
This is where behavioural insights (BI) can help. BI is a tool

based on the idea that context and behavioural biases
influence our decision-making. As a policymaker, you can use
BI to address “wicked problems” like obesity to better design
and deliver policy outcomes.
To integrate BI into your day-to-day work, you can use

the BASIC process to analyse Behaviours, conduct
an Analysis, develop Strategies, test them with
Interventions, and scale up results for policy Change1.
1. Behaviour, Analysis, Strategic, Intervention and Change are the five

stages of the BASIC process for applying behavioural insights to
public policy
INTRODUCTION:5
For example, if your desired policy outcome is to lower adult l Information: Residents do not know exactly how calorific
obesity rates, then you can start by selecting a relevant, burgers are.
specific behaviour (i.e. proportion of healthy items ordered
from restaurant menus). l Cost: Residents find burgers to be cheaper than the
healthier options.
Say you learn that 60% of residents who eat out frequently
intend to take the healthy options but end up choosing l Access: Residents cannot easily access restaurants that
burgers. You start by writing your assumptions to explain why: serve healthy options.
TRADITIONAL POLICY INSTRUMENTS BEHAVIOURALLY INFORMED POLICY INSTRUMENTS
l Require calorie labelling on restaurant menus. l Require calories labelling on restaurant menus that put
calorie counts before the food item because people give
disproportionate weight to the first piece of information
they see. Dallas (2019) found that displaying calories first
resulted in a 16.31% decrease in ordered calories
l Implement a junk food tax on burgers to make them l Implement a junk food tax that requires the price hike
less affordable. of burgers to be clearly marked on the menu because a
price difference is more salient at the point of decision-
making. Chetty (2009) found that tax-inclusive prices
reduced demand by 8%.
l Provide a tax credit to restaurants that provide l Provide a tax credit to restaurants that only provide
healthy alternatives to increase availability. healthy options because adding healthy items next to
burgers can vicariously fulfil healthy-eating goals and
increase indulgent eating habits. Wilcox (2009) found
that adding a healthy alternative increased unhealthy
ordering by 230%.
In theory, better information, prices, or access should lead to adult obesity rates before setting policy and full-scale
healthier eating habits. This is aligned with classical economic implementation.
theory that assumes individuals will choose the rational
decision that maximises their utility. People use information This approach is not only limited to healthy eating or
to make better decisions so you may assume that the more complementing traditional policy levers. By integrating BI
information they have about how unhealthy burgers are, the from the start of the policy cycle, policymakers can design
more likely they will choose a healthier option that will benefit behaviourally informed policies on a variety of issues that
them in the long-run. go with the grain of how people actually behave rather than
go against it, and ultimately improve outcomes without
Unfortunately, we know, even from our own personal compromising people’s autonomy.
experience, that this is not always the case. This is
the central idea behind BI, which is built on extensive This guidebook helps you get started by breaking a policy
research from the field of behavioural economics and the issue down to its behavioural components and identifying
behavioural sciences that have repeatedly found that people potential behavioural barriers that can undermine the
systematically deviate from traditionally explained rational intended policy outcome or enablers that can ultimately
behaviour. This is not to say individuals are irrational but enhance the effectiveness of the policy. It uses a process
rather that you cannot always rely on your assumptions that that guides the policymaker through Behaviour, Analysis,
people will always make the decision that leads to the best Strategies, Interventions and Change (abbreviated “BASIC”)
outcome for them. to apply BI to any policy problem from start to finish
(see Figure 2.1).
By understanding how people actually react and behave
in different situations, you can better anticipate the WHAT IS BASIC?
behavioural consequences of your policy and ultimately
design policies that can help citizens make the healthy BASIC is a toolkit that equips the policymaker with best
choice. practice tools, methods and ethical guidelines for conducting
BI projects from the beginning to the end of a public policy
At this point, you can choose which solution(s) is/are the cycle. Earlier BI frameworks have primarily focused on the
most appropriate in your context, and test which is the end stages of the policy cycle such as experimentation or
most effective in increasing the proportion of healthy items compliance while less emphasis is placed on the behavioural
ordered from restaurant menus. Through testing, you will analysis of a policy problem (OECD, 2017a). BASIC aims to
gain evidence-based results to inform your policy to lower bridge this gap by providing guidance on how to apply to
INTRODUCTION:7
FIGURE 2.1. The BASIC Framework
B
Behaviour
A
Analysis
S
Strategy
I
Intervention
C
Change
Identify and target Scrutinise target Identify and Design experiments Plan for
Step
crucial behavioural behaviours through conceptualise to evaluate the implementation, scale,

aspects of the policy the lens of behavioural behaviourally informed effectiveness of the monitoring, evaluation,
problem. science. policy solutions. strategies. maintenance and
dissemination of results.
Increase pension Individuals tend to Change the default. Test whether allowing Share results with
Example
savings by stick with defaults Automatically enrol individuals to opt-out citizens, apply
encouraging more and choose inaction individuals into increases pension findings to system-
citizens to enrol in over action. pension plans and savings rather than wide reminders and
pension plans. allow them to the current practice monitor long-term
opt-out. of opt-in. consequences of the
intervention.
BI to ex ante appraisal as well as the ex post evaluation and context but have limited or even no experience with BI.
stage of a policy cycle. This approach is reflected in the five You can find approaches, proofs of concepts and details on
stages of BASIC (Table 2.1). By understanding how and under methods for designing and implementing a behaviourally
what circumstances BI can be applied to cause behaviour informed policy intervention in the BASIC Manual
change, policymakers are far more likely to design and deliver accompanied by an introductory guide.
more effective policies.
As you read through the BASIC Guidebook, you will gain

an introduction to behaviourally informed policymaking
and a brief overview of testing and implementation. This is
geared towards policymakers who know the policy problem
l
BASI C
Specifically, the guide will give you:
A practical and in-depth look into the first three sections,

First, ethics should be a priority from the onset. The BI
approach has specific ethical concerns that are different from
traditional public policy because it often involves the use of
Behaviour2, Analysis and Strategies to identify a primary data of individual- or group-level behaviours and
behaviour that is driving the policy issue and why, and leverages behavioural biases to inform policies. As public policy
design actionable strategies based on the behavioural operates within a transparent setting and has farreaching
analyses. implications, it is important to integrate ethical considerations
when applying BI from the start to the end of the policy cycle.
l Outline of the Intervention section that provides general At the end of each section you will find ethical considerations
guidance on engaging with behavioural experts and for every stage of BASIC.
stakeholders at the testing stage.
The final section includes overall ethical considerations and
l High-level recommendations for the Change section specific guidelines for every stage of BASIC.
so policymakers can make an informed decision when
planning to scale and disseminate results after testing. Second, you should be aware of both the benefits and
limitations of BI. Table 2.2 gives a high-level summary of
WHAT YOU NEED TO KNOW BEFORE YOU KEEP READING considerations before deciding if BI is the right fit for
your project.
Integrating BI throughout the policy cycle can enhance the
design and delivery of policy outcomes, but it has several As a reminder, the guide is by no means the only resource to
areas that you should consider as a policymaker before apply BI to policymaking. In addition to the BASIC Manual,
moving forward. there are other useful frameworks and reports to aid you in
your BI project. Box 2.1 shares some key resources that can
2. In the BASIC manual, the core stages of BASIC are referred to in small
caps (i.e. “Behaviour”) to distinguish the stage from the regular use of
provide additional tools and examples to complement the
the word (i.e. in Behaviour you diagnose the behaviour problems) approach provided by BASIC.
INTRODUCTION:9
TABLE 2.2. Considerations before applying Behavioural Insights (BI)

WHAT BI IS WHAT BI IS NOT
Problem-solving method Silver bullet

BI is a powerful method to better understand policy BI is not a silver bullet that solves all policy challenges.
problems and pre-test solutions before they are Some policy issues may benefit more from traditional
implemented across a wide range of policy issues. policy levers (i.e. financial, regulatory or awareness-raising
approaches) or alternative non-traditional tools (i.e. human
centred design or machine learning).
Way to learn “what works” One-size-fits-all

The BI culture of empirically testing solutions and Replicating what works in one environment does not
disseminating results allows practitioners and academics guarantee success in another environment. Ethical
to exchange evidence on lessons learned to inform considerations should also be adapted to the context.
policymaking. Pretesting solutions in the context where you plan to
implement the policy minimises this risk.
Beyond nudging 1.0 Only for behavioural experts

