About the author: Stephanie Tam specializes in behavioral approaches to sanitation and hygiene in international development, with a focus on how cultural practices mediate the impact of technology in operations and maintenance. A graduate of the Harvard Graduate School of Design, she has been working on human behavior through the lenses of performance studies and behavioral economics, and thanks Dilip Soman at the Rotman School of Management at the University of Toronto for support on this current project. She is deeply invested in monitoring the socioeconomic impact of WASH projects, and has worked with various NGOs, human rights advocates, anthropologists, environmental engineers, and urban planners to develop technologies, infrastructure and behavioral practices that empower the under served in a measurable and accountable way. Stephanie can be contacted via her website, Linkedin, or Academia.
* * *
Behavioural change is a crucial site for WASH interventions, and in Part I I’ve argued that practitioners should focus upon reflexive behavioural influences instead of reflective ones. While existing WASH behavioural change strategies are based upon public health models that target reflective behaviour, I’ve suggested that we turn towards behavioural economics for reflexive behavioural insight. That said, not all behavioural economics studies are strictly about reflexive behaviour. The field itself is shaped in refutation of rational choice theory, and consists of diverse studies unified by their interest in deviations from traditional economics. Behavioural economics cannot be directly mapped onto WASH behavioural change, and what I seek to do here is to elucidate paths for action research rather than prescribe ways of applying behavioural economics to WASH. The fundamental tenet of behavioural economics is that context matters. WASH interventions must therefore develop from the site-specific findings of action research, and eschew a copy-and-paste approach to behavioural economics studies.
The “principles” of behavioural economics are less universal rules than particular kinds of contextual information that we should pay attention to when designing behavioural change interventions. It is tempting to extrapolate localized findings into universal principles, and even though some behavioural economics studies have successfully replicated their findings in various settings, I strongly contend that WASH projects need to invest the time and money into developing behavioural influences specific to site conditions. The consequences of shortcuts are too costly for those whom WASH serves for us to not take extreme care in crafting interventions.
Furthermore, reflexive behavioural influences need to be located within a participatory planning process where target audiences can choose what kinds of behaviours they want to adopt. Given the power of subconscious behavioural influences, it is ethically imperative for target audiences to consciously decide what nudges they desire. This is where extant behavioural change approaches based on public health education are helpful in offering people knowledge about the repercussions of WASH (or lack thereof), so that audiences can make informed decisions. Behavioural economics is also important here in ensuring that decision-making biases are neutralized so that audiences are not predisposed towards certain choices that they would not otherwise prefer. That, however, will entail a whole other article on decision-making, so I will leave it at that and move on to the kinds of contextual cues that action researchers should document and test when an audience has identified WASH behaviours they want to take up.
WASH behaviours are ingrained through implicit or procedural learning, such as classical conditioning or priming, where a person is predisposed towards certain behaviours through strategic exposure to an earlier stimulus. For example, in an oft cited social psychology experiment, certain participants were exposed to words associated with a stereotype of the elderly and then shown to walk more slowly down the hall in comparison to other participants who were primed with non-age-specific words (Bargh, Chen, Burrows 1996). Of course, there are many variables that have to be controlled and many limitations to experimental results (see the recent scholarly scuffle about the validity of priming experiments), but the gist is that exposure to particular images and words activate associations that manifest in our subsequent behaviours. Behaviours can be encouraged or discouraged if project planners seek out an audience’s already learned behaviours and transfer them to a WASH context through planned exposure to stimuli. Action research needs to pay attention to local stereotypes, association clusters, and already conditioned reactions, no matter how seemingly unrelated they may be to WASH. Memories and emotions triggered previous to an encounter with soap or a toilet can predispose their uptake. Deploying images, sounds, or smells that an audience already associates with a certain kind of behaviour, or having an audience perform a gesture that is laden with emotions and memories can suggest behavioural reactions that we can leverage.
In another priming experiment, Nobel laureate Daniel Kahneman instructed certain participants to hold a pencil in their mouth so that their facial muscles were unconsciously forced into expressions of happiness (a smile) or dismay (a frown), and found that their emotional reactions to respective humorous or upsetting images were more elevated compared to other participants (Kahneman 2011). Unconscious acts and gestures induce emotions in us that transfer to our subsequent experiences. Affect studies scholar Sara Ahmed describes emotions as “sticky”: they maintain connections between ideas, values and objects, orienting us towards certain behavioural trajectories over others (Ahmed 2010). WASH hardware and behaviour introductions need to ensure that an audience is emotionally biased towards WASH so that they can take advantage of what Kahneman calls the “halo effect” – our desire for emotional coherence that results in us liking or disliking everything about something once we’ve jumped to the conclusion that it’s “good” or “bad” (2011). Our first impressions of new things come to dominate how we react to them, overshadowing our subsequent experiences even if these counter our initial judgment. Action research can help to identify what primes will resonate most with a target audience, and thereby increase the likelihood of them subconsciously preferring WASH interventions. Even though the audience may have consciously stated that they desire a WASH intervention, subconscious judgment is key to influencing reflexive reactions to sanitation and hygiene practices.
