Some learnings from one of this year’s Behavioural Science-based Beesley Lectures

Written by Alex McCluckie, Associate Director. Contact Alex here

 

Behavioural science and the water sector are two, once disparate areas, that are slowly becoming better acquainted with each other. Given the great impact that behavioural science has been having over the past few years (as evidenced by the Behavioural Insights Team or “Nudge Unit”), this article’s first sentence I would hope will be encouraging to many that work in the industry. To try and investigate further how behavioural science could be leveraged for the good of the water sector, on Wednesday the 10th of October, I jetted off from Manchester down to the Institute of Directors in London to attend one of this year’s Beesley Lectures, succinctly titled: Behavioural science in water and energy markets: lessons for evidence-based policy making. 

 

The night was chaired by Dr Xeni Dassiou of City, University of London with the star attraction being a lecture given by Professor Robert Hahn from the University of Oxford. 

 

As appears to be customary at the conferences and speaking events I have attended of late, the session began with a quick overview of behavioural economics and its Nobel-winning popularisers, Kahneman and Thaler. After bringing everyone up to speed, however, Professor Hahn jumped straight into some meaty examples of nudges using real-world examples and demonstrated how insights from this exciting sphere have been harnessed across numerous areas within both water and energy.

 

Whilst it wasn’t an all-encompassing overview of what can influence behaviour, part of the session did provide an interesting examination of some of the ways that Hahn’s own field experiments have been altering behaviour, with the key tenet for this portion of the talk being that information matters! Indeed, as I have seen from my own work in the water industry, from asking customers about how acceptable they deem their water company’s proposed plans for the next five years to be, to their willingness to support a social tariff, how information is framed can make such a difference. Let’s take a look at some of Professor Hahn’s examples…

 

Some work undertaken on behalf of British Gas that sought to understand how best to get people to take up smart meters provided support for the view that neoclassical economics doesn’t always ring true. In this example, the researchers offered a randomly selected group of people a £5 voucher and a second randomly selected group of people a £10 voucher to take up a smart meter. Now firstly, it should be noted that these two monetary incentives resulted in a difference compared to the baseline (that is to say that the offer of a cash incentive did increase people’s likelihood to take up smart meters) - nothing overly surprising there I hear you say! What was interesting however was the finding that there was absolutely no difference between the amount of take up between the two amounts, something that is contradictory to what classical economics would teach us; that we should expect more people to take up smart meters with the higher incentive. Now, why is this first snippet of information interesting? Well, imagine you are the operations manager for your water or energy company who yourself is attempting to increase the uptake of smart meters. Falsely assuming that doubling the monetary incentive would double (or at least relatively increase) the take up of meters would lead to a tremendous waste of money that could otherwise be put to great use elsewhere in the business! Can you think of a better example of why experimentation when planning an intervention is crucial?!

 

I should make clear, behavioural science isn’t about throwing money at problems. It is about recognising that there are a number of mental biases that human beings tend to succumb to and that these biases can be utilised. One such way of doing this is by presenting people with information that is framed in ways that play on these biases.

 

For clarity’s sake, let’s take a closer look at two of these now:

 

• Loss aversion: is the principle that people react to losses more strongly than gains and they try to prevent losses more than they try to make gains.

 

• Social norms: is the principle that people want to be like everyone else and are heavily influenced by what they perceive everyone else is doing. 

 

These two biases have been utilised to influence people to behave in different ways across a range of sectors and industries and here Hahn demonstrated their power in the water sector specifically. 

 

The City of San Antonio, Texas wanted to motivate people to take up rebates for drought-resistant landscaping. To do this, they sent out letters that were framed in different ways but which utilised the aforementioned biases. For instance, asking people to take up an offer of drought-resistant landscaping because they are using more water than their neighbours uses a social norm frame which is different to asking people to take up an offer of drought-resistant landscaping because if they don’t, they will lose this offer and by the way this is what your neighbours are doing in terms of consumption, which uses both the social norm frame and the loss frame. Interestingly, the combination of a social comparison and loss framing led to a 36% increase in take up of the offer relative to a benchmark letter that had neither framing effect applied!  Now there are a whole host of other fascinating titbits from this study that time and space restraints won’t let me dive into here, such as how social norms alone may play better in some domains whilst loss frames alone may play better in others. However, I have included a link to the article here and I urge you to follow this up because it makes for some really interesting reading.