BI goes beyond nudging or small policy tweaks. BI is not limited to behavioural experts. A multi-disciplinary
BI represents a wide range of tools to use evidence to approach is key for BI projects. BI brings together diverse
diagnose problems, bridge the gap between research and expertise such as knowledge of the policy context,
practice, and inform comprehensive policy solutions. behavioural science and first-hand experience with public
service.
Policy tool Irrationality

BI should be considered every time you are designing BI does not suggest that humans are fundamentally
or evaluating a policy. Even in cases where you may not irrational creatures. Rather, it argues that deviations from
be able to start with a behavioural analysis or run a full “traditionally explained rational” behaviour are not the
experiment, BI can still be used to complement traditional result of flawed reasoning but rather adaptive forms of
policy tools and levers throughout the policy cycle. reasoning that can also constitute efficient heuristics (i.e.
mental shortcuts or intuitive judgments) in an uncertain
world.
10 : T H E B A S I C G U I D E B O O K
ADDITIONAL BI RESOURCES
With the rise of BI around the world, a number of useful frameworks have been developed by both government
and non-government agencies. BASIC has been developed to fill a need in the community for how to implement
behaviourally informed public policy.
Below is a non-exhaustive list of widely referenced frameworks that complement BASIC and could be a resource for
policymakers looking for different ways to analyse a behavioural problem.
l MINDSPACE (The Behavioural Insights Team, 2010): Provided an early checklist for thinking about how nine well-
evidenced behavioural insights may inform public policy development, design and delivery.
l Test, Learn, and Adapt (The Behavioural Insights Team, 2013): Gave an accessible introduction to the basics of
using randomised controlled trials in policy evaluation.
l EAST Framework (The Behavioural Insights Team, 2014): Provided a simple framework considering how behavioural
insights may help design policies based on leveraging convenience, social aspects of decision-making and the
attractiveness and timeliness of policies.
l World Development Report Mind, Society, and Behavior (World Bank, 2015): Gave a comprehensive overview of
how the BI perspective on human decision-making is of relevance to development policy.
l Define, Diagnose, Design, Test (ideas42, 2017): Provided a practical framework for thinking through a problem and
identifying behaviourally informed solutions.
l US Internal Revenue Service (IRS) Behavioral Insights Toolkit (IRS, 2017): Created to be a practical resource for
use by IRS employees and researchers who are looking to use BI in their work.
l Assess, Aim, Action, Amend (BEAR, 2018): Presented a playbook developed for applying BI in organisations
outlining four steps for applying BI.

S T A G E 1 – I N T R O D U C T I O N : 11
Ethical considerations for APPLYING BEHAVIOURAL INSIGHTS
Applying BI to public policy can raise specific ethical to be carefully considered since
concerns as BI uses data on individual or group behaviours, individual are inherently not ideally
as well as experimental methods for testing these theories rational.
at small scales before implementing more broadly. As a
result, issues related to privacy, consent or the ethics of You should then discuss and consider the ethical
applying certain solutions to only some groups arise. implications of the intervention.
First, consider some general principles to dispel some BEFORE STARTING A BEHAVIOURALLY INFORMED
misconceptions on the ethics of BI: INTERVENTION
Consider establishing an ethical review board from day
l While we are always being behaviourally influenced, this one. If time and resources do not allow it, then outline the
does not exempt behaviourally informed interventions ethical issues associated with the project, how to address
from ethical evaluation. When applying BI, you are them and continuously consider where ethical approval
intentionally trying to intervene to change the behaviour may be required. A university ethical review board may be
of citizens. This means citizens will experience influences considered for expert advice..
that they would not otherwise, requiring an ethical
evaluation. Appoint ethical supervision of data collection, use
and storage. BI often involves data collection and analysis
l Public acceptance of BI does not make it always ethically that goes beyond what is standard in public policymaking.
permissible and ethical considerations should be Consider appointing at least one responsible – to supervise
discussed for each intervention. ethical aspects of data collection, use and storage.
l While people may avoid a behavioural intervention in Observe existing ethical guidelines and codes of
principle, this does not mean that they can in practice. conduct, which are often already present in public
BI interventions neither force individuals to act a certain institutions. Where existing standards are not sufficient
way nor sanction them for not acting a certain way. This for BI, flag these issues and establish procedures for these
does not mean they are always free to choose how they instances. Ensure appropriate procedures are in place to
want and issues related to consent and awareness needs protect whistleblowing and ensure anonymity is respected.
Stage 1: Behaviour... BI is more effective the earlier you integrate

it into your policy initiative. The first stage,
Behaviour, focuses on problem definition.
Applying BI at this early stage is ideal because it
is less likely that concrete solutions are developed
and more likely that innovative approaches
are welcomed. As a starting point, this section
provides you with the tools to think through four
key questions to help identify and define the
behavioural aspects of the policy issue:
1. What are the behaviours driving the policy

issue?
2. Which behaviours should you target?
3. What is your desired outcome?
4. What is the context shaping your target

behaviour?
S T A G E 1 – B E H A V I O U R : 13
1. What are the behaviours driving the policy issue?

Before applying BI to any policy issue, it is important to Throughout this process, it is important to remember that
define in as much detail as possible the behavioural elements every item may not be behavioural (i.e. generate interest in
of the problem. You can start brainstorming by using a composting). Although these are important to identify, the
Behavioural Reduction tool to identify relevant concrete focus of this exercise should be on behaviour(s) that can be
behaviours that are relevant to your policy issue. Engaging measured (i.e. register for composting programme) and not
stakeholders, citizens and/or behavioural experts even on opinions, values or structural aspects of the problem.
at this early stage can help generate insights into the
behavioural aspects of the problem that you may have not
been aware of as a policy official (Figure 2.2).
FIGURE 2.2. Example of a behavioural reduction
BEHAVIOURAL REDUCTION
1. Write the policy area ENERGY

USE
at the top .
2. Draw relevant strategic Workplace Residential

behaviour behaviour
domains.
Switch to Do laundry
3. List as many concrete Bring own
Take the train energy-efficient during off-peak
decisions, behaviours, coffee mugs
for work trips appliances vs on-peak
to meetings vs
and procedures for each vs planes vs keep old Image here.
single-use cups
appliances hours
strategic domain as possible.
2. Which behaviour(s) should you target?

When choosing which relevant behaviour to pursue, in question, you can rate the target behaviour (i.e. on a scale
addition to financial and political feasibility, there are several from 1 = “definitely not” to 5 = “definitely”), pre-determine
considerations that are specific to BI projects. The priority a cut-off (i.e. questions that score at least a 4.8 will be
filter questionnaire can be a guide to generate a more holistic considered) and calculate the overall score for each.
discussion on which elements are important to the project
and apply this to the selection of the target behaviour. Although each questionnaire is tailored to the project, Table
2.3 presents some general questions to consider. This can be
The priority filter questionnaire is a decision-tool composed an iterative process, so keep coming back to these questions
of weighted questions that reflect important considerations as priorities may change as you learn more about the
for the success of the behavioural project. For each behaviour, context and relevant behavioural biases.
TABLE 2.3. Sample questions for the priority filter questionnaire

Areas Sample questions
Importance Is a change in behaviour an institutional priority?
Ethics Are there any potential risks or unintended consequences when pursuing the desired
behaviour?
Are there uneven risks (i.e. positive for the majority but harmful risks for
minority groups)?
Impact Will changing the target individual behaviour translate to a significant

societal impact?
Feasibility Is it politically feasible? Are resources available? Is it controversial?
Data access Is baseline data readily available? Can you collect individual or group-level
prospective data?
Frequency Does the behaviour occur frequently? Is there a reasonable base rate for
the preferred behaviour?
S T A G E 1 – B E H A V I O U R : 15
3. What is your desired policy outcome?

