Behavioural insights and the policy process

Activity 1.3

Behavioural insights and the policy process


Behavioural Insights (BI) is the study of human behaviour and draws from psychology, cognitive sciences, and social sciences including economics. Applying BI to policy helps you better understand how biases and behaviours shape people’s choices, and therefore how to better develop policy to engage and change behaviours. Even if you cannot apply BI straight away to your policy cycle, it is an important and growing area to know about.


How is Behavioural Insights useful for gender analysis?


BI helps policy professionals understand how context and behaviour biases affect people’s choices. These can include gender social norms, as well as biases directed toward ethnicity, age, disability or indigeneity. These can influence the effectiveness and implementation of your policy. By applying BI, and empirically testing solutions, policy professionals can develop innovative, targeted and measurable policy that supports gender equality.

How is Behavioural Insights useful for gender analysis?

Our beliefs and attitudes shape our action and decisions. These can shape biases and stereotypes that can be explicit (we are aware of how these shape our behaviour) or implicit (we are unaware of these shape our behaviour).

Biases not only shape our individual behaviour, but also shape institutional behaviour including processes and practices that are taken-for-granted or seen to be neutral.

Applying BI requires time, expertise and resourcing but can help you understand more clearly what is driving behaviour, the underlying biases and stereotypes that shape action and decisions, and create policy to counter these and support a fairer society.

There are many existing frameworks that can be used to guide your planning and thinking around BI. The OECD’s BASIC toolkit is one framework. As an introduction, the five BASIC steps to applying BI to your policy are outlined below, starting at the beginning of the policy cycle:

1. Behaviour

The first step is to identify behaviours to determine those that are relevant for your policy project, and to decide which behaviours to target that might have best potential. This will help you identify your policy outcome and next steps.

2. Analysis

At this step you learn about the behaviours you are targeting and why people make these choices. This is evidence-based using a range of theories or methodologies. The ABCD framework in the BASIC toolkit will help focus your thinking and methodological approaches around investigating four key areas of behaviour,

  • Attention: People’s attention is limited and easily distracted.
  • Belief formation: People rely on mental shortcuts and often over/under-estimate outcomes and probabilities.
  • Choice: People are influenced by the framing, social and situational context of choices.
  • Determination: People’s willpower is limited and subject to psychological biases.

3. Strategy

Based on step 1 and 2, this step supports developing policy interventions that are targeted and attractive for people to change their behaviour.

4. Intervention

Before implementing solutions it is important to empirically test them to help you understand the most effective ways to influence the behaviours you are targeting. This includes user testing, engaging with experts, and considering stakeholder views. Not all solutions will work when tested but this saves time and resource at a later date.

5. Change

Finally, you can check if your initial assumptions about behaviour were correct (before rolling out your policy); and consider how you can best implement and scale up your chosen solution. This may require new legislation, or changes to existing legislation. Make specific plans for monitoring policy changes over time.

These steps are detailed in full in BASIC document, showing the complexity and payoff that BI can deliver to policy, challenging the more standard approach.

Case study: Understanding the impact of biases for indigenous women in Bolivia to obtain microloans

Researchers in Bolivia examined whether microloan applications were approved based on the ethnicity and/or gender of potential borrowers.

The study included 70 credit officers from six Bolivian microfinance institutions. Participating credit officers were given similar credit applications from four potential borrowers: an indigenous man, a non-indigenous man, an indigenous woman, and a non-indigenous woman. The ethnicity of applicants was indicated by including photos and surnames in the applications. Information about the (equal) payment capacity of the potential borrower was also included.

The study found that non-indigenous women had a higher probability of receiving preference for a loan than indigenous women. Non-indigenous women had twice the chance of obtaining a loan compared with men, whereas Indigenous women had 1.5 times the chance of obtaining a loan compared with men.

The authors conclude that affirmative actions towards women must not overlook ethnicity as an important factor for financial policies, and that policies need to be shaped to target and reduce certain biases and structural barriers for indigenous women.

Gonzales Martinez, Rolando, Gabriela Aguilera‐Lizarazu, Andrea Rojas‐Hosse, and Patricia Aranda Blanco. "The interaction effect of gender and ethnicity in loan approval: A Bayesian estimation with data from a laboratory field experiment." Review of Development Economics 24, no. 3 (2019): 726-749.

© Copyright APEC 2022