 

So what is the point here? Well, on the one hand, there is the fact that we know there are certain interventions that can lead to changes in people's behaviour and that the recognition and utilisation of various mental biases can help facilitate such changes. However, I think there is a broader point here and that is that experimentation and evaluation are key. Notice that the two examples I detail above each had a control plus various experimentation groups which due to the randomised design adopted allowed for a direct comparison of cause and effect that otherwise would have been lost. If you’re thinking this is something of a ‘stating the obvious’ type of moment, then spare a thought for California… 

 

In 2015 California was in the sways of a worrying drought. So much so in fact that the Governor, in an attempt to sort this problem out, set about trying to achieve a 25% reduction in water usage by, amongst other costly things, adopting a turf removal programme. Now, these initiatives resulted in the Governor largely meeting his goal, however, there was a problem; there was almost no evaluation or experimentation of the effectiveness of these various programmes undertaken. What this meant was that they were effectively blind as to what worked best and what would work best should a new crisis/drought strike. As Hahn pointed out, it seems that this episode holds true to the phrase coined by Stanford economist Paul Romer: a crisis is a terrible thing to waste! 

 

So, have we found a solution to the myriad issues currently facing the industry?

 

Through experimentation and evaluation, can we slowly begin to form a perfect industry operating in the most efficient and optimum way imaginable? Really, the answer is no, at least not yet. It has been recognised for some time that different nudges work on different margins (although why this is the case is something that we’re still trying to learn more about). The truth is, there is no complete theory of human decision making, but we are developing a toolkit that is growing stronger with time. 

 

Indeed, I am conscious that this article thus far has been relatively glowing about behavioural science’s possibilities and I feel that in the name of balance it is only right to cover some of the gripes brought to the fore by the audience too. In fact, given the buzz that often follows any discussion of behavioural science and its potential impact, it was interesting to hear a couple of audience members question its usage given that its effects are often quoted as delivering relatively small percentage changes in whatever the target behaviour may be. This even led to a discussion around whether or not behavioural science’s effects are at times oversold. Well, as Alex Plant of Anglian Water pointed out, small effects across a large population are probably still worth having and given that running such experimentation to establish how these small effects can best be brought about through skilfully crafted nudges is relatively inexpensive, they can still be well worth the effort.

 

Alex, in his role as respondent for the night’s discussion, did provide a well thought through response to Professor Hahn’s talk in which he held strong to the belief that we maybe shouldn’t be drawn moth-like to the shimmering flame of behavioural science at the expense of traditional economics altogether. Rather, he claimed, we should think about those key tenets of competition regulation theory that have served us well in the past as, he claims, companies will still tend to act in economically rational ways. Policies that recognise this along with incentive-based regulation will still drive efficiency and can drive innovation in ways that deliver for all customers regardless of underpinning sub-optimum choices that may be being made by customers at the individual level.

 

And so to conclude, born out of a realisation that neoclassical economics needed a fundamental re-think, behavioural science does offer insights into how people can be nudged into behaving in ways that are ‘good’, however that may be defined, whether that be for themselves, the environment or society. 

 

It is well recognised that nudges have been working well in numerous areas of society and I see there to be no reason why behavioural science, if deployed effectively cannot play a really important role in promoting outcomes that are in the public interest in the water (and energy) industry. 

 

Indeed, as we are faced with the growing challenges of population growth and climate change, the fundamentals of evidenced-based policy are only going to increase, in both their appeal and their importance from initial diagnosis, to experimenting with RCTs and studying your outcomes through adequate evaluation and then looping back round to diagnosis. So, whilst no panacea, armed with the growing learnings from behavioural science, we are certainly entering into an incredibly exciting time within the industry of which we are only just beginning to scratch the surface of what seems, at least for now, to be vast possibilities.