New frontiers: The holistic impacts of nudging

Researchers have identified a series of wider-reaching holistic effects from nudging and steering behaviour which can have a significant impact on the overall outcome of a nudge or intervention. Having a broader understanding of the effects of a nudge on all related behaviours – not just the behaviour being nudged – is important to evaluate if it has a positive outcome on overall behaviour.

This is particularly relevant for some behavioural change interventions, notably behaviours which are repeated frequently over the long term, such as household energy consumption, taking daily medication or exercise. Without this wider understanding an intervention might, at first glance, seem like it has had the desired impact, giving us a false sense of success.

Over the past decade, behavioural scientists have identified five different holistic effects which can all impact on the overall effectiveness of a behaviour change intervention. Some of these effects or concepts can be positive, whereas others may end up neutralising the effect of any nudge, or worse, having a negative impact:

1. Licensing effects
2. Compensating effects
3. Positive spillover effects
4. Displacement effects
5. Systemic effects or what we are calling ‘nudge fatigue’

Where we were

Behavioural scientists have been aware of the first three of the five for some time now and they are frequently noted and measured.

Licensing and compensating effects: Sometimes people may ‘nudge back’ in a series of connected behaviours throughout the day or week, for instance, feeling permission to do something ‘bad’ because they have done something ‘good’ – a concept known as a licensing effect (see cartoon).

Image source: Sarah Lazarovic, 22nd October 2018, Behavioural Scientist 

Conversely, if we feel our behaviour has been less than exemplary we might try to compensate by doing something more worthy – a compensating effect.

Journalist Michael Rosenwald of the Washington Post sums up the concept of licensing and compensating effects particularly well. He says: “We drink Diet Coke – with Quarter Pounders and fries at McDonald’s. We go to the gym – and ride the elevator to the second floor. We install tankless water heaters – then take longer showers. We drive SUVs to see Al Gore’s speeches on global warming.” Each behaviour offsets the other, producing a net impact of zero.

Energy economist Lucas Davis observed licensing effects in energy consumption. He has shown that consumers who bought energy-efficient products increased their energy usage so far that they offset any potential gains. For example, buying an energy efficient washing machine led US consumers to increase clothes washing by 5.6% – which offset any potential energy-savings.^[ 1 ]

Positive spillover effects: Spillover effects occur when we initiate a change in one behaviour which then creates changes in other related behaviours. For example, researchers found that prompting hotel guests to pledge to reduce their impact on the environment by reusing their hotel towels led to spillover effects during their hotel stay. Guests were not only more likely to reuse their towels, but that they were also more likely to carry out other more environmentally friendly behaviours such as turning the lights off. So, the overall behavioural impact might be much more powerful than the initial objective.^[ 2 ]

Current developments

On top of this, researchers have been adding to the collection of holistic effects as they design and run increasingly varied trials and studies and investigate potential secondary outcomes to better understand new facets of applied behavioural science. As the application of behavioural science grows and becomes more widespread, new areas are being discovered.

Displacement effects: This can occur when we appear to ‘nudge’ behaviour successfully but, in fact, have merely displaced the behaviour, shifting the location or time in which it is performed. This ultimately results in a ‘zero impact’ outcome in terms of behaviour change.

A fascinating and recent example of a displacement effect comes from research by the UK regulator, the Financial Conduct Authority (FCA) on credit card repayments. Credit card debt is a growing concern in the UK. As of May 2019, credit card debt now stands at nearly £73bn across a total of 30 million card holders – nearly half the population.[ 3 ] Each household in the UK now has an average of £2,688 unpaid on credit cards.[ 4 ] Around 3.3 million credit card holders are in what’s been termed ‘persistent debt’ by the FCA. With this alarming situation in mind, the FCA wanted to test whether making changes to the design and message of someone’s credit card bill might nudge them into repaying debt faster.

