The concept of ‘Day Zero’ slipped into everyday language in October 2017 when the City released its Critical Water Shortages Disaster Plan, which spelled out clearly what the emergency rationing measures would be if the dams ran down to that critical threshold of 13.5 percent of usable water. This disaster plan turned out to be the single most effective intervention to bring about behaviour change from water users, as dam levels ran perilously low. Spelling out the urgency of the drought, and how imminent the emergency really was, seemed to send a shockwave through communities, and people were compelled to change how they used water.

Some considered the use of this term through media platforms as an irresponsible ‘fear-mongering’ tactic that led to panic. And yet, according to local behavioural economist Professor Martine Visser, the threat of extreme rationing measures worked well to change people’s behaviour.

Although in hindsight, Visser argues that a more moderate and transparent communication drive, started earlier on in the drought, might have been less costly for businesses and citizens, and would have built more trust.

Theories from the field of economics, about what drives behaviour, supports the idea that if people have clear information about how serious a threat is, but still have a sense of agency in terms of knowing what they can do to help tackle the problem, they’re more likely to act.

Visser is Director of the Environmental Policy Research Unit (EPRU) at the University of Cape Town’s School of Economics. When the drought peaked in late 2017, she and her team had already been tracking Capetonians’ water use behaviour from before the drought began and before anyone had an inkling of what was to come.

Early in the summer of 2015, the EPRU researchers were about to start a behavioural nudging experiment to see how people would respond to certain interventions that were geared towards encouraging water-wise behaviour (see ‘Nudging’ down water use through positive messaging, below). They began monitoring water use behaviour from the start of the experiment, which they ran through that summer. But once the experiment was done, they continued tracking people’s behaviour. As the drought worsened into 2017 and 2018, and as the City rolled out various other campaigns that were aimed at getting people to cut their water use, they could track how people responded.

Looking back over the course of the drought, the researchers could see what sort of interven-tions worked best to encourage behaviour change, and what didn’t.

The combination of interventions included the researchers’ behavioural nudging experiment, and the City’s various approaches to encouraging water use cuts, amounting to a series of carrot-and-stick approaches.

Some of these involved economic instruments like the ‘inclining block tariff’. This is a pricing system that charges low rates for the initial water that a household uses, but then increases the rate significantly after that, per block increase, so that higher water users pay much more per unit of water for their non-essential water needs.

Theories support the idea that if people have clear information about how serious a threat is, but still have a sense of agency in terms of knowing what they can do to help tackle the problem, they’re more likely to act.

Other interventions included water restrictions, such as a ban on hosepipes for cleaning cars or watering gardens, or setting limits per household, per month, or per person, per day. These were enforced through higher tariffs or fines. There were also threats of naming-and-shaming high water users who flouted restrictions, as well as the possibility of having water shut-off devices installed in their homes.

Once the drought was over, the researchers could compare how effective these different ap-proaches were. The lessons they’ve learned, they say, can apply to more than just public behaviour around water use during a time of drought crisis – they can also be useful to drive positive and sustainable household behaviour relating to energy use, or reducing household waste, or recycling.

‘Nudging’ down water use through positive messaging⁹

In the summer of 2015/16, Visser and the EPRU team ran a behavioural or ‘green nudging’ experiment to see if positive messaging could be a way to get Capetonians to be more water-wise. How well would this kind of messaging work, compared with the City’s other efforts to cut water use, such as the price hikes, water restrictions, fining, and so forth?

The economists drew up nine individual messages, each one framed in a different way. Each was generally affirming, subtle, and positive. Targeting 400 000 households, they slipped these messages into people’s utility bills that went out through the City’s mailing system over the course of the six summer months when water use generally goes up across the city. They targeted households through a random selection process. These were all freestanding homes, across different income brackets.

Some messages involved tips on how to cut water use at home. Others used graphics to better explain the block tariff pricing system, and then either showed people how much money they could save if they cut their water use, or how much it would cost them if they didn’t. Some messages tried to bring out the competitive streak in people, by showing in the bill how a family’s water use compared with others’ in their neighbourhood. Other messages used a value-systems approach to rally people around the idea of a ‘common good’, hoping to appeal to the idea that people are more likely to work cooperatively when they know it’s for the good of the wider community. One message appealed to being socially recognised: it promised people that they would be praised publicly for their water-wise behaviour, through having their names published on the City’s website.

The EPRU team tracked how the different groups responded to these messages, and com-pared these responses to those of a control group, another group of households that didn’t receive any messaging in their monthly bills.

