Going to the supermarket when I’m really hungry, and without a shopping list, is a recipe for disaster. It will take an act of iron will to avoid returning without some kind of junk food. Later, after eating, I’ll wonder how I could have bought junk food but forgotten healthy staples like rice and pasta.
Why should this be? After all, I know very well what sort of food I should buy; I’ve been hungry in the supermarket before and bought junk food and regretted it later. The reason is that in the moment, when I’m hungry at the supermarket, I’m out of touch with my future emotional self – something that psychologists have confirmed experimentally.
Research has shown that we can have considerable difficulty predicting our future requirements because our current emotional states override them. This is called the projection bias and it occurs despite the fact that we have plenty of experience of the problem and its undesirable consequences.
So, what is it about the projection bias that means I choose junk food in the supermarket when I’m hungry? Well, what studies have shown is that when we are in a ‘hot’ emotional state – in this case hungry – we tend to start ignoring the long-term in favour of short-term self-gratification. We think that because we are desperate for a quick fix of calories right now, we’ll always be desperate for a quick fix. This means that when I’m hungry I forget my commitment to eat healthily and concentrate on what will give me an immediate rush of calories and pleasure: junk food.
But this still doesn’t fully explain why I’m not able to override my preference for buying junk food when I’m hungry. Why can’t I just imagine my future self – after lunch – looking at the junk food I’ve bought and empathise?
Well, I can, but only a little. What studies have shown is that not only do we have this projection bias but we have what researchers have a called an ’empathy gap’ – this gap is effectively a failure to empathise with our future selves. In the moment, driven by our emotions, our present desires can suddenly override our long-term goals and we no longer care about healthy eating or any other promises we have made ourselves.
Healthy or unhealthy snack?
This effect is neatly demonstrated in a study conducted by Read and van Leeuwen (1998). Office workers were told they could have a free snack in a week’s time and were asked to choose between a healthy snack (e.g. an apple) or an unhealthy snack (a Mars bar). The trick is that they were either asked just after lunch or in the late afternoon. This meant they were either satiated or hungry. Before deciding they were also told at what time of day they would receive their free snack: either after lunch or late afternoon.
A week later experimenters returned at the time they stated, depending on which experimental condition each person was in. But this time participants were given the option to change their choice from a week ago. Again they were either asked after lunch when they were satiated or at the end of the day when they were hungry. This created four different experimental conditions.
Give me chocolate right now!
Predictably they found that people were much more likely to choose the healthy snack for consumption in a week’s time, than a week later when the snack could be eaten right now. Overall people predicted that they would be considerable more health-conscious than they actually were.
But what the researchers were interested in was how big the empathy gap was depending on whether people were hungry or not at the two initial time-points. Exactly how much empathy did people have with their future selves, depending on whether they were hungry or satiated, and whether they imagined their future selves as hungry or satiated?
To understand the results, let’s compare two pairs of conditions. In the first pair of conditions participants were asked after lunch, when they were not hungry, which snack they would like in a week. But in one they were told they would be given their snack after lunch and in the other condition late afternoon. So people who were not hungry now were trying to predict their preference for firstly when they are satiated in the future and secondly when they are hungry in the future. 56% predicted their future self would choose an unhealthy snack while hungry, but when they were actually presented with the choice (when hungry), 88% chose the unhealthy snack.
In comparison, when they thought their futures self would NOT be hungry only 26% predicted they would choose the unhealthy snack, but at the actual time of choosing 70% chose the unhealthy snack.
There are two interesting things about this result. The first is that people found it difficult to predict their future desires for food, even when they weren’t currently hungry. This suggests that the projection bias works both ways. It isn’t just in our hot, emotional moments that we’re poor at predicting our future desires, it’s also when we’re in a relaxed, neutral state as well.
The second interesting thing is that our empathy gap with our future selves seems to be about 30% (56% minus 26%). Here the empathy gap refers to the difference between the judgement we make when in the same state as our future selves and when in a different state. It’s important to note that in defining the empathy gap the authors are comparing two different advance choices and not the advance choice with the actual choice. This means the results can’t be explained by saying that people were just trying to make themselves choose the healthy option by predicting that’s what they would have.
