As designers, we influence important decisions in people’s lives. Let’s do it more ethically.
Research has demonstrated that shaping the environment where decisions are made can greatly influence people’s behaviours. As designers, we are in the business of shaping those environments, giving us great power over what people decide to buy, how much they save, or how they interact with each other online. It’s high time we treat this power with more careful consideration of ethics and systemic effects. This is how.
By Sofie Jensen
The use of behavioural design, a practice in which design is enriched by insights from behavioural science, is on the rise. It can be used to inform how people make decisions, and what drives their actions. Often this is purely profit-driven, but behavioural design is also increasingly being used as a method to design for sustainability. This is a welcome development, as many of the SDGs (Sustainable Development Goals) are linked to human behaviour and norms, which require behavioural changes from individuals, organisations and societies.
Even when working for sustainability, the old superhero adage applies — with great power comes great responsibility. Behavioural designers can affect important life choices through choice architecture, which refers to conscious adjustments of the context in which decisions are made, for instance by making some alternatives easier or harder to choose. If decisions concerning such behavioural nudges are taken without the user, choice architecture can quickly become paternalistic (reducing people’s freedom and responsibilities) and even unethical. This isn’t hypothetical; stores already use nudges to make us spend more money, e.g. by labelling objectively large products as the “small” option. Even good intentions can end badly. An example is including calorie-labels on menus to nudge healthier choices. For people with low eating-control it just makes them feel bad, instead of resulting in healthier eating. So how do we use choice architecture in an ethical way? The answer could be co-design, candour and consent.
Context is there already, so we can’t ignore it
The issue of ethics in behavioural design is difficult to approach in a constructive manner. Asking designers to be mindful of the unintended effects our work may have can easily translate to “it’s better to do nothing, then you’re safe”. It seems better to just not do it. But doing nothing is also a choice. If we do not design the decision-making context consciously, some alternatives will still be easier to choose. By making the decision-making context concerning organ donorship arbitrary, many countries today have less organ donors than they could have. Is that fairer? No. Doing nothing, allowing the effects of the context you are designing to be ruled by random effects, does not absolve you of responsibility for those effects.
At the same time, rushing in and thinking “it’s better to do something!” goes too far in the other direction. Take the case of fighting bullying in schools. Who can’t get behind anti-bullying programs, that encourage kids to speak up against bullying? Unfortunately, many of the existing anti-bullying programs increase bullying, because the mechanisms behind bullying behaviours are not fully understood. So, by moving fast and breaking things, we can, in fact, break potentially important things. Therefore, the mantra “first, do no harm” is key. Properly understanding the system, as well as how your actions might affect it, can help you to do good without having to break things first.
Designers, co-design!
So how to manoeuvre the ethical concerns of choice architecture in design processes? Luckily, designers need not learn new tricks. User-centricity is still king.
People often have a clear — if somewhat general — idea of what they want for their lives. They might want to eat healthier, spend less money on things they don’t really care about, or help save the rainforest. Of course, preferences will vary, yet in many instances people will be relatively uniform. Involving users when deciding which choices should be the easiest to make can thus empower them to achieve their personal goals. For example, if they want to exercise more, then including a nudge to share exercise goals with friends increases the likelihood that they manage to achieve their goal of a healthier lifestyle. Designing the service in a way that makes it easier to recycle than not helps people take care of the world. In this way the solutions can make our brains work with us, rather than against us.
But this only works if user involvement goes all the way. Some design processes involve users in a manner that resembles milking cows. Their lived experience is collected, but they aren’t involved in the important decisions. As long as the designer is “the expert” through which all insight and solutions must flow, the design process becomes a way of maintaining the status quo. Luckily, there are many useful tools available to help you empower users, for example through participatory design.
Be candid about where you’re coming from
In instances where the outcome is not as universal as good health, preferring one alternative could result in imposing our view of what is “right” on others. Examples could be choosing to save for buying a house and settling down or working a 9–5 job. Although the majority of the population in many countries consider this desirable, a sizeable minority don’t want this from life and their desires should be respected. An analysis of your own biases and assumptions — reflexivity — is thus a crucial step to achieve ethical behavioural design. To uncover your blind spots, is important to involve those that might illuminate the perspectives that are inaccessible to you. This has implications for recruitment. To reach representative conclusions, it is necessary to map out possible aspects of diversity and ensure that this is represented in your sample.
Be clear about what you know — and what you don’t
Uncovering your own — and your client’s — biases is one important step. The next is to be candid about the limitations of your data. When presenting findings, it can be tempting to focus exclusively on what is known. Unfortunately, such a focus can obscure the limits of your findings. For example, if the sample only included white, middle-class women, you do not know if your insights apply to working-class men from ethnic minorities. If 30 % of the employees responded to your survey, the majority might have a very different perspective than your results indicate. Consequently, when the client relationship allows, it is important to be honest about what isn’t known and for whom your findings might generalise to. By making such limitations visible and a part of the conversation, you can hopefully prevent one “truth” from becoming the only truth.
Nudging still works if people consent to it
Sometimes it seems that people think we have to keep nudging a secret in order for it to work. Yet most of the time these nudges and adjustments will work just as well if people are aware of them. Therefore, transparency concerning the implemented choice architecture, can and should be the norm. Also, telling users that you have designed the solution to make sustainable or healthy living easier is rarely badly received. The people who are most sceptical towards such transparency might be those that use choice architecture at the expense of the user or the planet. A good rule can therefore be: if you can’t tell people that you are doing it, you probably shouldn’t be doing it.
You’re not done at delivery
Even well-researched solutions will most likely have some unintended effects. What’s more, they might not always turn out to be effective in alleviating the targeted problem. It is therefore paramount to measure the effects of the change. Did it result in the desired effects? Were there any negative side-effects? Did resolving this problem create another problem? This resembles the work you do before acting when applying “do no harm”-principles: mapping out the wider system and exploring whether your hypotheses are confirmed or not.
Behavioural design — the synergies of two strong disciplines
As with all perspectives and methods, behavioural design is of course not a purely good, fix-all approach. Yet it is latent in all decision-making contexts, so it is imperative to at least be conscious of the effects. Through the synergies between behavioural sciences and design, we can also increase the likelihood that the decision-making environment is working for people and planet, rather than against them. If we are candid about our personal biases and the limits of our data, and involve diverse users in a meaningful way, we can reduce paternalistic tendencies. Mapping the potential effects beforehand, and measuring outcomes in retrospect, can help us avoid unintended harm. It does require conscious effort, but it can be done.
Let’s use our skills to ensure that we build a society working for the people living in it, and for the planet we depend upon, in a way that empowers, rather than infantilises. Let’s design ethically.
