I found myself slightly annoyed when I visited my local convenience store the other day. I purchased several items but wasn’t offered a bag at the checkout. I had to ask for one. "Does the storekeeper really expect me to carry all that shopping by hand?" I asked myself. The responsible consumer in me is of course aware that plastic bags produce unnecessary waste. The store’s default practice of not offering a bag is poor customer service at first glance, but makes sense from the perspective of cost reduction and pro-environmental business ethics.
My anecdote illustrates a simple but sometimes powerful mechanism at the core of ‘nudge’ theory in behavioral economics: defaults. These are pre-set options that take effect if we don’t make an active choice. Since humans’ default action is often inaction or plain old laziness (usually referred to as inertia), changing default options can make a difference to outcomes. Well-known applications of this principle include differences between opt-in choices (e.g. “Check this box if you’d like to receive special offers from us by email”) and opt-out choices (e.g. “Check this box if you would not like to receive offers from us”).
Defaults are ideally set to what’s good for the majority, or what helps the most vulnerable. The factory defaults on a smart phone, laptop or television, for instance, should cater to the way most people use the device, while information-disclosure settings on networks like Facebook may be more conservative in order to protect privacy.
Pension enrollment and organ donations are examples in favor of working with defaults to achieve desirable objectives. In both cases, a shift from opt-in to opt-out has the potential to dramatically increase retirement savings and organ-donation rates. These examples illustrate two conditions under which a change of default is effective:
Effort. Inertia is particularly strong if taking action requires effort, as often the case with investment paperwork. Automatic enrollment for retirement savings plans takes advantage of this problem. Unless people opt out, they will be enrolled.
Uncertainty. Individuals may be unsure about their own preferences and what’s expected of them, as when they consider organ donation. It’s usually perceived as a very personal choice. An opt-out approach signals that the social expectation or norm is (or should be) to donate your organs.
Since defaults signal norms, they have an implicit social character. Choices become even more strongly influenced by social forces if they are presented or mediated by other people rather than just impersonal application forms. Some decision-making contexts involve face-to-face communication, as in the shopping-bag example mentioned earlier. The choice is between a bag or no bag, but the storekeeper does not offer one by default. The customer has to make a basic but explicit request if s/he wants a bag. If the request is more unusual or even embarrassing (social norms will determine this), people are even more likely to go with the default option.
Consider service charges in restaurants: It’s the norm in many countries for customers to pay a tip to waiting staff. If you ask for the check in a UK restaurant, you’re increasingly likely to notice that a service charge of around 10-12.5% has already been added to your balance, sometimes with a remark like “Includes optional service charge." It’s a cheeky change of default that makes life easier for restaurants, and most customers. But the dissatisfied customer who does not wish to tip has to explicitly request that the service charge be removed.
Another example of socially-mediated choices and defaults are those encountered in British pubs and bars. If you ask for a draught beer, you will usually receive a full pint by default, unless you explicitly request a half. The requested size is tacit. It’s understood by pub-goers that they will be served a pint if they don’t request a smaller measure. Drinking pints, particularly among men, is a deeply-engrained social norm that has been around for ages. And of course it’s in the interest of businesses to sell more. It’s uncommon to hear bar staff ask customers whether they want a half or full pint. If bartender needs to clarify a drinker's preferred size, the question tends to focus on the most popular option—“Pint?” The response is usually yes, since it’s the implied choice. In certain masculine subcultures, it may even be perceived as embarrassing for a man to ask for a half pint of beer, a size often favored by women. Just like the beer that comes out of the tap, most people will go with the flow when others around them drink pints.
A few years ago, my colleagues and I reflected on ways behavioral economics could be used to reduce problem drinking in British pubs and bars. We suggested that reciprocal rounds-buying is an important normative issue that has to be tackled. Behavior change could also be achieved by helping consumers pre-commit to a certain level of drinking, by offering salient and easily accessible non-alcoholic alternatives, as well as by providing better feedback on the amount of alcohol consumed. In addition, the communication of drink sizes could be re-framed. For example, instead of calling drinks ‘half-pints’ or ‘small wines’ they could be referred to as regular-size, in contrast to the "large" pint option.
We also argued that “healthier” options in the form of smaller serving sizes can be made salient and that these choices could be presented as a default. The problem with using a default setting strategy to influence alcohol consumption relates to the two conditions outlined earlier. Making an active choice for consumer goods like alcohol does not involve much effort or uncertainty. With respect to the latter, changing defaults in the UK might work for better wine than beer, because wine serving sizes don’t appear to be associated with strong existing norms. Drinking establishments could communicate defaults through notices at the bar and verbally signal defaults when they take orders. If this practice were adopted as part of a mix of initiatives by a sufficient number of businesses, it may contribute to a decrease in binge drinking in the long run. The assumption, of course, is that smaller alcohol servings will sufficiently slow down the speed and amount of drinking in one sitting. There is some behavioral evidence suggesting that creating partitions slows the rate of consumption. Unwrapping another piece of chocolate, for example, breaks the mindless flow of consumption and incurs transgression costs, such as guilt. This may also apply when successive alcoholic drinks are ordered, though increasing intoxication is likely to neutralize some of the negative feelings.
Either way, nudges might have to occur alongside traditional economic interventions, such as pricing policies, in order to effect meaningful change. Behavior change through consumers’ wallets has already been successfully implemented for shopping bags by Marks & Spencers in Britain, for example, who introduced a 5-pence charge per bag a few years ago. In the case of alcohol, prices for a full (large) measures would have to be increased, creating a disincentive to buy large drinks in pubs and bars. The price would ideally be at least twice the price of a half (small) measure, which is not the usual business practice.
Alcohol is a domain of consumption that is highly hedonic, social and habitual. Behavior change is difficult to achieve. Along with the lack of incentives for businesses and the difficulty of implementing certain policies, I suspect that the powerful social norms that come with drinking in Britain may have been one of the reasons that motivated the UK government to propose a new size for draught beers. Instead of working with existing measures and defaults, the idea was to break out of existing norms altogether by introducing a size that has been absent in the UK, namely 2/3 pint ‘schooners.’ Judgments of size are relative and influenced by the range of measures that are available. In Scotland, a smaller measure of wine (125ml) is currently being promoted as an option that could be added to the widely available medium (175ml) and large (250ml) measures. In theory, this may change consumers' perception of drink sizes and norms, thereby potentially reducing large wine orders.
An important contribution of social psychological theory has been the finding that behavior can change attitudes (e.g., when we rationalize past behavior), just as attitudes affect behavior. According to behavioral scientists, traditional approaches to behavior change focus too much on attempting to change behavior by first influencing awareness and attitudes. Nudging people at the moment of decision making is often more effective and may even help new social norms to emerge. But sometimes existing social norms and associated preferences are bound to resist change. In those cases, attitudes may have to be influenced as well if nudging is supposed to lead to behavior change, especially in the long run. The relationship between defaults and social norms present us with an interesting chicken-and-egg problem in fostering responsible consumer behavior.
Available from July 2014: The Behavioral Economics Guide 2014 on BehavioralEconomics.com (free download)