As you begin to narrow down your target behaviour, it is l Specific to the target behaviour.
also important to start defining a roadmap for your policy l Measurable to assess and quantify results.
initiative. One way is to write down what a meaningful l Assignable to a specific group of individuals.
outcome would look like for your policy initiative. You l Realistic given the time, budget and resources available
can craft a well-defined outcome specific to your target for the project.
behaviour(s) by defining the following criteria (Figure 2.3): l Time-related to ensure outcomes are achieved within a
specified time period.
FIGURE 2.3. Example of SMART outcome
Specific: Define a quantifiable Realistic: If the baseline rate is low (i.e., 10%), then aiming
target at a key decision point. for a 20% increase (12.0%) is a realistic goal.
The outcome is to increase savings by 20% upon receipt of income among

low-income workers with mobile bank accounts by January 2020.
Assignable: Focus on a specific Measurable: Mobile phone Time-related: Clear timeframe
type of user. notifications and amount saved when the measurement will be
can be counted on an individual- made.
or group-level.
Source: Adapted from Doran (1981.
BI IS A DATA-DRIVEN PROCESS
Knowing which data can be used and generated will better prepare you for the Intervention section. When defining your
outcome, consider what kind of data can be collected, how much resources it costs and ways to ensure that collection
does not compromise data privacy. For example, determine if you can link who received which phone notifications
(intervention) with who transferred part of their income towards savings (outcome). Finally, understand the ethical
considerations or preparations (i.e. ethical review board) necessary to protect user privacy and your organisation.
4. What is the context shaping target behaviours?

Gaining a deeper understanding of where the target behaviour A process map is a visual tool that identifies touchpoints for
occurs or is the most likely to occur can shed light on how this the key actors engaging in your target behaviour(s). There
affects your desired policy outcome. A process map that are a number of popular process tools available from the
outlines decision points immediately before, during and fields of psychology and design research. These include:
after the target behaviour can help pinpoint areas where behavioural flowcharts (see figure), behavioural mapping,
you can potentially design a behavioural intervention. user journey mapping, service blueprint).
Symbols and sample Arrows that connect A decision which

behavioural flowchart the symbols and show represents a decision point
process flow (denoted as a diamond)
No Surcharge
Parking Parking
START 30 days delay added and
fine paid fine paid
reminder sent
Yes
The start and A process step Delay that
Yes No Handed
end of a process which represents an represents a time Reminder
END over to tax
(denoted as ovals) activity (denoted as period paid
authority
a rectangular box)
MAP THE “ACTUAL” BEHAVIOUR
The process map should reflect how people “actually” behave rather than how they should behave. Speaking to or surveying
relevant stakeholders and target individuals can generate helpful insights. Observations can expose new insights because people
may not provide honest answers, not remember past behaviour or not be consciously aware of their own behaviours or biases
(Ng, 2016). If you are using a user journey map, go to actual users, observe their process and listen to their real-time feedback.
If you are using the behavioural flowchart, observe friction points such as delays to know how long the delays really are.
S T A G E 1 – B E H A V I O U R : 17
Ethical considerations for BEHAVIOUR
Observe the limits of legitimate public policy Beware not to simplify behaviour
interventions. Not all behaviours driving a policy too much. Behavioural analysis of
problem fall within the legitimate confines of policy problems aggregate patterns of
policymaking. Make sure that you refrain from groups, yet individuals usually hold distinct
targeting and changing behaviours that cannot be preferences. Distributional impacts may also result in
defended as being in the public interest or aligned some citizens being influenced differently than others.
with government priorities. Always consider how to minimise potential side effects
and protect individual rights, values and liberties when
Secure acceptance when targeting behaviours. targeting behaviour change.
Policymakers suffer biases too, which can influence
the decision to target certain behaviour(s). To avoid
these biases, always evaluate the existing evidence
for targeting a given behaviour change.
Stage 2: Analysis...
A
People’s attention
is limited and
easily distracted.
Attention
Once you have identified the
behavioural problems at the People rely
heart of your policy issue, it is on mental
Be
D B
People’s
termination
shortcuts
lief f
important to understand why willpower is
and often
people behave as they do. The limited and The ABCD over/under
ormatio
subject to framework estimate
second section, Analysis, aims psychological
outcomes
De
to examine, through the lens biases.
and probab
n
of BI, which psychological and ilities.
cognitive factors are causing the
C h oice
targeted behaviours.
C
People are influenced by the
framing and the social as well as
situational context of choices.
1. Introduction to “slow” and “fast” thinking 4. Choice
2. Attention 5. Determination
3. Belief formation
S T A G E 2 – A N A LY S I S : 19
1. Introduction to “slow” and “fast” thinking

Analysing these behaviours starts by drawing a distinction FIGURE 2.4. Distinction between “slow” and “fast” thinking
between “slow” and “fast” thinking (Kahneman, 2011;
Figure 2.4). In broad terms, behavioural science shows that Thinking fast Thinking slow
we rely on two types of cognitive processes – a cognitive
process that is slow, deliberate and conscious (i.e. reason
through a math problem) and a cognitive process that is Unconscious Conscious
fast, automatic, intuitive and by and large unconscious
(i.e. recognise the emotion on the face of a friend).
Many of the behavioural problems that BI tackles are the Intuitive Reflective
result of intuitions which we form when we “think fast”. In
particular, we can identify four main aspects of behaviour
that tend to cause the biases involved in behavioural Automatic Deliberate
problems: Attention; Belief formation; Choice; and
Determination (ABCD). This “ABCD framework” assists you
Error prone Reliable
in analysing and diagnosing behavioural problems. The
framework, and how each aspect departs from rational
choice theory, is summarised in Table 2.5. Source: Adapted from Kahneman, D. (2011), Thinking, Fast and Slow, Macmillan
Familiarise with the behaviour Flexible methodologies

Before you delve into the Analysis stage using ABCD, use When studying behaviours, you can adopt flexible research
the process map that you created from the Behaviour designs. This means that the type and number of methods
stage to make sure you are familiar with the target used in your study might vary as data collection continues.
behaviour. If a process map is not feasible, it is still advisable
to examine past data on the behaviour or find ways to locally
observe or engage in the behaviour.
TABLE 2.5. Overview of ABCD framework

Aspect What rationality says What BI shows Example
Attention People should focus on what is People’s attention is limited and Forgetting an appointment.
most important in light of their easily distracted.
knowledge and preferences.
Belief People should form their beliefs People rely on mental shortcuts Underestimating how long a
formation according to the rules of logic or intuitive judgments and often task will take.
and probability. over/underestimate outcomes
and probabilities.
Choice People should choose so as to People are influenced by the Being influenced by what our
maximise their expected utility. framing and the social as well as social circle thinks is the right
situational context of choices. thing to do rather than choosing
the rational option.
Determination Provided that one decides to People’s willpower is limited and Failing to quit smoking.
pursue certain long-term goals, subject to psychological biases.
one should stick to the plan.
?
!
S T A G E 2 – A N A LY S I S : 21
2. Attention
Have you ever failed to take medication or to file taxes on by focusing on making the key information salient and
time? These are everyday examples of attentional problems understandable, seizing the person’s attention and planning
that originate from our cognitive limitations. Humans have for inattention in the decision-making process. In particular,
a finite ability to sift through a large amount of information getting the timing of an intervention right can really make
and to focus on what is important. the difference between the success or failure of a policy.
Attention is often scarce, easily distracted and quickly Guiding questions for ATTENTION biases
overwhelmed. This results in individuals having a hard time
making choices based on relevant information and aligning 1. Is the targeted decision point well-timed and placed in
their decisions with their intrinsic preferences. Common a context where people are in a suitable state of mind?
examples of inattention are forgetting commitments and
overlooking non-obvious information (or “salient,” in BI 2. What is seizing people’s attention in that particular
terms); as well as falling prey to distractions while working. context?
As we will see in the Strategies section, behaviourally 3. What happens if people are inattentive at the decision
informed policy solutions can target attention problems point? Is there a default safety mechanism in place?
HOW ARE ATTENTION BIASES RELEVANT TO POLICYMAKING?
Attention biases have been shown to affect numerous domains of relevance for policymakers. For example, they can
affect the number of people who enrol in pensions (Thaler and Benartzi, 2004) or donate organs (Johnston and Goldstein,
2004) when they do not actively exercise their attention and just choose the default option that a system offers them.
Equally, forgetfulness can have important effects on important health and legal effects if, for instance, people do not
show up to court (Ideas42, 2018) or doctor’s appointments because of memory limitations.
Sources: Thaler, R.H. and S. Benartzi (2004); Johnson, E.J. and D.G. Goldstein (2004).
3. Belief formation
It is a well-replicated finding in the social psychology As we will see in the Strategies
literature that, when asked to compare their driving skills section, behaviourally
to other people, the majority of participants (up to around informed solutions provide
90%, as in Svenson (1981), rank themselves in the top 50% – processes and tools to move
which cannot possibly be mathematically true. away from this confirmation bias with a view of supporting
good judgment and accurately assessing probabilities.
This form of overconfidence is in line with the behavioural
findings on belief formation, which show that individuals Guiding questions for BELIEF FORMATION biases:
tend to rely on a coherent worldview to make
predictions and decisions. In doing so, they ignore relevant 1. What are people’s pre-existing beliefs? What questions
information that goes against their views or only accept direct their search?
information that confirms these beliefs. The consequence
can be over/underestimation of outcomes, missing relevant 2. How does context interact with belief formation?
information and relying “too much” on heuristics (i.e. mental
shortcuts or intuitive judgments) to make decisions.
HOW ARE BELIEF FORMATION BIASES RELEVANT TO POLICYMAKING?
While they are only rarely within the scope of policy discussions, erroneous beliefs can be a real threat to policymakers.
For example, if a population misperceives the probability of high-consequence events such as natural catastrophes or
terrorist attacks (Sunstein, 2003), this might affect the appropriateness of the time and resources their community
spends to prevent them.
Moreover, cognitive traits linked to belief formation, and notably overconfidence, have been linked to issues of high
priority for policy, such as financial crises (Lo, 2013) or people’s beliefs about whether climate change will personally
impact them (Gifford et al., 2009).
Sources: Sunstein, C.R. (2003); Lo, A.W. (2013); Gifford, R. et al. (2009.
S T A G E 2 – A N A LY S I S : 23
4. Choice
Have you ever been confronted with too many options for Guiding questions for CHOICE biases:
you to possibly choose the best one? This phenomenon is
defined as choice overload and is one of the ways in which 1. What makes a given choice attractive to people?
our decision-making is influenced by behavioural factors.
The context and moment in which we make choices 2. How are choices framed?
have a distinct upshot on whether we will choose the
best option according to our preferences. Often, people
value more intrinsic factors and motivation like the feeling
of “doing the right thing” than purely material or economic
incentives, running sometimes counter to traditional
economic models. Failing to think of all rational and
irrational aspects of choice can lead to policies that miss
the driver of individuals’ decisions.
Behaviourally informed solutions test possible choice

