Dueling with Desire
The better we understand want/should conflict, the more successful we will be at designing effective interventions that shift behaviour.
IN OUR DAILY LIVES we frequently face a tension between what we want to do and what we believe we should do. After a long week at work, we may want to share an expensive dinner and a few drinks with friends — when we know we should go home early and get a good night’s sleep; similarly, we might be tempted to get caught up on the current season of Homeland when we know we should focus on drafting that report we promised our colleague. For decades, researchers have examined the battle between highly desirable options that provide immediate gratification (e.g., eating junk food, procrastinating, overspending) and options that provide more long-term benefits (e.g., eating healthy food, meeting deadlines and saving for retirement).
Harvard Professor Max Bazerman calls this common struggle ‘want/should conflict’. In this article, we will summarize the key findings on this pervasive conflict and discuss a series of interventions that policymakers, organizations, and individuals can use to promote more future-oriented choices.
A Struggle for the Ages
Individuals often evaluate decisions through two different lenses: The want self focuses myopically on the here and now, and thus strongly desires instant gratification; while the should self is more far-sighted, guided primarily by long-term interests.
Indulging in the wants that we most desire can cause us to feel wasteful, irresponsible and immoral, and as a result of a distaste for such feelings, some individuals under-consume want options. However, it is far more typical for people to feel that they have made the opposite mistake (i.e., over-indulging in wants at the expense of shoulds) and to regret this irrational behaviour later on.
Over-indulging in want options typically has a much greater cost than overindulging in should options. For example, failures to control one’s desires (e.g., choosing pizza over vegetables, watching TV instead of exercising, smoking rather than quitting, buying an unnecessary designer handbag rather than depositing the money in a savings account) can contribute over time to
The want self focuses myopically on the here and now, and thus strongly desires instant gratification.
serious individual and societal problems, such as obesity, higher cancer rates, and under-saving for retirement.
In many such cases, then, we can say that observed levels of indulgence in want options are sub-optimal (and thus not rational), as higher net utility would be obtained by selecting shoulds. Because the mistake of over-indulging in wants is generally more common and costly than the opposite error, when we discuss want/should conflict throughout this article, we will focus our discussion on how to increase the rate at which should options are selected.
Following are eight factors that have been shown to affect our choice of wants vs. shoulds.
1. CHOOSING FOR NOW OR LATER. A key finding is that people prefer should options at a higher rate when making decisions for the more distant future, but prefer want options more often the sooner a choice will take effect. For example, deciding to go to the gym tomorrow is easier than deciding to go this very minute, and committing to save more for retirement next year is easier than committing to forego a portion of today’s paycheque. This pattern has been demonstrated in decision domains ranging from those involving money to food and movie rentals.
Multiple studies investigating impulsiveness have confirmed that people show extremely high ‘discount rates’ for delayed rewards. In one study, the average participant opted to receive $50 immediately (a want option) rather than $100 in six months (a should option) but preferred to receive $100 in 18 months rather than $50 in 12 months. These results contradict the predictions of standard economic theory, which suggests that an individual’s preferences between two sure sums of money should depend only on the time delay that separates their receipt (six months in both cases).
These results are consistent with hyperbolic time discounting. In one experiment, subjects randomly assigned to select a film to watch that same day were more likely to select low-brow films ( want choices) than subjects randomly assigned to select a film they would watch several days in the future. And in the domain of online grocery shopping, one of the authors [ Katherine Milkman] found that the percentage of extreme should groceries (e.g., fruits and vegetables) in a customer’s basket tends to increase and the percentage of extreme want groceries (e.g., ice cream and cookies) tends to decrease, the further in advance of delivery a customer places her order.
2. COGNITIVE LOAD. Daniel Kahneman’s two-system model of thinking suggests that the relative strengths of System 1 reactions (which are automatic, emotional and instinctual) and System 2 reactions (slower, more deliberative and logical thinking) influence the outcomes of want/should conflicts. Building on this notion, want options are expected to be more likely to win out when the cognitive resources available to make a decision are limited (i.e. when System 2 is over-burdened), which allows System 1 to dominate.
In one study designed to test this prediction, researchers presented participants with two snack options: A piece of chocolate cake or a cup of fruit salad. They found that individuals who were randomly assigned to memorize a seven digit number (and who thus had reduced cognitive resources) were more likely to choose cake over the fruit than those who were assigned to memorize a two digit number. This finding highlights that the availability of cognitive resources is critical to making far-sighted and deliberative should decisions.
. CONSTRUAL LEVEL. Construal Level Theory (CLT) suggests that we prefer shoulds over wants more often when we are thinking more abstractly and thus focusing on the global features of options, rather than when we are thinking concretely and focusing on the contextualized, surface-level and goal-irrelevant features of options. For instance, abstract representations of ‘exercising’ bring to mind its long-term benefits, while concrete representations remind us of its in-the-moment pains and required planning.
