A peek under the bonnet — when and where are nudges effective?

Even though nudges have been around for a decade, there is still debate around how they work and where they might be used most effectively. In this post, I review a proposed matrix classifying nudges based on their proposed mechanisms and in which circumstances they are likely to work best.

Earlier this year I came across a paper (Löfgren & Nordblom, 2020) proposing a theoretical framework of decision making explaining the mechanisms of nudging by mapping out decisions on two axes: importance of the decision to the person and their confidence in making the right choice. The model is meant to be used to predict under which circumstances, and in which choice situations, a nudge is likely to be effective.

As a quick recap, this is the original definition of a nudge by Thaler and Sunstein (2008):

A nudge… is any aspect of the choice architecture that alters people’s behavior in a predictable way without… significantly changing their economic incentives.

On the surface it seems straightforward and commonly the latter part of the definition (without economic incentives) has been considered crucial for defining what constitutes a nudge. In contrast, this paper classifies behavioral interventions as nudges or not based on the decision-making process, rather than on the motive behind the intervention. The definition for nudge proposed by the authors of this paper is therefore:

“a nudge is an alteration of the inattentive choice situation, which would not affect the attentive choice.”

The balancing act in every decision we make

Every decision we make includes a trade-off between speed, effort and accuracy: a faster, more intuitive, less attentive decision that is less accurate, or a slower, more deliberate and attentive decision that is (hopefully) more accurate. When making a decision, we — or our brains — compare the benefits of making an effort or not, and the risks of getting it wrong. We might decide it’s not worth the effort because we already (feel like) we have enough knowledge in this situation, or because it doesn’t really matter that much even if we got it wrong.

Although our goal is to maximise the benefits, it often takes a lot of cognitive effort to work out what the right choice is so our brains ‘cheat’ by making decisions on gut feel, our previous choices (habit) or by using heuristics. It is these situations where we pay less attention to our choices that we are most susceptible to behavioural interventions such as nudging.

However, what makes a choice effortful varies — it could be about how much information someone needs to find out before making a rationally optimised choice (friction), or their individual characteristics such as cognitive ability, personal traits or the choice situation itself.

In general, a decision is more likely to be made inattentively if:

• A lot of effort is required to make the best decision
• The behaviour is habitual and the person has a lot of experience of the consequences
• The choice is not very important and the risk of making a mistake is considered small

Their definition of a nudge is some change in the situation where a choice is made inattentively — but not if the choice was made attentively. Therefore, for a behaviour to be susceptible to nudging, the choice has to be made inattentively.

A nudge is more likely to be effective if the choice is not very important to the person — in other words, the consequences of getting the choice wrong are not meaningful to them in some way. The authors suggest that for choices that someone considers to be important, nudges (of any type) are less likely to have any effect at all and other kinds of intervention should be considered (e.g. information, regulations or price).

So, what are the four types of decision making situations?

The two axes of the matrix are confidence of decision maker and importance of the decision:

• on the right are the situations where the decision maker feels confident that they know how to make the right choice (B & D)
• on the top (A & B) are situations where someone cares a lot about making the right decision for themselves — where stakes are high for some reason, for example

The authors also did a survey* where they asked people about how people in general would perceive the different choice situations in the above importance-confidence space. The results and examples are discussed below.

At the top (quadrants A & B) are important choices made by low-confidence decision makers — these can be made inattentively (and therefore nudged) if the required effort to make an attentive choice is very high. For example, buying a house is a high stakes decision but if you are a first time buyer you might be less confident than someone buying their third house.

Quadrant B is decisions that you care a lot about getting right and that you also feel pretty confident you’ll be able to make the best choice for yourself — in principle. In real life, there could be some individual variation in the degree of attentiveness for something like taxes where you might want to change the behaviour of a small minority who don’t pay on time. Habitual choices or routine behaviours are also in this quadrant — for example, if you usually drive to work it will be difficult to nudge you to change your commute.

According to the authors, decisions in the bottom quadrants C and D are most likely to be influenced by nudges of two kinds**:

• Pure nudges that change something in a choice situation that is seemingly irrelevant to the choice itself (most effective in quadrant C)
• Preference nudges that change the perceived value in a choice situation without actually changing the attentive choice. (most effective in quadrant D)

Commonly used pure nudges are default options and framing (e.g., placing apples at eye level) while preference nudges commonly include social-norm messages of others’ behaviour (e.g., a note in the shower indicating the share of hotel guests who reuse their towels).

Some examples from the paper:

• What to have for lunch was an example that most people saw as being neither very important nor unimportant — here, a pure nudge could be changing the order of dishes on the menu and a preference nudge a sign saying, “Chef’s recommendation” or “Vegetarian food is good for the climate”
• In electricity use, a preference nudge could be to provide information about the mean energy consumed by a household comparison group
• Although pension saving is an attentively made decision, low-confidence decision makers can make inattentive decisions — for a target group like this, nudges can be effective

There are more examples in the paper itself as well as supporting evidence from other research.

My thoughts on the proposed ideas

To some extent, this makes sense — there are lots of things I have strong preferences for and very little can sway my choice for those even if they are categories someone else might describe as low stakes or trivial like shampoo, oat milk or dog food — to me all of these are important because poor quality shampoo makes me look like a sheep in rainy season, all other oat milk except Oatly tastes terrible and dogs… well, it’s no surprise the wellbeing of my dogs is hugely important to me and I know enough about the topic to have a lot of confidence in my choices.

For a practitioner the question still remains how we’ll know the degree of confidence and importance — especially as they might be influenced by contextual factors as well as evolve with over time. For example, hand soap was probably a relatively inattentive choice before March 2020, but much less so a year later.

You can find more details and references in:

Löfgren, Å., & Nordblom, K. (2020). A theoretical framework of decision making explaining the mechanisms of nudging. Journal of Economic Behavior & Organization, 174, 1–12.

* “Survey conducted among a representative sample of Swedish citizens during fall 2019. In the survey, 1,533 respondents answered questions about different choice situations in terms of perceived importance and confidence about the outcome if making the choice inattentively. Since answering a survey is by definition made in an attentive state, we did not ask how the respondents themselves would react, but rather how they thought people in general would perceive different choice situations.”

** “Note that the same intervention could be regarded as a pure nudge by some and a preference nudge (or, even information) by others”.