Similar results were seen when hungry participants tried to make predictions about their future selves, either when hungry or satiated. Of those who were hungry 78% predicted they’d go for the unhealthy snack, while 92% actually chose it. Of those who were satiated, 42% thought they’d go for the unhealthy snack while 82% actually chose it when hungry. The empathy gap for those who were hungry while making their prediction was about the same as the previous pair of conditions at 36% (78% minus 42%).
One positive to emerge from the results, however, is that our predictions, inaccurate as they may be, do have some control over our future behaviour. The researchers found that people whose intention was to eat healthily were less likely to change their mind than those whose intention was to eat unhealthily. So it’s still better to promise yourself you will eat healthily because it gives you a better chance of resisting temptation when it comes along.
Summary and explanation
Overall this study demonstrates that people directly project their current emotional state (of hunger or satiety) into the future, forgetting they will probably feel differently when the future becomes the present. This is the projection bias. On top of this we seem to project both neutral or ‘cold’ emotional states into the future as well as ‘hot’ ones.
The exact nature of the projection bias depends on whether our current state is compatible with the future state we are imagining. If it is incompatible (actually hungry versus predicting satiety), we are consistently less accurate.
Despite this study’s focus on hunger, these findings probably do extend to other domains, some of which, such as happiness, have already been documented.
false consensus bias, the hindsight bias and the knowledge bias – is suggested by Loewenstein, O’Donoghue and Rabin (2003). They argue that these phenomena have an ‘anchoring and adjustment’ characteristic. We are each anchored in our current emotional and cognitive state and the only adjustments we can make are with respect to this anchor.A useful way to think about how the projection bias works – and, actually, many other common biases such as the
Consequently people tend to assume that others have the same knowledge as them, that other people tend to generally hold the same opinions as them and, that, broadly speaking, they will be in the same emotional state in the future as they are right now. And that is the projection bias and just one of the reasons why we tend to have trouble predicting exactly what it is that will make us happy in the future.
[Image credit: funadium]
Loewenstein, G., O’Donoghue, T., & Rabin, M. (2003). Projection Bias In Predicting Future Utility. Quarterly Journal of Economics, 118(4), 1209-1248.
Read, D., & van Leeuwen, B. (1998). Predicting Hunger: The Effects of Appetite and Delay on Choice. Organizational Behavior and Human Decision Processes, 76(2), 189-205.
→ This post is part of a series on cognitive biases:
- The Dunning-Kruger Effect: Why The Incompetent Don’t Know They’re Incompetent
- The Worse-Than-Average Effect: When You’re Better Than You Think
- Why You’re a Sucker for the Impact Bias
- The Hindsight Bias: I Knew It All Along!
- How to Overcome the Egocentric Bias
- See How Easily You Can Avoid The Memory Bias
- Why Your Future Self is an Emotional Mystery: The Projection Bias
- How To Avoid Choosing the Wrong Job or House: Fight the Distinction Bias
- 4 Belief Biases That Can Reduce Pleasure
- Does Delaying Decisions Lead to Better Outcomes?
- The Belief in a Just World: A Fundamental Delusion
- Why Society Doesn’t Change: The System Justification Bias
- The Availability Bias: Why People Buy Lottery Tickets
- The Illusion of Transparency
- The Illusion of Control: Are There Benefits to Being Self-Deluded?
- The Endowment Effect: Why It’s Easy to Overvalue Your Stuff
- Illusory Correlations: When The Mind Makes Connections That Don’t Exist
- Anchoring Effect: How The Mind is Biased by First Impressions
- The Confirmation Bias: Why It’s Hard to Change Your Mind
- The Well-Travelled Road Effect: Why Familiar Routes Fly By
- How a Psychological Bias Makes Groups Feel Good About Themselves And Discredit Others
- The Sobering Up Effect: Why People Get More Pessimistic As The Moment of Truth Gets Closer
Published: 12 May 2008