mechanisms and use the results to inform decisions.
HOW ARE CHOICE BIASES RELEVANT TO POLICYMAKING?
Biases in choices can have negative consequences on important policy outcomes. Think, for example, of how biases
are used to sell us things we do not want or guide us to certain choices over others. Instances include restaurants that
structure their menus strategically or airline companies that exploit people’s inattention and set the purchase of flight
insurance as a default option (European Commission, 2014). Having awareness of these choice biases is fundamental in
order to create responsive regulatory regimes that effectively protect consumers in the market. It is in this context that
behaviourally informed interventions can be leveraged for “better information disclosure, access to customer service,
usage and consumption of data and understanding of bundled services” (OECD, 2016).
Sources: European Commission (2014); OECD (2016).

5. Determination
Taking the right decision is not enough, as anyone who Guiding questions for DETERMINATION biases:
has ever subscribed to a gym and then hardly attended
can easily understand! Individuals can have a hard time 1. What are the points of friction relative to the desired
sticking to some of their choices over time because behaviour? Is it too easy to do the wrong thing?
of issues with will-power, self-regulation, self-control
and choice architectures that guide people away from 2. Do people have plans and are they given feedback?
their decisions. Problems with determination involve the
psychological discomfort of not being able to achieve a 3. How do performance and goal achievement interact
long-term goal and guide people to search for immediate with the social context? For example, do people
gratification. Determination challenges can also create commit to their long-term goals privately or publicly?
mental taxation or exhaustion, which has been shown And what kind of expectations do such commitments
to decrease our decision-making abilities. Finally, the create in other people?
determination biases can create a climate of inertia and
procrastination and eventually lead to excessive self-
directed blame.
As the Strategies stage suggests, behaviourally informed

solutions take into consideration these commitment biases
and provide plans and feedback to increase determination,
for example by using reminders and commitment devices.
HOW ARE DETERMINATION BIASES RELEVANT TO POLICYMAKING?
Intention-action gaps can bring about many policy problems and make existing policies ineffective. For example, they have
clear implications for public health, which is hugely impacted by people’s limited ability to act upon their desire to lead a
healthy lifestyle, in terms of exercising and eating healthy but also, for instance, quitting smoking.
Similarly, it may impact people’s ability to think long terms, in terms of monetary savings as well as in the case of
sustainable behaviour.
S T A G E 2 – A N A LY S I S : 25
Ethical considerations for ANALYSIS
Behavioural analyses usually observe or study Consider what guidelines must be

human behaviour close up and often in their followed when studying behaviour
individuals’ everyday environments, running the up close. These include collecting and
risk of affecting participants’ personal lives and documenting consent, revealing the purpose of
colliding with people’s privacy. the study, ensuring participants are voluntarily
participating and additional safeguards are in place
Seek ethical approvals and competencies where when studying vulnerable populations.
necessary. Use the ethical review board or relevant
authorities within which the behaviour is studied Only collect data that is necessary and ensure
to grant approval. If using a third party to conduct secure handling. Ensure that those handling the data
the study, this ethical responsibility cannot be are properly instructed in the secure collection and
transferred. Ensure appropriate training to develop handling of data.
sufficient competencies for data use and analysis.
Stage 3: Strategies...
Building on your behavioural analysis, the
next step is to identify behaviourally informed
strategies that will effectively change the
identified behaviours that you wish to or can
address, at the root of your policy problem.
Translating your analysis to actionable

strategies inevitably requires some level The goal of this chapter is to get overview of
of familiarity with previous research in the the concepts that you can use to generate
behavioural field, so it is advisable to refer to behaviourally informed solutions. Because these
the expertise of a behavioural team or expert. strategies are strictly grounded in the first two
Nevertheless, having a preliminary understanding stages of BASIC, the section will also follow the
of the key principles behind the most common structure of the ABCD framework.
behavioural strategies will facilitate your
relationship with behavioural experts and allow 1. Targeting attention
you to better understand how to reduce the
behavioural barriers at stake. 2. Targeting belief formation
3. Targeting choice
4. Targeting determination
S T A G E 3 – S T R A T E G I E S : 27
1. Targeting attention
Attentional issues are rarely at the centre of the l Timing and placement of an intervention play a huge
development, design and delivery of public policies. Yet, as role in whether people will pay attention. For instance,
mentioned in the Analysis stage, inattention is widespread to increase the likelihood of people paying fines, it might
and it can make the difference between failure and success be strategic to time the deadlines of fines and charges
of a policy. For this reason, it may prove effective to revise relative to when people receive their pay. Similarly, placing
and design policy interventions so that they become more healthy food at eye-level rather than far from sight could
relevant, seize attention and, if this is not possible, so that positively impact people’s choices (Thaler and Sunstein,
they plan for inattention. 2008).
MAKE IT RELEVANT SEIZE ATTENTION
For interventions to work effectively, it is first of all One of the ingredients that can make an intervention
important to engage with people in impactful ways. This successful is simply making sure to seize people’s attention.
means engaging with them at the right time, in the right Behavioural insights shows us that people often fail to
place and at the point where they are most willing to enact attend to what may be perceived to be important in
the behaviour that you aim to promote. Some behavioural a given context. In light of this knowledge, you should
insights to keep in mind in this regard: always carefully consider how to design the details of your
policy intervention so that people will not overlook what
l State of mind: People’s abilities and motivation are not is important for the intervention to succeed. Three BI
constants (Loewenstein, 1996) but rather are influenced strategies can help to do so:
by their current state of mind. For example, if you are
hungry or tired, you are more likely to make mistakes, l Make it salient: Activate, guide or retain focus on a
make worse decisions and eat bad food. particular aspect of the choice architecture so that
people attend to it. A famous example of this principle
is a behavioural experiment in the Schiphol Airport in
Amsterdam, where engraving silhouettes of flies into the
urinals reduced spillage by 80% and cleaning costs by 8%
(Evans-Pritchard, 2013).
l Reminders: Send reminders and trigger an association in 2. Targeting belief formation