Research has shown that the tendency to make should choices can be enhanced by inducing abstract, high-level representations of events — for example, by focusing people on more distal events in time and space, rather than more proximal events. In one study, researchers primed participants to think abstractly by asking them to describe why they maintain good physical health, and primed others to think concretely by asking them to describe how they maintain good physical health. They found that when
a more abstract, high-level mindset was activated, people exhibited stronger preferences for immediate want outcomes over delayed should rewards. Inducing people to adopt a higher-level ‘construal mindset’ is therefore one way to increase future oriented decision making.
4. DEPLETION. As indicated, a growing body of research suggests that exerting willpower comes at a cost, and that cost is a reduction in available self-control resources for use in future choices. In other words, individuals have limited self-regulatory resources, and exerting self-control to avoid wants in one situation can decrease one’s subsequent ability to exert self-control.
In one study designed to test this theory, researchers showed that participants who resisted eating chocolate chip cookies (an obvious want for most people) quit working on unsolvable puzzles earlier (where persistence is a should behaviour) than did individuals who resisted eating radishes — an activity that for most people, requires little self-control. This and other studies illustrate the paradox that by exercising self-control now, we increase the likelihood that we will give in to our desires to indulge later. Fortunately, although the self-control muscle can be weakened through repeated use, it can also be strengthened through proper exercise: One study found evidence that practicing small acts of self-control greatly increased smokers’ chances of successfully quitting.
5. JOINT VS. SEPARATE EVALUATIONS. The outcomes of want vs. should conflicts are also influenced by whether we evaluate options one-at-a-time or simultaneously. Although want options tend to be preferred at a higher rate than should options in isolation, we are more likely to think about the costs and benefits of each option and make far-sighted choices when multiple options are evaluated at the same time. For example, when viewed in isolation, a charity that saves baby polar bears may seem more alluring and receive more donations than a charity that funds skin cancer research. However, when these choices are compared side-by-side, people tend to donate to the charity that helps people, viewing its mission as more important, albeit less emotionally resonant. This research highlights that presenting want and should options simultaneously rather than sequentially is one way to promote more should choices.
6. MOOD EFFECTS. Positive mood has been shown to facilitate future-oriented decision making, and a number of explanations have been proposed for this. One account is that experiencing positive emotions signals to decision makers that their current situation is non-threatening, which reduces discounting of the future and thus makes shoulds relatively more attractive.
Another reason is that positive emotions can counteract ‘ego depletion’, restoring the depleted willpower resources necessary for selecting should options. In one study, researchers demonstrated that after randomly assigning participants to watch clips of videos that induced either a happy or neutral mood, those placed in a positive mood were more likely to select a should option (grapes) over a want option (M&MS). However, they found that the tendency for positive moods to increase should choices is attenuated by elevated arousal because arousal is depleting.
Research has also explored the impact of negative emotions on want/should conflict. Recent studies have demonstrated that these emotions can lead to self-control breakdowns. For instance, sadness increases decision makers’ tendency to focus on immediate gratification and to dramatically discount future outcomes. Together, this research shows that people who are relaxed and happy are more likely to make should choices, whereas individuals who are emotionally aroused or in a negative mood are more likely to reach for instant gratification.
7. LICENSING EFFECTS. Interestingly, our choice between a want and a should can also be affected by decisions we have made in the past, as well as those we anticipate making in the future. Research shows that people feel ‘licensed’ to make (or justified in making) want choices if they believe they have previously engaged in should behaviours or if they anticipate having opportunities to engage in should behaviours in the future.
For example, people who were asked to imagine they would partake in a should behaviour (e.g., donating part of their tax rebate to charity, volunteering for community service), relative to a control group who did not imagine any such future good behaviour, were more likely to select a desirable want product (e.g., a pair of designer glasses) over a cognitively favourable should product (e.g., a less expensive but more utilitarian pair of glasses).
Furthermore, want products are more likely to be selected when individuals make what they believe is the first of a series of similar decisions rather than a single, isolated choice, presumably because individuals making repeated decisions believe that they will have the opportunity to choose shoulds in the future to compensate for current indulgences. These findings highlight that choosing between wants and shoulds is often not done in isolation, but instead, hinges on an individual’s past choices and anticipated future decisions.
8. FRESH STARTS. Recent research suggests that there are naturally-arising points in time when people are particularly motivated to pursue their long-term interests and prefer shoulds. Temporal landmarks, which include personally-relevant life events (e.g., anniversaries, birthdays) and reference points on shared calendars (e.g., holidays, the start of a new week, month, year or semester), demarcate the passage of time and help us organize our activities and experiences. Field research has shown that temporal landmarks magnify people’s virtuous intentions and increase their engagement in should behaviours. Researchers analyzed (a) daily Google search volume for the term ‘diet’, (b) undergraduate students’ gym attendance records, and (c) a wide range of goals (e.g., pertaining to education, health, finance, etc.) that Internet users committed to pursuing on a goal-setting website ( stickk.com). These studies revealed that people engage in should behaviours (i.e., dieting, exercising and goal pursuit) more frequently following temporal landmarks, including the start of the week, month, year and academic semester, as well as immediately following a birthday, a national holiday, or a school break. The authors refer to this phenomenon as ‘the fresh start effect’.