It drew on research published in 2009 by Neil Stewart and other behavioural economists, which found that consumers seem to anchor to the minimum repayment figure on their credit card bill, using that as a reference point or guide for how much to repay.

Analysis by the FCA of people’s existing repayment behaviour also showed that many people usually set up a direct debit to repay the minimum repayment and then make ad-hoc manual repayments throughout the year to repay the principal debt chunk by chunk.

When the FCA trialled removing the anchor of the minimum repayment figure from people’s actual credit card bills, it was surprised by the results: typical repayment behaviour did change, but not as expected. Promisingly, one in five people switched from paying the minimum repayment to a higher fixed repayment, resulting in a total of nearly 50% of card holders making an automatic fixed payment. A few more people also chose to repay in full.

However, after seven months, the picture was less promising. As people’s credit card debt continued to grow, the required minimum payment caught up with the previous fixed payment some people had set so only seven in 100 people were now paying more than the minimum repayment. Overall credit card debt had also not declined due to two reasons:

• consumers offset higher automatic payments with lower manual, or ad-hoc payments; and partly because
• it discouraged more people from setting up a direct debit at all ( 24.9% rather than 19.8%), perhaps leaving them more susceptible to forgetting to make any repayment.^[ 5 ]

Nudging people to make higher automatic repayments displaced the ad-hoc manual repayments they had been making and the nudge even put some people off setting up direct debit repayments altogether. This finding shows why it’s important to track behavioural outcomes both over a sufficient amount of time and more broadly in comparable contexts. Christopher Woolard, executive director of strategy and competition at the FCA said: “This is something that we would never have discovered without choosing the right measures and monitoring them over a decent amount of time.” He advised: “choose your outcomes carefully. Unless we measure the precise outcomes we are aiming for, it is easy to see success where – in fact – there is none.”^[ 6 ]

Displacement effects have also been illustrated in an earlier field trial looking at flu vaccination uptake. At first glance, it appeared that automatically opting people in to receive a flu shot at a clinic – by sending them a letter notifying them of a scheduled appointment date – had a big impact on behaviour. 56% of patients who received the letter kept their given appointment as opposed to just 5% of a second group of patients who were instructed to phone and schedule their appointment. However, when researchers tracked the uptake in vaccination appointments over the longer term, they found that much of the uptake was explained by patients who usually received their flu vaccination from another clinic and had merely switched surgery.^[ 7 ]

In the trials above, the overall impact was neutral; credit card holders were not repaying their debt any faster following the intervention and no more patients were receiving their flu jab. In some cases, displacement effects might even have negative consequences in the long term; nudging in the short term might annoy people to the extent that they take action to ensure they can’t be nudged in the future.

In another study, Christina Gravert and Mette Trier Damguard partnered with one of Denmark’s largest charities to see if they could boost donations from existing donors on its mailing list. They tested whether sending an email reminder one week after the charity’s first email request for donations had any effect. While it did increase donations on that occasion, the number of un-subscriptions to the charity’s mailing list also increased from 2.1% to 3.7%. Moreover, those who unsubscribed were valuable donors who had previously donated to the charity.^[ 8 ] Therefore, in the long run, the charity lost out. Nudging in the short term may have cost them valuable donations in the longer term. Without accounting for these hidden costs, the charity may have overestimated the impact of the email reminders and assumed the nudge was successful.

‘Nudge fatigue’: The final effect is what we might call systemic effects – or perhaps ‘nudge fatigue’. Current interventions by behavioural scientists often involve things like extra pop-up computer menus, hover messages, micro-animation, feedback, mailings, additional post-it notes, financial and other incentives, or even signing pledges to put up on the wall.

Practitioners and academics are now recognising the potential impact of multiple nudges implemented at the same time. Given the growth of the science, we could soon get to a point where people are so overloaded with simultaneous nudges, none are salient or attention grabbing. Rather like the phrase ‘urban clutter’ used by traffic and city-design experts, we may end up with so much ‘nudge-clutter’ that the impact of a nudge could be muted.

