This article is written by James Drummond, Policy Analyst, and Anna Pietikainen, Senior Policy Analyst, Public Governance Directorate, OECD.

Can behavioural insights (BI) be used to change the behaviour of organisations?

The rise in the use of BI around the world has shown the many ways governments can effectively influence individual behaviour to improve policy outcomes. However, many policy problems facing governments today involve addressing the behavioural drives of group behaviour, especially organisations.

At its core, safety culture is about an organisation’s values, beliefs, norms, practices, competencies and behaviours related to safety

This is especially true in the field of regulation. Economic regulation aims to promote market efficiency by companies and to create the best outcomes for customers. Technical health and safety regulation aims to protect citizens, workers and the environment across sectors of the economy. While individuals benefit from these forms of regulation, interventions are aimed at organisations or regulated entities, such as those that supply electricity, gas or water to households.

In the energy sector, evidence shows that many high-profile incidents have occurred — at least in part — due to poor organisational behaviour, including poor safety culture. For instance, one study analysed 15 major petrochemical accidents between 1980 and 2010 noted that poor safety culture contributed to 12 of the 15 accidents. Moreover, a review of literature showed that behavioural concerns have traditionally not been included as part of a well-rounded response by regulators to promoting safer outcomes.

Since 2013, the OECD Public Governance Directorate has been developing work on applying BI to regulatory policy. In April 2020, the OECD released its newest publication Behavioural Insights and Organisations: Fostering Safety Culture, which presents the full results and analysis of experiments with energy regulators testing the way organisations could be influenced by BI, particularly in safety culture.

This article looks at why we conducted this work, how we went about it, and what we found out.

Why apply BI to safety culture?

Despite the importance of safety systems and risk management for policymakers around the world, safety culture and the behaviours that drive it are relatively underexplored. While there is no concise internationally agreed-upon definition of safety culture, at its core, it is about an organisation’s values, beliefs, norms, practices, competencies and behaviours related to safety.

You can influence organisations via the people within them, including through organisational policies and procedures

We wanted to see if BI could be used to promote organisational cultures that could lead to safer outcomes. More broadly, we hoped that this could provide a possible methodology and lessons learned to encourage further experiments on changing organisational behaviour in regulated sectors.

Through its Network of Economic Regulators, the OECD worked with the Canadian Energy Regulator (CER); the Commission for the Regulation of Utilities (CRU) in Ireland; the Agency for Safety, Energy and Environment (ASEA) in Mexico; and the Authority for Electricity Regulation (AER) in Oman to test the application of BI to safety culture.

How did we implement the experiments?

At first glance, it may seem that applying BI to organisations is a different ballgame than applying BI to individuals — and sometimes it can be. However, research conducted by Industrial-Organisational psychologists suggests that “people make the place” and that you can influence organisations via the people within them, including through organisational policies and procedures. Where groups function similarly to individuals, many of the same behavioural insights could apply.

At the OECD we worked with a team of behavioural scientists from the London School of Economics and Columbia University to design experiments that would test applying BI to safety culture. Based on a literature review and scoping exercise with the regulators, three insights were chosen to study:

  • Messenger effect: People process information differently depending on whom we receive it from;
  • Social benchmarking/feedback: People pay attention to feedback in almost everything they do and often cannot adjust their behaviour without it;
  • Social norms: People take cues from their social and physical environment and care deeply about what their neighbours do.

An online questionnaire presented respondents with behavioural vignettes (scenarios) to test each of these insights, as well as perceptions of safety. A total of 1,033 respondents from regulators and regulated entities responded to the survey, broken down by 28 from Canada, 92 from Ireland, 409 from Mexico, and 504 from Oman. Given the small sample from Canada, caution is recommended on inferring conclusions in their setting. A second phase of experiments focusing on small scale gas and electricity installers was conducted with Ireland.

What did we find out?

If you are hoping for a neat picture of how humans behave in different contexts, we are sorry to disappoint. Human behaviour is influenced by social context, and the behavioural insights we tested performed differently in each country and context.

The first insight we can draw is that caution should be exercised when interpreting comparative results and in case of interest to replicate the experiments in a new context. The report notes all these findings in detail; however, three thought-provoking lessons can be drawn out for policymakers.

First, don’t take a one-size-fits-all approach to communication. We found that front-line workers in both regulated entities and regulators had the lowest perception of safety culture, while middle and senior management chains inside regulated entities had higher perceptions. These differences may influence the way people identify problems with safety and messages to each group should take this into account. This may make intuitive sense for many — in every workplace or organisation, people have different points of view formed based on their own experiences. Solutions to issues often rest in understanding and addressing these experiences. For policymakers, this suggests the need to take a similar approach as they seek to improve safety in their sector.

Behavioural insights intended for individuals can also be applied to influencing organisational behaviour, providing a new set of tools to support better regulatory design and delivery

Second, know who you are talking to and adjust your message accordingly. Results generally show that those with lower perceptions of safety seem to be more responsive to social norms, while average to high perceivers appear to be more responsive to feedback and messengers. However, this varied greatly by context — Mexico seemed highly influenced by descriptive norms throughout the perception levels, while Ireland responded better to messengers and Oman to feedback for average and high perceivers. The takeaway message for policymakers here is to understand clearly which insights drive actions for various members of a group and tailor policy responses accordingly.

Third, think about who is sending the message. Messages coming from someone of higher status but socially close (think: direct manager or trainer for front line staff, or regulators for senior managers inside regulated entities) were most effective. For policymakers, this suggests investing time in identifying, developing and utilising the most appropriate channels to communicate important messages — it may not be as obvious as it seems.

Behavioural insights intended for individuals can also be applied to influencing organisational behaviour, providing a new set of tools to support better regulatory design and delivery. However, there are caveats: results varied by context, organisation and people within them — demonstrating the need to always test in each context. As well, this field is understudied and the sample sizes varied greatly, affecting the design and results. Additional research is certainly needed to dive deeper and gain a better understanding of the behavioural drivers of safety. — James Drummond and Anna Pietikainen

Note: This work was carried out by the OECD Public Governance Directorate in collaboration with Mary MacLennan, behavioural science consultant and former member of the Impact and Innovation Unit, Canada; Morgan Tear, Research Fellow, London School of Economics and BehaviourWorks Australia at Monash University; and Daniel Shephard, PhD Candidate, Columbia University and former member of the Social and Behavioural Sciences Team, United States.

(Picture credit: Unsplash)

Make sure to share your own thoughts with the author by leaving a comment below