Should Prescriptions
Faced with the above challenges to our decision making, what can be done? Following are three ‘nudges’ that have been shown to successfully increase the rate at which people choose shoulds over wants.
PRESCRIPTION 1: PROMPTS Prompting people to stipulate when, where and how they will enact their goals is one of the oldest prescriptions for increasing engagement in should behaviours — dating back to research conducted in the 1960s. Plan-making has been shown to improve our likelihood of achieving goals in a diverse array of domains, including exercise, dieting, smoking cessation, academic performance, test preparation, recycling and voting.
Planning prompts are effective for a number of reasons, one of which is that they reduce forgetfulness. When people take the time to create and even write down the when, where and how of a plan, they mentally associate their target actions with cues relating to the when and where of execution.
Planning also discourages procrastination by creating explicit commitments to oneself and sometimes to others. For example, people feel internal pressure to follow-through on their plans and seek to avoid breaking explicit commitments to themselves because behaving inconsistently with their past actions, beliefs, and attitudes creates discomfort. Further, some plans (e.g., to get a mammogram) may literally require making an appointment, which may be difficult to cancel or delay.
Recent large-scale studies have demonstrated the effectiveness of plan-making as a means of increasing take-up of two important should behaviours: receiving flu shots and colonoscopies. In a study by one of the authors (Prof. Milkman), planning prompts were demonstrated to significantly increase take-up of colonoscopies. In this study, those predicted to be the most likely to forget to follow through (e.g., older adults, adults with children, and those who did not comply with previous reminders) benefitted most from the planning prompt.
PRESCRIPTION 2: COMMITTMENT DEVICES Many people are sophisticated about preventing their self-control problems from getting in the way of their good (or should) intentions. As a result, another way to increase engagement in shoulds is by providing individuals with access to commitment devices — or a means of voluntarily (a) enforcing restrictions on themselves until they have done what they know they should or (b) imposing penalties for failing to do what they should. Commitment devices have existed in many forms throughout the years. For instance, the piggy bank is a commitment device that encourages us to commit to saving by setting aside a certain portion of earnings for future use.
More modern forms of commitment devices include Antabuse, a medication that makes alcoholics physically ill after consuming even a small amount of alcohol, and the aforementioned stickk.com, a website that takes users’ money if they fail to achieve their goals. Ultimately, commitment devices are mechanisms that allow people to prevent themselves from giving in to
unwise wants, and the research shows that they can be used to reduce procrastination, smoking, failures to achieve work goals, and succumbing to repeated temptations in a laboratory setting.
PRESCRIPTION 3: TEMPTATION BUNDLING Temptation bundling seeks to increase should behaviours by bundling them with tempting wants — a strategy that can simultaneously reduce engagement in wants and increase engagement in shoulds. For example, a doctoral student may have the goal of spending more time writing a manuscript (a should behaviour) while recognizing that he has been consuming too many Starbucks white chocolate mochas (a want behaviour). Using temptation bundling, the student might commit to only consuming white chocolate mochas while working on his manuscript, thus increasing time spent writing and reducing white chocolate mocha consumption. In addition to simultaneously tackling two types of self-control problems, temptation bundling has the potential to harness consumption complementarities: Working while drinking mochas may make work more enjoyable and efficient, as well as reducing the guilt (and therefore overall enjoyment) associated with chocolate mocha consumption.
Temptation bundling has proven an effective means of increasing engagement in one important should behaviour: exercise. Prof. Milkman et al. demonstrated the effectiveness of temptation bundling in a field experiment where study participants were randomly assigned to one of three conditions:
(a) A full-treatment condition, in which access to tempting, lowbrow audio novels (wants) was restricted to the gym; (b) An intermediate treatment condition, in which participants were simply encouraged to self-restrict their enjoyment of tempting audio novels to the gym; or (c) A control condition.
Their finding: Initial gym attendance among individuals in the full treatment condition was 51 per cent higher than attendance in the control group, and participants in the intermediate treatment condition showed a marginally significant 29 per cent initial increase in gym attendance. At the conclusion of the study, 61 per cent of participants were willing to pay to have their access to an ipod containing tempting audio novels restricted to the gym. In other words, people would pay to have access to a possession they could otherwise use freely restricted so they could only enjoy this desirable want while exercising (i.e. engaging in a should behaviour).
In closing
The better we understand want/should conflict, the more successful we will become at designing effective interventions that promote should choices and help people avoid the temptation to give in to harmful cravings and desires. As indicated herein, even minor interventions can shift behaviours in beneficial directions.