Written in the data: what drives behaviour change

Visser and her team were already tracking water use behaviour before the drought started, and before anyone had even conceived of a Day Zero-type scenario. They started their behavioural nudging experiment in the summer of 2015, and then continued monitoring people’s water use behaviour once the experiment had run its course and as the drought tightened its hold into 2017 and 2018.

By the time the drought finally broke with the first rains of winter 2018, the EPRU crew had nearly three years’ worth of observations and data. They were able to look at the data from the City’s utility records, and see how households’ behaviour responded to the various interventions: the ‘green nudging’ experiment methods, the price hikes, water restrictions, fines, the public honouring, or the name-and-shame threats used by the City. In the end, they could also see how these methods compared with behaviour change that corresponded with the heightened sense of panic that came with the knowledge that Day Zero might actually happen.

What they found was a surprisingly positive trend from Capetonians: in just two and a half years, people reduced their water consumption by half, far faster than other cities that have faced similar crises in recent decades.

This is the kind of evidence-based research that Visser says needs to inform city-level policies around water and energy use, not only in a crisis situation, or in Cape Town, but for other cities in the country and the Global South.

Looking back, these are some of the take-home messages about how people responded:

‘Give us a gold star’

People want to be recognised for being water-wise. Those households who opened their monthly utility bills and found a message from the City saying that they would have their names published on the municipality’s website if they cut their water use, were more likely to do so.

‘Hitting us in our wallets doesn’t work’

Middle and higher-income households didn’t respond significantly to tariff increases, and the nudges that focussed on the high cost of excessive use didn’t have much impact on their behaviour, either. Basically, rich people don’t feel the pinch of the block tariffs; poorer households are usually low water users already, and can’t afford to pay for more than they currently use and so are unlikely to respond to this sort of messaging.

‘Bring on the bans’

Water restrictions worked well: banning hosepipe use, only allowing people to water their gardens or wash cars using buckets, insisting that private swimming pools be covered – these all brought on fairly significant responses from the public.

‘Give us easy-to-use info tools’

The City set up a few online tools that helped plug the information gaps (see Creating a ‘water literate’ city). The Water Dashboard and the Think Water website gave daily updates about dam levels, and were a key source of information on how the public could save water.

The Cape Town Water Map, which the City developed with EPRU as another ‘green nudging’ tool, was also designed around the idea that drawing on social comparisons and social recognition may be good ways to encourage behaviour change. The online map flagged households that stayed within the City’s target water levels per household, per month, by awarding them a green dot.¹⁰

The EPRU team wasn’t able to track how this map impacted on behaviour within households, as they had with the other nudging experiments, but their survey of people’s engagement with the map through conventional and social media suggests that it got plenty of attention. The Unpacking the Cape Town Drought: Lessons Learned analysis also shows how important the Water Dashboard and Think Water websites were, in terms of getting technically complex content out to an information-hungry public.

‘Name-and-shame’

As the drought worsened in 2017, the City released the names of the streets where the top 100 most wasteful household water users were, across the suburbs, threatening to name-and-shame those who ignored the restrictions. Next, the mayor sent out letters to individual households that were using more than 50kL of water per month, reprimanding them and asking them to lower their usage and behave in a more pro-social way. The letters also threatened them with the installation of water-restricting devices (see Water Management Devices). EPRU found that people responded quite positively to these letters, cutting their water use for up to seven months after receiving the letters.¹¹

‘It’s time to panic’: fear works, but with instructions on how to respond

When the City released the Critical Water Shortages Disaster Plan in October 2017, and with it the notion of Day Zero, Capetonians jumped into action: domestic water use dropped drastically.¹² This was the single biggest reduction across the two-and-a-half year drought period. The take-home message: sharing clear, salient information with citizens during times of crisis is crucial in driving behavioural change, particularly if it also gives people a sense of what they can do to help fix the problem.¹³

Rich people don’t feel the pinch of the block tariffs; poorer households are usually low water users already, and can’t afford to pay for more than they currently use and so are unlikely to respond to this sort of messaging.

The take-home message for cities in the Global South

‘Cities around the world face resource scarcity like this, not just in times of crisis. Most have limited funds to roll out campaigns that encourage people to use less water or electricity, or change their behaviour around household waste,’ Visser explains. Many cities have tried different models to drive behaviour change, but without real evidence on what causes these changes, it’s hard to attribute change to any specific drivers. What works in one city, can’t necessarily be copied directly into another city context. It may not work, may be a costly waste of time, or it may bring about unexpected negative consequences.

‘Using randomised controlled trials to test different models for behavioural change in the beginning of the drought allowed us to show which approaches worked amongst a diverse population of Capetonians,’ Visser says. ‘Now these can be used to target larger groups, and used in other cities.’