people’s memory. Reminders are becoming increasingly
relevant due to digitisation and have been shown to What is the best way of preventing the formation of
successfully induce behaviour change in a vast array of erroneous beliefs and inaccurate interpretations of
domains, including health outcomes (Stubbs et al., 2012). probabilities? Effective strategies include guiding search,
making inferences intuitive and supporting judgments.
l Prompts: Using prompts is another powerful way of
grabbing people’s attention. Prompts are now especially GUIDE SEARCH
common in the online universe in the form of pop-up
boxes and they work by interrupting people’s ongoing Problems in belief formation usually go hand in hand with
action and forcing them to make a decision before being vast amounts of information and possibilities: people have
able to proceed. too much information and too little time to process it. Tips
for avoiding this include:
PLAN FOR INATTENTION
l Searching by aspects. Allow people to use filters to
Examining what happens when attention fails and then partition the set of options they are given. This kind of
planning and designing for inattention is also a central “search by aspects” has proven useful to guide people
strategy for dealing with attentional problems. through complex informational sets, such as in the online
search for flights, hotels or job-openings where people can
DEFAULTS prioritise which features are most
important in their search.
We increasingly rely on defaults or pre-set choices to decide 1 2 3
for us when we do not have the time or capacity to carefully l Question trees. We are all 4 5 6
7 8 9
examine the vast array of choices available. It is, therefore, familiar with call centres 0 #
*
crucial to get the arrangement of defaults right and to guiding us to the right service
prevent their misuse. In particular, if your policy issue involves section by using prompts
defaults, ensure that these are well aligned with individual such as “press 1 for English”.
and societal preferences, in order to guarantee compliance Using this type of Q&A tools
with ethical standards. It is also important to consider (sometimes called
active decision-making, by sometimes still offering a choice, “question trees”) is
otherwise, defaults could backfire. another powerful way
to guide users to the right answer and help people find Similarly, adjusting
their way around vast and complex sets of information. information architecture and
In the public policy domain, this could, for example, be layout on public websites so
applied to guide citizens to the information they need that, in BI terms, it conforms
when interacting with public bodies. to their “mental models”
(i.e. resembles what they
MAKE IT INTUITIVE are used to, searching for,
or browse most frequently)
If you want to help people navigate information, it is also may significantly improve the
important to structure information in a way that is easy for functionality and experience
them to understand. The concept of “intuitive coding” might of the service.
prove crucial for putting in place user-centric interfaces in
public policy, such as creating medical prescription forms As in the case of King et al 2014, where prescribing errors
that are intuitive for any citizen and will help them adopt the were significantly reduced through slight changes in the
right behaviour (King et al., 2014). format of the prescription form.
Intuitively-coded prescription form

SUPPORT JUDGMENT
New camera
People inform their beliefs from pre-existing beliefs. In

doing so, people use their intuition and an array of mental 3,792
shortcuts or intuitive judgments (i.e. heuristics) to make
decisions in uncertain contexts. This plays a part in forming
beliefs in specific contexts. This knowledge can be used to
design effective Strategies:
l Adapt to heuristics: Your policy interventions should

strive to present information in a way that allows for
the appropriate application of heuristics. The setup of
the intervention should match, rather than conflict with,
people’s intuitions.
l Leverage social proof: When people are uncertain about

what is the right thing to do in a given context, they often
look at the behaviour of others in an attempt to make
sense of the world. By highlighting a positive behavioural 3. Targeting choice
norm, practitioners may support judgment by “de-biasing”
the existing misperception or potentially promote the MAKE IT ATTRACTIVE
misperception that the positive behaviour is more common
than it actually is, which might result in people adopting “The fundamental law of choice is that of attraction”
the desired behaviour. (Hansen, 2018). When faced with a set of options, people
usually select the option that they find most attractive. If you
want to make your desired policy outcome more appealing,
consider the following insights:
l Consider motives. What are people’s drivers and FRAME PROSPECTS

incentives in your policy context? Getting a sense of their
extrinsic and intrinsic preferences will allow you to align Whenever we face a set of options, we are also confronted
your intervention with people’s strongest motives. with a set of possible futures, i.e. prospects. Arrangement
and formulation of these options matters in determining
l Create perspectives. When people hold weak preferences which choice will be chosen.
over options, highlighting an attribute that may provide a
secondary motive for choosing among options might be To illustrate the potential effect of arranging choices,
an effective way of helping people make choices. consider the choice options in the two arrangements in
Figure 2.5. As can be understood intuitively, some people
l Trigger emotions. Our emotions are a fundamental factor who prefer the small coffee in the first setting actually have
when navigating choices. To choose, we internally simulate a preference for the medium size in the second setting, as
the consequences of making one choice over another. people have a tendency to choose the middle option rather
Anticipating your public’s emotions can thus help you than extremes (the compromise effect).
understand what kind of strategies will work best in your
environment. Ethics is very important in this strategy.
Figure 2.5. Arranging choices – Which do you prefer?
A. Which do you prefer? B. Which do you prefer now?
€2.50 €3.50 €2.50 €3.50 €4.50

Prospect theory, one of the theoretical pillars of MAKE IT SOCIAL

behavioural economics, provides an explanation for this
and shows that people think in terms of expected utility Humans are, first and foremost, social animals. Yet, this is
relative to a reference point rather than absolute terms. often overlooked in public policy, where they are treated as
Furthermore, the theory indicates that “losses loom larger individual citizens and consumers. Connecting with the social
than gains”, meaning that people are more strongly affected identities and norms that informally regulate human groups
by the prospect of loss than the prospect of gains. An is an invaluable strategy to spark behaviour change. Two
implication of this theory is the so called “endowment main insights to do so include:
effect”, by which we attach more value when selling an
object than when buying it (Kahneman et al., 1990). l Connect with social identities and consider peer
comparison. This mechanism is what drives people’s sense
Practitioners might use prospect theory when deciding of status, recognition and identification with a group and
how to formulate simple prospects, such as those faced by a powerful force behind their behaviours.
citizens when making everyday decisions in their interaction
with public bodies. l Create a sense of community. Would you rather go to a
popular theatre or restaurant or to an empty one? Most
of people’s choices ultimately have deeply ingrained social
The endowment effect dimensions to them. Observing the role that a sense of
community may play for how people make choices and
creating a sense of community around certain activities
may hold the key for influencing and creating behaviour
change. This is evidenced by big marathon events,
communal eating events, and charity fundraising.
0 0
€4.5 €2.5
MINE NOT MINE
People value things they posess much more

than those they don’t
4. Targeting determination
Finally, we overview Strategies to overcome the intention- l Commitment devices. Make use of strategies that will
action gap. Simply, if you want people to achieve a goal, help people follow their plan of action. For example,
you should strive to make it easy for them to achieve it. encourage people to publicly, rather than privately,
Otherwise, even relatively small obstacles may become a commit to a certain goal so they will have the added
reason for people to postpone action. incentive of maintaining their reputation.
l Work with friction. Reducing the hassle-factor and l Leverage social norms. These are the mutual
simplification can make it easier to take up a preferable expectations that govern the behaviour of members of
service or performing an action. This might translate into groups and societies. These can lead people to forego
reducing the number of actions, clicks or questions people self-serving behaviour in favour of fairness and persist
need to go through to achieve a goal. The simpler it is to even when everyone in the group would prefer the norm
perform an action, the more likely it is that people will do it. did not exist. In some situations, these norms can help
achieve policy goals – especially when promoting pro-social
l Provide plans and feedback. As anyone who has been behaviour – but care should be taken as they intervene
on a diet can understand, behaviour change requires and make use of the structures and foundations of society.
that goal-directed behaviours are not just initiated or
considered once or twice, but are also continuously Future lock-in, holding policy constant
maintained over time. Mental taxation and balancing Distant future emphasis increases strength of support for gas tax increase
of competing goals can easily lead to failure. On the 18%

contrary, making concrete and specific action plans 16%
towards a goal can help people harness their inner 14%
resources (boost) and accomplish their goal. For example, 12%
% supporting tax
Rogers and Bazerman (2007) show that individuals are 10%

more likely to select choices that serve their intentions 8%
6%
when the choices will be implemented in the distant rather
4%
than the near future (See figure on Determination)
2%
0%
Near future Distant future
Source: Rogers and Bazerman (2008).
34 : T H E BB AASSIICC GGUUI IDDEEBBOOOOKK
Ethical considerations for STRATEGY
Some behavioural insights rely on mechanisms that Prioritise transparency. Is your

