By Laura Mneimneh
Behavioral scientists design and implement nudges to steer our decisions away from what is ‘bad’ for us towards what is ‘good’ for us. At least they try to do so. Sometimes, they fail. They do not fail to steer us – no. Steering is often the easy part. Steering for ‘good’ is the challenging part, because ‘good’ is an abstract concept. We know steering is measurable with statistics. Steering for ‘good’ is also measurable with statistics. But statistics are not enough. Discussing nudges and their ethical implications is an essential exercise to ensure that we are indeed nudging for ‘good’.
Behavioral scientists are thinking of a nudge to make us stick to a ‘good’ diet. Many of them are thinking: ‘Let’s put a label on unhealthy items in the vending machine. A candy bar could bear a label stating that one tiny candy bar is equivalent to a kilogram of carrots in terms of calories. Consumers will steer away from the candy bar and switch to a healthier item! Bingo!’
This nudge seems like a nice idea, but it does not nudge for ‘good’. It sheds light on quantity (the number of calories), rather than quality (whether the ingredients are natural or processed), which reinforces many misconceptions we might hold.
Debunking common misconceptions
A frequent misconception we hold is believing that what makes food healthy or unhealthy is the number of calories it contains. We know that soda is unhealthy and has a lot of calories and that chamomile infusion is healthy and does not have a lot of calories. What we observe is a negative correlation. The negative correlation between healthiness and the number of calories of food might be strong, but it is not a causation. Instead, quality is the cause for the healthiness of a product. A product is unhealthy because it is processed, and it happens to contain a lot of calories. What a quantity nudge does is to tell us to switch to another item because the one we initially wanted is high in calories, thus reinforcing the false causation.
Shedding light on quantity and omitting quality implies that it does not really matter whether the item we pick is natural or processed. In terms of calories, an avocado can be substituted by a candy bar, and vice versa. This equation can be misleading because it only holds if the calorie is the unit of measurement. However, for the same number of calories, an avocado is preferred to its equivalent in calories – a candy bar. An avocado is still preferred to half the candy bar, even if the avocado has the double of the number of calories included in half a candy bar. Of course, this only holds up to a certain extent (Will a whole avocado be preferred to one crumble of the candy bar?) To a large extent, quantity can be compromised for the sake of quality. When the nudge implies that natural and processed products are interchangeable options based on calorie-content, it is reinforcing the misconception that quantity matters over quality.
Going back to the vending machine, the calorie label on the item does not say anything about correlation or causation, it appears to be simply descriptive. So why delve into all of these details? Are we not putting too much weight on behavioral scientists’ shoulders?
Behavioral scientists are not merely giving us a small push
Behavioral scientists are shaping behaviors and their role should not be underestimated when it is convenient to do so. The choice of the label in the vending machine is not simply descriptive. It gives us a sense of what a ‘good’ diet is, by shedding light on some aspects of the item and leaving out other important aspects.
We tend to underestimate the extent to which nudges affect our thinking. We falsely believe that the automatic and reflective systems are separate. We think of nudging in contrast to education because of the different systems they target. Nudging targets the automatic system whereas education targets the reflective system. Because we distinguish the two systems when we speak and write, we tend to think of the distinction as material. We should not forget that the two systems are ‘interrelated’ to a large extent. A nudge targeting the automatic system can and will involve the reflective system. In our case, the consumer who reads the number of calories on the candy bar might still purchase it (possibly exhibiting the automatic system), only to add another half an hour on the treadmill in compensation (exhibiting the reflective system at a later stage).
You may argue that the consumer who purchased the chocolate bar despite the nudge will statistically decrease the impact of the nudge. Still, the nudge would not be for ‘good’. People carrying misconceptions regarding nutrition will have their misconceptions reinforced and will have a harder time sticking to a ‘good’ diet. For example, individuals who eat a chocolate bar despite being nudged might end up skipping dinner out of fear of exceeding their daily calorie intake.
The impact of nudges is not frozen in time
Whether the people nudged actually end up buying the chocolate bar or not, we have little knowledge of what happens past the vending machine. This is something that we have no control of and that applies to all nudges, the ‘good’ and ‘bad’ ones. Luckily, our lack of information does not prevent us from implementing nudges in general. It rather highlights the importance of the ‘sustainability’ of the message conveyed. In our example, it highlights the importance of quality over quantity.
The ideal nudge would highlight both quality and quantity. One example could be a ‘high in sugar’ label. ‘Sugar’ can refer to both the quality (processed) and quantity (high in calories) of the food product in question. However, where it becomes tricky to combine both aspects of a healthy diet in a nudge, quality should be prioritized over quantity, even if it yields weaker statistical results.