Not only does the audience need to be emotionally primed prior to the introduction of a WASH intervention, the emotional arc of any WASH session and how it draws to a close need to also be carefully designed. Kahneman distinguishes between our “experiencing self” and our “remembering self”, with the latter steering our future reactions based upon our selected memory of a past experience. In an experiment where participants had their hand painfully submerged in cold water with some moments of relief at different points of the trial, Kahneman found that memories of an experience are “strongly influenced by the peak and the end” (2011). To create a positive memory of WASH that will dispose audiences towards engaging with WASH in the future, practitioners should aim to program sessions with a positive emotional peak and ending, regardless of whether or not the positive emotions are rationally connected to WASH.
The most difficult part of WASH behavioural change is daily maintenance. New WASH practices need to be perceived as the norm in order to be practiced as the norm: representation of what the status quo consists of strongly influences what actually becomes the status quo. As Richard Thaler and Cass Sunstein suggest in Nudge (2008), “[i]f choice architects want to shift behavior and to do so with a nudge, they might simply inform people about what other people are doing”. Rather than using Behaviour Change Communication to elaborate upon the benefits of WASH, messages should seek to communicate how most of the target audience already practices WASH behaviours. Note that it is perception of peer pressure that is at work here, not actual enforcement through peer policing. We are leveraging existing desires for social conformity (“social contagion” in nudge-speak), and the existing belief that other people are paying close attention to what we are doing (the “spotlight effect”) to establish new default behaviours. We are also leveraging representation in a situation where people are not certain as to what the status quo is. Nudges are most effective in situations where people are uncertain as to what to do or believe. Those who have staunch counter-beliefs will not be as susceptible to nudging, which is why it is important to ensure that the audience consciously desires to adopt new behaviours in the first place.
When new WASH behaviours are perceived as the norm, alternative behaviours come less to mind when encountering a sanitation or hygiene situation, thereby narrowing the margin for alternative practices. Samuelson and Zeckhauser argue that “[i]n day-to-day decision making…a decision maker may not even recognize the potential for a choice” so that “the status quo is then even more likely to prevail” (1988). While new WASH behaviours will not become the status quo overnight, we can leverage information distribution to accelerate changes in perception. Behaviour Change Communication can be a powerful determinant of information availability. Readily available information about a particular behaviour creates a perception that the behaviour is common. Tversky and Kahneman call this the “availability heuristic” (1974). The easier it is for us to retrieve an example of something in our memory, the more we are prone to believe that it happens a lot. Retrievability is affected not only by repeated exposure to a particular piece of information, but also the vividness or personal impact of the information. We can recall a burning house we’ve seen in person better than a report about a burning house we heard on the radio, and will therefore be more likely to believe that our own house is at risk of fire if we have seen a burning house in person. Behaviour Change Communication can take advantage of the availability heuristic by developing content and modes of delivery that speak to the audience personally. Information about prior or alternative behaviours should be suppressed, even when such behaviours are portrayed negatively.
Unlike public health models and behavioural change theories that offer a general action plan which can be adapted for specific cases, individual nudges have to be developed for each scenario, and what works in one context will not necessarily work for another. While behavioural economists have the luxury of spending time and resources devising and evaluating controlled trials, WASH behaviouralists will have to limit their design process and work with small sample sizes in an uncontrolled test environment that will likely generate a lot of data contamination. The challenge for WASH behavioural change is to stay context sensitive and to develop criteria for adequate testing without sacrificing validity.
Ahmed, Sara. “Happy Objects” in The Affect Theory Reader. Ed. Melissa Gregg and Gregory J. Seigworth. Durham: Duke University Press, 2010.
Bargh, John, Mark Chen and Lara Burrows. “Automaticity of Social Behavior: Direct Effects of Trait Construct and Stereotype Activation on Action.” Journal of Personality and Social Psychology 71.2 (1996): 230-244.
Kahneman, Daniel. Thinking, Fast and Slow. Canada: Doubleday, 2011.
Samuelson, William and Richard Zeckhauser. “Status Quo Bias in Decision Making.” Journal of Risk and Uncertainty 1 (1988): 7-59.
Thaler, Richard H. and Cass R. Sunstein. Nudge: Improving Decisions about Health, Wealth, and Happiness. New Haven: Yale University Press, 2008.
Tversky, Amos and Daniel Kahneman. “Judgment under Uncertainty: Heuristics and Biases.” Science 185.4157 (1974): 1124-1131.