are not fully accessible to consciousness or under intervention clearly communicated,
people’s conscious control, while others involve including being transparent about its
counter-intuitive and theoretical insights whereby purpose and nature?
moral intuitions are not well adapted.
Offer a way out. Can citizens avoid the intervention?
To ensure the responsible use of BI in public Does the intervention offer easy pathways to
policy, you should always evaluate the morality objections and complaints?
of a policy strategy with regards to transparency
and “avoidability”. Transparent insights are when Ensure the policy intervention serves the public
citizens can identify: i) who is trying to influence interest. Is it in line with public sentiments? Does it
them; ii) what this means; and iii) what purpose prevent harm against others?
is being achieved. Alternatively, behaviours
that people cannot control are referred to as Ensure citizens are not being held responsible for
unavoidable, while those that make possible or consequences that they did not consciously select.
depend on conscious control are referred to as In your context, are they able to fully understand the
avoidable. implications of their choices? Are they considered
legally accountable for these?
When assessing the transparency and
“avoidability” of policy intervention, keep in mind
the following considerations:
S T ASGTEA G4 E– 3I N
– TSETRRVAETNETGIIOE N
S : 35
Stage 4: Intervention... After developing your BI strategies, the next

stage is to actively participate in the design of
an intervention that will test which strategies
are the most effective to reach your policy
outcome. Your understanding of the context and
relationships with key stakeholders are essential
to identify where, when and among whom the
intervention can take place. In parallel, this is
another opportunity to rely on behavioural experts
who have rigorous research skills and subject
matter expertise to lead the intervention design,
consult on its implementation and conduct analyses
of the results.
At this stage, you can work with stakeholders and

behavioural experts to choose which BI strategies
are the most relevant because testing too
many strategies can become expensive and
messy. To increase the likelihood of a successful
1. Preliminary considerations for your behavioural intervention, below are some preliminary
intervention considerations to discuss with your key stakeholders
and behavioural experts. This will enable you to align
2. Ways to know “what works” expectations and make necessary changes early on
in the experimental design.
Preliminary considerations for your intervention

✓ Define success. There no single definition for a successful ✓ Be realistic about the timeline and budget. In
BI project. Take the time to understand what success addition to necessary approvals, consult the experts on
looks like for you, your stakeholders and the experts at how much time and resources are necessary for your
the start of your project to manage expectations. intervention to have its desired effect (or not). Learn
from stakeholders on what is necessary to generate
✓ Involve user-tests early on when piloting BI strategies. the desired sample size. Build this into your timeline and
When moving into the research and design stage, a first budget.
step is to involve users in testing aspects of the solutions
that the strategies give rise to. ✓ Secure legal, ethical and other approvals. Seek
necessary approvals required to carry out the
✓ Explore research designs. Seek consultation on whether experimental design. If you are working with external
an experiment or another design is the best fit for the experts understand what is required to obtain their
project. Explore digital platforms that may make it easier institutions’ approvals to consult on the project (i.e.
to randomly assign and deliver the intervention in a institutional review board).
cost-effective way. If testing in a real-life setting (field
trial) is not possible, discuss possibilities of conducting a ✓ Be transparent about data sharing and publications.
laboratory or online experiment. When engaging with external experts such as those
from academic institutions that value novel research and
✓ Know your sample size. The bigger the sample size, the scientific publications, it is important to be explicit about
better, but it comes at a cost. Work with stakeholders what data can be shared and the user-rights of results
to determine what you both find to be a meaningful for external publications.
difference. Consult the experts who can calculate how
many participants are necessary to achieve it and what ✓ Consider options for communicating results, long-
is the statistically meaningful difference. term monitoring and scaling. Outline a game plan on
how to move forward after data collection and analyses,
✓ Assess the risks of the intervention. Conduct a risks accounting for the possibility of null or negative results.
assessment with stakeholders and experts to manage Refer to the Change section for guidelines.
risks and potential unintended outcomes. This may
include a null result where no intervention has an effect.
S T A G E 4 – I N T E R V E N T I O N : 37
Ways to know “what works”

Randomised controlled trials (RCTs) to take part in the pilot. From there, you randomly assign 150
How do you know whether your new strategy works or schools to offer breakfast in the classroom (“the treatment
if other uncontrollable factors are driving your policy group”) and the other half to the normal practice of breakfast
outcome? An RCT is the ideal way to solve this dilemma. in the cafeteria (“the control group”) (Figure 2.6).
The key feature of an RCT is the use of a random
assignment to create at least two groups that closely The beauty of random assignment is that it eliminates
resemble each other. The only difference is that one group uncontrollable pre-existing or external factors
is exposed to the new strategy while the other does not. (i.e. improvements in public transportation) that may affect
By comparing identical groups, chosen at random, an RCT school attendance because the schools in the treatment
enables you to understand which strategies, if any, are group are not systematically different from those in the
working, and eliminate pre-existing or external factors that control group. At the end of your experiment when you
normally complicate the evaluation process. You may be compare school attendance, the difference between the two
more familiar with RCTs in medicine where RCTs are used to groups should only be a result of your intervention. If you
compare a new treatment with the current one or a placebo. are testing multiple strategies, it is still necessary to include
a control group to know how the new strategies did against
Often you may want to determine whether a new policy is the status quo. In the above example, you can see that four
more effective in achieving your measurable policy outcome schools saw an increase in school attendance while one
compared to the status quo. Say you implemented a city- school saw a decrease. How do you know if you are better
wide school breakfast programme with the aim of increasing or worse off under the new policy? The control group shows
school attendance. The current policy requires schools to us that only two schools saw an improvement while two
provide free breakfast in the cafeteria before school starts, schools saw a decline in attendance. So, you can say that
but student participation remains low. Your behaviourally the classroom breakfast policy led to an overall improvement
informed strategy is to “make it relevant” by changing the in school attendance compared to the cafeteria breakfast
placement and timing. Your new policy is to offer breakfast in policy. If the new policy was no better than the current one,
the classroom at the start of the school day. you would have seen a similar pattern in both the treatment
and the control groups.
To test the effectiveness of the new strategy against the
existing policy, you identify 300 schools (“sample size”) who
already run the school breakfast programme and have agreed
FIGURE 2.6. Simple randomised controlled trial (RCT) Alternative to RCTs

example RCTs are the ideal way to determine if your intervention led
Visualisation of an RCT to test the effect of the new school to your desired policy outcome but random allocation may
breakfast policy on school attendance not always be logistically, ethically or politically feasible. In
this case, consult a behavioural expert about conducting a
SAMPLE
quasi-experiment – an experimental approach that is similar
to an RCT, but participants are not randomised into control
and treatment groups (Campbell and Stanley, 2015). You can
no longer eliminate pre-existing or external factors that may
influence your outcome but you can still generate evidence
R A N D O M I S AT I O N
to learn which strategies are more effective. Examples of
valuable quasi-experimental designs include:
GROUP 1 GROUP 2
l Regression discontinuity (RD): where participants are
assigned to treatment and control groups based on a
cut point of an assignment variable. The discontinuity
between the treatment and control trends is then
measured.
l Propensity score matching (PSM): where participants

in the treatment group are paired to participants in the
Intervention No Intervention
No change in school control group based on the similarity of their scores to
attendance
account for selection bias.
T R E AT M E N T CONTROL
GROUP GROUP
l Difference in differences: where the effect of a
Decrease in school treatment or of a policy is estimated by comparing the
attendance pre- and post-treatment differences in the outcome in the
treatment and control group.
Increase in school
attendance
S T A G E 4 – I N T E R V E N T I O N : 39
Ethical considerations for INTERVENTIONS
Be aware that interventions unavoidably Always consult experience. Make

intervene in people’s lives. Experiments intentionally sure that experiments are conducted
give one group a treatment that is believed to have by people with experience in experimental
a positive impact, while withholding this treatment design, intervention and reporting to ensure proper
intentionally from another group. Consider three protocols are followed.
sources for help ensuring interventions are carried
out ethically: i) ethical codes of conduct; ii) informed Ensure justice, fairness and distributional impacts
consent; and iii) institutional review boards. These are considered. You need to consider and address
sources may not be uniformly applicable to all the potential ethical issues that arise from one group
experimental research. You must orient yourself receiving treatment, and the other not. This may
within the standard ethical guidelines and codes require deploying safety valves for discontinuing the
that fit into the special circumstances of the specific experiment for ethical reasons or compensating/
behavioural interventions. offsetting groups after the experiment.
Consider whether legal permission is required and Take all measures to protect confidentiality and
demonstrate the necessity of the experiment. ensure ethical data analysis. You should carefully
You should consider if the laws in your country deem consider using procedures and protocols that ensure
experimentation as legally permissible in public the confidentiality of participants’ responses, e.g. by
service. It may also be necessary to demonstrate that using randomised response methods or determining
the intervention will improve a policy situation, reveal not to collect or connect any data about potential
knowledge not currently known, provide necessary identifiers. Ethical data analysis can be strengthened
data, be used to inform policy and protect the rights by pre-registering studies, over accounting for
of individuals. data outliers and truthfully reporting on attrition,
to strictly follow standards of statistics and their
representation.
Stage 5: Change Your team has implemented the Intervention

