Behavioural Exchange 2017 Takeaways

Photo by Ken Bohn/San Diego Zoo

Some notes and thoughts on the annual conference on behavioural insights a.k.a. ‘nudging’

Last week, I had the privilege of volunteering at BX2017 as a liaison officer with the Civil Service College, and was able to attend most of the talks and sessions the conference held. The field of behavioural insights, at least under this title, is fairly new in the policymaking realm, and is defined by the OECD as:

“one discipline in a family of three, the others being behavioural sciences and behavioural economics, which mix traditional economic strategies with insights from psychology, cognitive science and other social sciences to discover the many ‘irrational’ factors that influence decision making” and aims “at improving the welfare of citizens and consumers through policies and regulations that are formed based on empirically-tested results, derived using sound experimental methods.”

BX2017 is the fourth annual behavioural exchange conference which strives to “showcase the best examples from around the world of applying behavioural science to public policy”. For those who did not have the time or the two grand to attend, I have distilled what notes I was able to take and some of my thoughts for you below.

Dean Karlan’s Keynote on Using Commitment

The conference began with Dean Karlan going over the fundamentals of behavioural insights for the uninitiated. The three he chose to focus on were:

  1. to watch what people do, and not what they should do, or say they will do,
  2. to keep interventions simple (as we can’t always get people in a state of deep reflection), and
  3. to remember evidence-based policymaking cannot be done without evidence.

He then talked about how hard is is to predict when we’ll have self control given individual differences and influences from a plenitude of circumstances. He defines commitment as voluntarily opting into some contract which alters the frequency of behaviour, and advocated for self awareness first, and then a focus on raising the price of vice as one the most effective way of attaining it on an individual level. Dean then shared various examples where BI was used to change behaviour through affecting vice, the last of which involving a website he helped to create, which allows individuals to make a ‘commitment contract’ to aid in achieving their desired goals.

Colin Camerer’s Keynote on how to Nudge the Habitual Brain

Colin Camerer’s talk discussed getting people weaned off incentives to habits in the process of constructing long term behavioural change.

He laid out a model that distinguishes between three brain modes: model-based, model-free, and habit. He then discussed what he thought were the two hallmarks of habits: the high costs of inertia, and the unavailability of habitized good. An interesting point he raises is that unavailability interrupting habit can motivate the search for better choices, giving the example of the southern rail strike in London leading some commuters to seek and transition to better travel routes with more time value savings.

Panel on Public Communication and Engagement

Jerril Retcher

Jeril Retcher shared the successes of VicHealth’s adoption of BI in the past year, in the areas of water consumption and unconscious recruitment bias. She highlighted stakeholder negotiation as a challenging but important part of rolling out BI trials.

Philip Ong

Philip Ong talked about current research in Singapore using BI, including MCCY’s active mobility trial and MHA’s attempt to increase participation in the Community Emergency Preparedness Programme (CEPP). Some of the techniques discussed seemed like a rehashing or reiteration of conventional psychology concepts like compliance techniques, but perhaps their usage as part of policymaking is fairly new.

There was also an interesting discussion started by a question regarding social media. While he cited the example of communities benefiting from forming WhatsApp groups to formalise themselves and allow for the emergence of leaders that are able to take charge of the groups, Philip described social media as a combination of the worst traits of the telephone and television: an echo chamber that leads to an overconfidence in one’s point if view and an inability to detect truth. He creatively suggested creating a Facebook button that could switch your feed to only display opposing or random viewpoints but lamented on the difficulty of creating something like that. Jerril was more optimistic and suggested that influencing individuals would lead them to spearhead these opposite views and dialogues on the feeds of their friends. Philip also talked about needing to find preexisting spaces where people mix and make them more vibrant or accessible.

Break-out Session on Understanding Bandwidth Tax

Commonly used in loan sharks’ harassing graffiti

This session was a slight departure from behavioural insights techniques and success stories, but rather allowed us to better understand the challenges those with lower income might face.

We played a few rounds of O$P$, a game developed by the Civil Service College, to elucidate and allow us to partially experience the compounding effects of having to deal with a scarcity of time and money over a period of time.

I thought that the topic of bandwidth tax had a place in the conference because it illustrates more human reasons why those who with deal with scarcity on an everyday basis might make poorer decisions overall, as compared to a ‘rational’ economic perspective which dictates that they are caused by calculated adjustments, or a purely meritocratic standpoint which suggests people have dug their own holes due to their own dispositions.

Panel on Lessons Learnt

The panelists’ presentations in this session were extensive and there were only a couple of minutes left in the session when the Q&A finally began. However, I thought this was one of the more productive sessions in the conference. Here are some of my key takeaways and thoughts:

Faisal Naru

  • Behvioural insights has been mostly applied in the implementation stage of the policy cycle, but other areas such as enforcement and agenda setting could benefit from its application as well.
Image from OECD
  • Behavioural insights can be applied beyond individual behaviours and to organizational behaviour (not only to individual behaviours within an organization, but to organizations’ cultures and structures).
  • Instead of solely focusing on pricing, the traditional economic indicator, focusing on behaviour after/besides pricing opens up more possibilities.
  • More effort needs to be placed in generating trust from the target of policymaking.
  • We should begin from the peoples’ needs instead of from an agency’s perspective (sounds familiar) and realize that collaboration across more agencies is required.