The take-home message about how to design these kinds of strategies is also clear: for any kind of behaviour change instrument or model or approach to be effective, a city needs to be transparent about what it’s going to do to deal with the crisis, it needs to show its data and models, and it needs to be clear in its message to the public. People want real information.

‘If a municipality wants people to change their behaviour, they have to give the public credible and trustworthy information,’ says Visser.

Researchers working with the City: easier said than done

The process of working with the City in order to do this kind of research highlighted various institutional, legal, and relationship challenges, which researchers now say are worth capturing so that others can learn from these lessons if they want to try similar collaborations elsewhere in the country or the world.

Departmental ‘silos’

One of the most difficult parts of getting an experiment like the behavioural nudges one up and running, is that it involves a small team of academics at a local university roping in the cooperation of a number of large municipal departments, where each one works in its own jurisdiction and is often hogtied by its own institutional rules and responsibilities.

‘We had to work with the City’s utility directorate, the legal department, the billing department, and the revenue department,’ says Visser, reflecting on the process. ‘We also needed to get permission to access confidential household data, and we had to guarantee the City that this information would be kept safe.’

It was difficult liaising between so many different municipal ‘silos’, and figuring out exactly who in each department should be the ‘go-to’ person. Since the start of the experiment, the City has now set up a designated research unit, which can help smooth over these sorts of practical rumble-strips. But during the crisis, many of the people who needed to be involved with the research were overwhelmed by their main responsibilities and had other priorities.

‘Coordination at this sort of scale is a huge job. Unless researchers are willing to do most of this work themselves, it can be impossible to pull off a project as big as this.’

Managers versus politicians

The City’s decision-making process is also shaped by specific legal parameters, which limit its agility in a time of crisis when decisions need to be made quickly. These constraints also made the researchers’ job difficult at times.

‘The City runs itself through two arms, essentially: there’s the managerial side of the city, which is bureaucratic, and there’s the council, which is the political side. A lot of design, thought and implementation happens on management side. This is often where you get a deadlock. So even if you’re sitting inside the City administration, you can’t necessarily implement something until it’s passed council approval, and that’s often highly politicised.’

Money doesn’t matter: lessons from an energy-saving experiment

Having a ‘champion’ for a cause within a community is more likely to get people to voluntarily change their behaviour relating to water or energy use, particularly if people aren’t paying for the resource and therefore wouldn’t have an economic incentive to cut their use.

This is the finding from another of the EPRU team’s behavioural nudging experiments¹⁴, and gives a further glimpse into how to encourage wiser resource use during times of electricity load shedding or a water crunch.

From June 2015 to October 2016, the EPRU economists ran a behaviour-change experiment in a 24-storey office block in the centre of town where people were employed by the provincial government.

From a researcher’s perspective, trying to tweak people’s behaviour in an office block context like this is a tough nut to crack: compared with experiments involving households, where you’re generally only dealing with four or so people, an office block context could have between 50 and 200 people taking part, per floor. And because office workers aren’t footing the bill for their water or electricity, researchers can’t use money as a motivator to drive behaviour change. In this experiment, researchers divided up the staff in the building by floor into three groups, and used smart meters to track energy use per floor.

Group 1: The ‘control’

The first group of seven randomly-selected floors was the ‘control’ group, and were left to carry on as normal, without any ‘intervention’ targeted at them.

Group 2: Tips and competitions

The staff from another seven floors were targeted with energy-saving tips, sent by email, with reminders to turn off the lights, or tips on how to cut energy use in the office kitchen. Researchers also stoked the staffs’ competitive flame, by setting up competitions between different floors.

Group 3: Tips, competitions, and a ‘champion’

The staff from another seven floors got the same treatment as above, but then also had someone on their floor appointed as the energy-saving ‘champion’ to help rally colleagues around this common good.

Championing a cause

The staff who only received emailed information and competed against other floors, reduced their energy use by 9 percent, says Visser. But the appointment of a ‘champion’ seemed to raise enthusiasm levels even more – these floors reduced their energy use by 14 percent.

‘This is a surprising find,’ say the EPRU researchers, ‘and it turns on its head the conventional wisdom that people won’t respond to energy or water-saving efforts if they’re not paying for the resources themselves.’

The EPRU team ran a similar kind of experiment targeting water use during the height of the drought. Working with researchers from Stellenbosch University, they ran another series of ‘green nudging’ interventions in a selection of schools in the city, and monitored water use with smart water meters that they’d installed in the schools’ water systems. Again they found that giving information to people, creating accountability in the community, and getting people to compete with each other in water-saving efforts was very effective in lowering water consumption.