and examined the findings – now what? You
want to avoid stopping at the proof-of-
practice stage if indeed the findings suggest
that the intervention could have an impact.
The fifth and final stage, Change, is the time for
you to look back at your BI exercise and think of
its long-term implications.
By this stage, you will know whether the

tests have produced promising results and a
behavioural insight can be scaled up into a full
policy intervention – or whether repeated failure
brings the project to an end and the community
can learn from what did not work. Change is
a crucial decision point for policymakers – at
this time, you really have the decision power to
determine which direction to take the project.
The goal of Change is to inform public policies

on the basis of the results of the project and to
ensure that society gains the broadest possible
value from the insights of the experiment. BASIC
suggests five final considerations on this note:
S T A G E 5 – C H A N G E : 41
1. Revisiting the political context and project level. 4. Maintaining the policy initiative. While crucial to
Are the interventions aligned with the current context avoid watering down behavioural policies, sometimes
such as political, and technological, state of affairs? maintenance of BI interventions may be neglected.
Are the results sufficiently promising to motivate policy This can happen because BI features may appear as
action, in light of the current legal landscape? Would the unimportant or may be in conflict with what seems
intervention be in line with public opinion or feasible in the necessary from a more rational perspective. Therefore,
current climate? Are the resources in place to implement it is important to have instructions for the proper
the intervention in the long term? maintenance – physical or systemic – of the policy. To
avoid problems with maintaining a policy initiative over
2. Implementing and scaling behaviourally informed time, practitioners should consider what audiences need to
policies. How can you best implement and scale up your be involved in the maintenance and produce material and
intervention? Should you change a law, regulation or instructions that fit these audiences and the situations in
regulatory regime (OECD, 2016)? Is it possible that the which this material is to be used (Hansen, 2018).
results might fail to generalise when scaled? Keep in
mind that through the implementation and scaling up of 5. Disseminating knowledge widely. While the idea of
a behaviourally informed policy, certain groups could be disseminating results is common in the behavioural
more or less affected than what was suggested by the community, it is still not so in most public institutions.
intervention. This, in turn, may lead to further iterations As a result, many early BI projects were not reported at
and tweaks in the design of the policy in question. all or only for internal use. In particular, null results have
not been widely publicised leading to publication bias.
3. Setting up monitoring of long-term and potential Likewise, the lack of standards has led to non-transparent
side effects. Because most BI experiments are limited in reporting; reporting without moderators; reporting only
time span, plan to have specific plans for monitoring the in local languages; overstatement of effects, savings
scaled-up policy in the long term as well as its potential and revenues; and understatement of true costs (for
side effects. This may be done by integrating an ex post examples, see OECD, 2017a, and Osman et al., 2018). For
evaluation or review of a given policy as a required step this reason, it is crucial that researchers and practitioners
of the policymaking process. In this way, evaluations or participate, support and systematically share and report
reviews will help ensure the quality of policy over time. their work in national as well as international networks of
both researchers and policymakers.
Ethical considerations for CHANGE
Adhere to principles of proper stakeholder Report on what works and what

engagement. Make sure to involve public bodies, does not. Other policymakers, society
staff, citizens, businesses and other affected parties and the research community should learn
are properly consulted and the results of this from your efforts. This includes reporting on null
consultation are clearly communicated. results and unexpected effects to avoid exposing
citizens to interventions that have already been
Follow principles of transparency and shown to fail.
accountability. Results of experiments and
consultations should be shared with executive and Monitor long‑term and side effects. In implementing
legislative branches, as well as with broader society. behaviourally informed interventions, you also have
This includes ensuring proper credit is given to the the responsibility for developing plans for monitoring
policymakers and government agencies who ran the the effects of the interventions to protect citizens
experiments. from the potential negative consequences.
C O N C L U S I O N S : 43
CONCLUSIONS
What are the ingredients of good behavioural This introductory guide provides a tool for
public policies? When and how should BI be incorporating these insights when designing and
applied to policymaking? Can we use our implementing public policies, by overviewing the
knowledge of how people think and make basic steps and best practices of behavioural
decisions to create more effective interventions? insights. Once you have gained an initial
The BASIC Guidebook addresses these understanding of these through the guide, we
fundamental questions and provides a starting recommend referring to the BASIC Manual for a
point to learn about the process of incorporating more detailed and comprehensive depiction of the
BI into policymaking. BI approach, that will allow you to fully grasp the
rationale and mechanisms of the BASIC framework.
The guide offers an introductory look at how
to identify, scope and address the behavioural
aspects of a policy problem and emphasises the
importance of applying BI from the beginning to
the end of the policy cycle. It aims to show how a
nuanced understanding of human behaviour can
provide new tools for policymakers. Introducing
subtle changes to choice of architecture, Identify
Behaviours
Analyse
Biases
Design
Strategies
Test
Interventions
Scale
for Change
factoring in people’s social preferences, working