Kok Ping Soon

Kok Ping Soon discussed how behavioural insights has been around for a long time in different forms; case in point: Singapore’s Electronic Road Pricing (ERP) system, developed almost 20 years ago in 1998 to reduce traffic congestion. What has changed since then?

Satirical depiction of a little too many ERPs in Singapore; source unknown.
  • Problems today are more complex (such as an aging population and climate change)

This echoes for me Levin et al.’s characterization of the environmental problem as a super wicked problem, which necessitates less conventional approaches such as a behavioural insights one, along with applied forward reasoning as mentioned in Levin et al.’s paper.

  • Citizens are more demanding.
  • Tighter resources but increased demand for government intervention.

Ping Soon also introduced the formation of the Co-Lab, the behavioural insights and design unit within the Ministry of Manpower, and the acronyms SEED and HEART, which are being used by different organizations to build BI capacity:

  • Strategic leadership
  • Experimentation
  • Execution
  • Diversity
  • Hear them out
  • Make it Easy
  • Anticipate their needs
  • Respect every individual,
  • be Timely

Closing Panel on Frontiers and Cutting Edge Research

David Halpern

David Halpern took us through a number of examples of BI application that have progressed the field.

One interesting study he shared was on having prize draws as a default state was more effective than setting it up as a reward when trying to nudge a certain behaviour, challenging the reflexive slapping on of incentives to change behaviour, when there exists an arguably stronger force that is reciprocity. This isn’t new stuff. Cialdini’s well established principles of persuasion along with the entire field of social psychology point toward strong social forces that move people. But policymaking has long been dominated by economic frameworks and would benefit from consideration of these forces.

Like Mr Naru in the previous panel, Dr. Halpern also highlights the application of BI beyond individual nudges, as depicted in the case of taxing companies based on the sugar levels of their drinks. Because of the time frame of the tax implementation and the two tax tiers based on sugar level, while consumers are classically incentivized to not purchase drinks with sugar, companies are in fact being nudged towards reducing the sugar levels in their drinks by a tier and focusing marketing efforts on drinks with less sugar — further cascading the effects of the initial policy!

Figure detailing the two sugar tiers; taken from BIT’s report on sugar tax

Colin Cameron

Colin Cameron also shared his hopes and beliefs about BI’s future:

  • There is a need to draw from domain knowledge.

This echoes Faisal’s point on beginning from the people and engendering more collaboration between different fields/organizations.

  • RCTs are still the gold standard.

Critical global public health has something to say about this though. RCTs might not work for every population or scenario.

  • UI, UX, and design will feature prominently.

This goes back to the beginnings of BI, in which we see how small changes such as font size and word placement on a page can result in drastic changes in behaviour regarding things like tax payment rate.

  • We have to continue measuring aggressively, leveraging on improved technologies such as facial tracking, and the internet of things.

Colin ended his presentation emphatically by declaring our present time the “golden age of social science”, with the current tools and institutions in place to study and implement behavioural changes.

Final Thoughts

In his panel presentation, David Halpern quoted the following from E. O. Wilson as documented in Harvard magazine, which I believe makes an important case for the continued development and application of behavioural insights in policymaking. I have included the extended quotation below:

“The real problem of humanity is the following: we have paleolithic emotions; medieval institutions; and god-like technology. And it is terrifically dangerous, and it is now approaching a point of crisis overall.” Until we understand ourselves, concluded the Pulitzer-prize winning author of On Human Nature, “until we answer those huge questions of philosophy that the philosophers abandoned a couple of generations ago — Where do we come from? Who are we? Where are we going? — rationally,” we’re on very thin ground.”

Things like aggressive measurement sound a little scary, reeking of a sort of Huxlian manifestation of biopower. In fact, the whole business of subtly changing behaviours without really the consent of people does actually seem quite insidious and maybe tyrannical. However, it might not be that people’s choices are being limited or manipulated by the use of BI, but rather, that the use of BI can prevent people from falling prey to other forces that are perpetually exerting their influence, and often in a far less positive manner.

To loosely paraphrase Michael Maniates, an environmental studies professor at Yale-NUS College, choice editing only seems unethical until you realize that our choices are already and perpetually being edited by corporations that are primarily focused on their own interests, the design and architecture of spaces, the concentric circles of social structure, or the inherently political nature of technology. The adoption and growth of BI in Singapore’s policymaking bodies (gives me hope that I might find meaningful employment upon graduating with a Bachelors in psychology, and) is important in acknowledging and leveraging otherwise latent or ignored forces in everyday living for policymaking.

Also published here on LinkedIn.