For example, Peter Ubel, physician and behavioural scientist at Duke University makes a valid observation regarding the nudges increasingly used in the healthcare sector, questioning whether we may start to see ‘nudge fatigue’: “There are only so many pop-up windows, so many feedback letters, a physician can receive before ignoring all incoming signals.”^[ 9 ]

Additionally, ‘pressure selling’ has become ubiquitous in online retail. Go to buy airline, train or concert tickets, or make a hotel booking, and you’ll likely be bombarded by messages such as ‘Last booked 10 minutes ago!’ or ‘25 people currently looking at this route!’ which draw on concepts such as social norms and scarcity bias or the anticipation of loss and the fear of missing out.

While there is evidence that in isolation these types of messages can spur us into making a booking earlier and more quickly than we might otherwise have done, the proliferation of these messages on so many platforms is causing the tide to turn. Consumers are becoming numb to their original effects, especially if they are exaggerated – or as one recent study found, in some cases false and actually randomly generated.[ 10 ]

As the application of behavioural science becomes more widespread, it will be valuable to try to measure these longer term, broader systemic impacts. To evaluate any of these effects, interventions have to be designed in a way that allows them to monitor a wide breadth of behaviours, not just the immediate impact of a nudge – but also the longer term impacts over time and the impact on other related behaviours.   

Conclusion

There is a clear need to further research holistic effects, particularly systemic effects and displacement effects which are relatively new and not yet fully explored. As Professor Daniel Read says, “If we nudge one behaviour, we may be ‘un-nudging’ others”. All of these effects increasingly deserve our attention, both for how to design and effectively evaluate any intervention and for how to best shape and guide the application of behavioural science as it evolves and becomes even more frequently applied.

By The Behavioural Architects’ Crawford Hollingworth and Liz Barker

Reference:

[ 1 ] Davis, L. W. ( 2008 ) Durable goods and residential demand for energy and water: evidence from a field trial. The RAND Journal of Economics, 39 ( 2 ): 530-546.

[ 2 ] Baca-Motes, K., Brown, A., Gneezy, A., Keenan, E.A., and Nelson, L.D., â??Commitment and Behavior Change: Evidence from the Field Journal of Consumer Research, Vol. 39, Feb 2013

[ 3 ] https://www.fca.org.uk/publications/research/research-note-helping-credit-card-users-repay-their-debt-summary-experimental-research

[ 4 ] https://www.independent.co.uk/news/business/news/credit-card-debt-household-bank-of-england-consumer-november-uk-economy-a8712271.html

[ 5 ] Paul Adams, Benedict Guttman-Kenney, Lucy Hayes and Stefan Hunt, "Helping credit card users repay their debt: a summary of experimental research" Financial Conduct Authority, July 2018; https://www.fca.org.uk/publication/research/research-note-helping-credit-card-users-repay-their-debt-summary-experimental-research.pdf

[ 6 ] https://www.fca.org.uk/news/speeches/prisoners-wellness-programmes-and-rats-hanoi-why-fca-tests-its-interventions

[ 7 ] Chapman, G.B., Li, M., Leventhal, H., Wisnivesky, J., Leventhal, E.A. ( 2014 ). Defaults for influenza vaccination appointments in Judge, Nudge, Dodge presented at the SJDM conference, Nov 2014

[ 8 ] Damguard, M.T. and Gravert, C. "The hidden costs of nudging: Experimental evidence from reminders in fundraising" Journal of Public Economics, Volume 157, January 2018, Pages 15-26

[ 9 ] https://www.forbes.com/sites/peterubel/2016/07/15/what-behavioral-economic-interventions-get-wrong-about-improving-healthcare/#21c3a6f42ff2

[ 10 ] Mathur, A. et al ( 2019 ) Dark Patterns at Scale: Findings from a Crawl of 11k shopping websites