to facilitate people’s decision-making, can all be
powerful drivers of behaviour change. A more detailed how-to manual of the BASIC Toolkit is
available on the OECD iLibrary: http://oe.cd/BASIC
REFERENCES
Allcott, H. (2011), “Social norms and energy conservation”, Gifford, R. et al. (2009), “Temporal pessimism and spatial
Journal of Public Economics, Vol. 95(9-10), pp. 1082-1095, optimism in environmental assessments: An 18-nation
http://dx.doi.org/10.1016/J.JPUBECO.2011.03.003. study”, Journal of Environmental Psychology, Vol. 29,
pp. 1–12.
Campbell, D.T. and J.C. Stanley (2015). Experimental and
Quasi-experimental Designs for Research, Ravenio Books. Hallsworth, M. et al. (2015), “Stating appointment costs
in SMS reminders reduces missed hospital appointments:
Chetty, R., A. Looney and K. Kroft (2009), “Salience and findings from two randomised controlled trials”,
taxation: Theory and evidence”, American Economic PLOS One, Vol. 10(9),
Review, Vol. 99(4), pp. 1145-77. https://doi.org/10.1371/journal.pone.0137306.
Dallas, S.K., P.J. Liu and P.A. Ubel (2019), “Don’t count Hertwig, R. (2017), “When to consider boosting: Some rules
calorie labelling out: Calorie counts on the Left side of for policy-makers”, Behavioural Public Policy, Vol. 1(2),
menu items lead to lower calorie food choices”, pp. 143-161.
Journal of Consumer Psychology, Vol. 29(1), pp. 60-69.
ideas42 (2018), Using Behavioral Science to Improve Criminal
Doran, G.T. (1981), “There’s a SMART way to write Justice Outcomes Preventing Failures to Appear in Court,
management’s goals and objectives”, Management https://urbanlabs.uchicago.edu/attachmentsstore/9c86b
Review, Vol. 70(11), pp. 35-36. 123e3b00a5da58318f438a6e787dd01d66d0efad54d6
aa232a6473/I42-954NYCSummonsPaperFinal_Mar2018.
European Commission (2014), “Taking consumer rights into pdf.
the digital age: over 507 million citizens will benefit as
of today”, Press Release, http://europa.eu/rapid/press- ideas42 (2017), Define, Diagnose, Design, Test,
release_IP-14-655_en.htm (accessed 7 November 2018). http://www.ideas42.org/blog/first-step-towards-solution-
beta-project/ (accessed 6 November 2018).
Gibbons, S. (2018), Journey Mapping 101, Nielsen Norman
Group, https://www.nngroup.com/articles/journey- IRS (2017), Behavioral Insights Toolkit, US Internal
mapping-101/. Revenue Service, https://www.irs.gov/pub/irs-
soi/17rpirsbehavioralinsights.pdf.
R E F E R E N C E S : 45
Johnson, E.J. and D.G. Goldstein (2004), “Defaults and OECD (2016), Protecting Consumers through Behavioural
donation decisions”, Transplantation, Vol. 78(12), pp. 1713–6. Insights: Regulating the Communications Market in
Colombia, OECD Publishing, Paris,
Kahneman, D. (2011), Thinking, Fast and Slow, Macmillan. http://dx.doi.org/10.1787/9789264255463-en.
Kahneman, D. et al. (1990). “Experimental tests of the OECD/OPSI (n.d.), Practical Service Blueprinting – The
endowment effect and the Coase theorem” Journal of Guide, Practical Service Design, OECD Observatory for
Political Economy, 98(6), 1325-1348. Public Sector Innovation, https://oecd-opsi.org/toolkits/
practical-service-blueprinting-the-guide/.
King, D. et al. (2014), “Redesigning the ‘choice architecture’
of hospital prescription charts: A mixed methods study Nudge blog (2010), Measuring the LSD Effect: 36 Percent
incorporating in situ simulation testing”, Improvement, http://nudges.org/?s=lake+shore+drive
http://dx.doi.org/10.1136/bmjopen-2014-005473. (accessed 7 November 2018).
Lo, A.W. (2013), “Fear, greed, and financial crises: A cognitive Osman, M. et al. (2018), “Learning lessons: How to practice
neurosciences perspective”, in Handbook of Systemic nudging around the world”, Journal of Risk Research, pp.
Risk, Cambridge University Press. 1-9, http://dx.doi.org/10.1080/13669877.2018.1517127.
Ng, C.F. (2016), “Behavioural mapping and tracking”, Rogers, T. and Bazerman, M. (2007). Future Lock-In: Future
Research Methods for Environmental Psychology, Vol. 29. Implementation Increases Selection of ‘Should’ Choices.
SSRN Electronic Journal.
OECD (2017a), Behavioural Insights and Public Policy:
Lessons from Around the World, OECD Publishing, Paris, Sunstein, C.R. (2003), “Terrorism and probability neglect”,
http://dx.doi.org/10.1787/9789264270480-en. Journal of Risk and Uncertainty, Vol. 26(283), pp. 121-136.
OECD (2017b), Obesity Update 2017, OECD Publishing, Paris, Svenson, O. (1981), “Are we all less risky and more skilful
http://www.oecd.org/health/health-systems/Obesity- than our fellow drivers?”, Acta Psychologica, Vol. 47(2),
Update-2017.pdf pp. 143-148,
http://dx.doi.org/10.1016/0001-6918(81)90005-6.
Thaler, R.H. and S. Benartzi (2004), “Save more tomorrow︎: The Behavioural Insights Team (2010), MINDSPACE,
Using behavioral economics to increase employee The Behavioural Insights Team, https://www.
saving”, Journal of Political Economy, Vol. 112(1), behaviouralinsights.co.uk/publications/mindspace/
University of Chicago. (accessed 11 April 2019).
Thaler, R.H. and C.R. Sunstein (2009), Nudge: Improving Wilcox, K. et al. (2009), “Vicarious goal fulfilment: When the
Decisions about Health, Wealth, and Happiness, Penguin mere presence of a healthy option leads to an ironically
Books, New York. indulgent decision”, Journal of Consumer Research,
Vol. 36(3), pp. 380-393.
The Behavioural Insights Team (2018), Applying
Behavioural Insights to Labour Market, The Behavioural World Bank (2015), The World Development Report
Insights Team, https://www.bi.team/publications/ 2015: Mind, Society and Behaviour, World
applying-behavioural-insights-to-labour-markets/ Bank, Washington, DC, http://www.worldbank.org/
(accessed 6 April 2019). content/dam/Worldbank/Publications/WDR/WDR%20
2015/WDR-2015-Full-Report.pdf (accessed 11 April 2019).
The Behavioural Insights Team (2014), EAST Four Simple
Ways to Apply Behavioural Insights, The Behavioural World Obesity Federation (2015), About Obesity, World
Insights Team, https://38r8om2xjhhl25mw24492dir- Obesity Federation, England and Wales.
wpengine.netdna-ssl.com/wp-content/uploads/2015/07/
BIT-Publication-EAST_FA_WEB.pdf
(accessed 11 April 2019).
The Behavioural Insights Team (2013), Test, Learn,

Adapt: Developing Public Policy with Randomised
Controlled Trials, The Behavioural Insights Team,
https://38r8om2xjhhl25mw24492dir-wpengine.netdna-
ssl.com/wp-content/uploads/2015/07/TLA-1906126.pdf
(accessed 11 April 2019).
N O T E S : 47
NOTES
OECD work on Behavioural Insights:
www.oecd.org/gov/regulatory-policy/behavioural-insights.htm
OECD work on Regulatory Policy:

www.oecd.org/gov/regulatory-policy
OECD Network of Economic Regulators (NER)

www.oecd.org/gov/regulatory-policy/ner.htm
For more information, please contact:

James Drummond
[email protected]
Nick Malyshev
[email protected]
OECD work on Behavioural Insights:
www.oecd.org/gov/regulatory-policy/behavioural-insights.htm
OECD work on Regulatory Policy:

www.oecd.org/gov/regulatory-policy
OECD Network of Economic Regulators (NER)

www.oecd.org/gov/regulatory-policy/ner.htm
For more information, please contact:

James Drummond
[email protected]
Nick Malyshev
[email protected]

BASIC Toolkit Web

Uploaded by

Copyright:

Available Formats

BASIC Toolkit Web

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

BASIC Toolkit Web

Uploaded by

Copyright:

Available Formats

THE

TOOLS AND ETHICS FOR

learn about the process of incorporating BI

The toolkit presented here guides the policy maker through a

l The BASIC Brief: gives a quick overview for senior

l The BASIC Guidebook: provides a practical instrument

scoped and addressed.

A more detailed how-to manual of the BASIC Toolkit is

TABLE 1.1. Examples of behaviourally informed strategies using the “ABCD” framework

It is easy to make unhealthy choices even when you choose

This is where behavioural insights (BI) can help. BI is a tool

To integrate BI into your day-to-day work, you can use

1. Behaviour, Analysis, Strategic, Intervention and Change are the five

TRADITIONAL POLICY INSTRUMENTS BEHAVIOURALLY INFORMED POLICY INSTRUMENTS

FIGURE 2.1. The BASIC Framework

crucial behavioural behaviours through conceptualise to evaluate the implementation, scale,

As you read through the BASIC Guidebook, you will gain

A practical and in-depth look into the first three sections,

TABLE 2.2. Considerations before applying Behavioural Insights (BI)

Problem-solving method Silver bullet

Way to learn “what works” One-size-fits-all

Beyond nudging 1.0 Only for behavioural experts

Policy tool Irrationality

Ethical considerations for APPLYING BEHAVIOURAL INSIGHTS

Stage 1: Behaviour... BI is more effective the earlier you integrate

1. What are the behaviours driving the policy

2. Which behaviours should you target?

3. What is your desired outcome?

4. What is the context shaping your target

1. What are the behaviours driving the policy issue?

FIGURE 2.2. Example of a behavioural reduction

1. Write the policy area ENERGY

2. Draw relevant strategic Workplace Residential

2. Which behaviour(s) should you target?

TABLE 2.3. Sample questions for the priority filter questionnaire

Importance Is a change in behaviour an institutional priority?

Impact Will changing the target individual behaviour translate to a significant

Feasibility Is it politically feasible? Are resources available? Is it controversial?

3. What is your desired policy outcome?

The outcome is to increase savings by 20% upon receipt of income among

4. What is the context shaping target behaviours?

Symbols and sample Arrows that connect A decision which

MAP THE “ACTUAL” BEHAVIOUR

Ethical considerations for BEHAVIOUR

1. Introduction to “slow” and “fast” thinking 4. Choice

1. Introduction to “slow” and “fast” thinking

Familiarise with the behaviour Flexible methodologies

TABLE 2.5. Overview of ABCD framework

HOW ARE ATTENTION BIASES RELEVANT TO POLICYMAKING?

HOW ARE BELIEF FORMATION BIASES RELEVANT TO POLICYMAKING?

Behaviourally informed solutions test possible choice

HOW ARE CHOICE BIASES RELEVANT TO POLICYMAKING?

Sources: European Commission (2014); OECD (2016).

As the Strategies stage suggests, behaviourally informed

HOW ARE DETERMINATION BIASES RELEVANT TO POLICYMAKING?

Ethical considerations for ANALYSIS

Behavioural analyses usually observe or study Consider what guidelines must be

Translating your analysis to actionable

MAKE IT RELEVANT SEIZE ATTENTION