Affective forecasting. When and why are our emotional predictions inaccurate?  

The process of predicting future emotional states in response to potential events or experiences is known as affective forecasting. The neural overlap between memory and mental simulation means that reconstructed memories, engaged by brain regions like the hippocampus and ventromedial prefrontal cortex, are not always reliable predictors of future emotions. Common errors include overestimating the intensity and duration of emotion (impact bias), lack of the detail and accuracy of the simulated events (misconstrual of imagined events) and focusing solely on one aspect play a crucial role in prediction accuracy (focalism), with more vivid and detailed simulations leading to more accurate forecasts. Cultural differences and individual cognitive styles also influence these forecasts, with holistic thinkers less prone to errors compared to their more narrow-thinking counterparts. This article will critically discuss how understanding the mechanisms and biases in affective forecasting can help us improve our predictions and lead to better emotional management and decision-making.

Introduction

Affective forecasting describes the process of predicting one’s future emotional states, such as how an event or experience will make us feel. It is a vital strategy we regularly employ – even if unwittingly – to protect ourselves from risky situations, unpleasant feelings, and uncertainty about the future (Wilson & Gilbert, 2003). We mentally forecast by drawing on previous experiences and memories, generating and simulating mental images of what we predict will happen and then assessing how that will affect us emotionally. It is crucial to explore this process, as simulated future emotional predictions are one of the core sources we use when making decisions, and they are often erroneous, thus providing an unreliable basis for our future behaviour. The more intense the expected emotional response to an outcome, the greater the effort and resources people will invest in either achieving or avoiding it (Levine et al., 2012), which is why understanding affective forecasting provides valuable insights into human decision-making, risk assessment, and psychological well-being.

Affective forecasting and the brain

The brain’s emphasis on memory stems from its role in generating predictions, which provide us with a proactive advantage in assessing risk and reward in each situation (Gilbert & Wilson, 2005). Affective forecasting and memory share much of the same neural architecture, meaning we mentally simulate the future by employing many brain regions associated with memory recollection, such as the hippocampus. The hippocampus is crucially involved in memory storage and formation and also plays a significant role in constructing imagined future scenarios (Schacter & Addis, 2007; Voss et al., 2017). By generating events in our minds, we predict future feelings. However, because we have an easier time remembering than imagining, affective forecasting often yields inaccurate predictions when our imagined events differ from actual situations (Schacter et al., 2008).

Future emotions are forecasted by drawing on past experiences, which makes the quality of the memories on which predictions are based essential. The constructive nature of memory entails that simulations of future events are constructed around and based on autobiographical memories. Autobiographical memories are liable to distortions in emotional memories and affective forecasting errors because recollections are not replayed but reconstructed in the mind (Conway & Pleydell-Pearce, 2000). Practically, this means that, for example, when we remember breaking a bone, we can be misled into thinking we are remembering how it felt, whereas we are actually recalling the event and emotionally reacting to its memory (Schacter, 2012). This happens because, with increasing time between the event and recall, emotional reports are generated from semantic networks where knowledge is general and stored conceptually, independently from the event from which it was extracted. As a result, episodic details fade and are no longer an accurate basis for emotional predictions (Robinson & Clore, 2002). Therefore, when people engage in affective forecasting, they draw on memories that might not be reliable representations of past events. Affective forecasts are made from simulated reactions to memories, not the memories themselves, which can mislead our emotional predictions.

Because emotions are one of the primary drivers of decision making which results from affective forecasting, the somatic marker hypothesis offers insight into how emotions influence affective forecasting and decision making by suggesting that somatic markers are bodily sensations associated with emotions, which form the basis of our decisions, for example goosebumps associated with excitement. A crucial element of the hypothesis is the ‘body-loop,’ which posits that emotional events manifested in the body can impact decision-making through sensory feedback to the brain (Poppa & Bechara, 2018). Especially involved is the ventromedial prefrontal cortex, which is key for processing emotional arousal by regulating and inhibiting responses to it, thereby contributing to decision-making (Damasio, 1998). Lesion studies have shown how participants with damage to the ventromedial prefrontal cortex exhibit a general reduction in emotional experience, poorly regulated emotional responses, impaired decision-making, particularly in social contexts, difficulties in goal-directed behavior, and a notable lack of self-awareness (Barrash et al., 2000). Illustratively, those with damage to the area exhibit a complete lack of awareness of the future consequences of their actions and appear to only account for immediate outcomes (Bechara et al., 1994). The ventromedial prefrontal cortex, therefore, appears significantly involved in affective forecasting through its link to emotion regulation and inhibition. This idea is further supported by empirical findings that individuals with damage to the ventromedial prefrontal cortex experience difficulties in predicting their emotional reactions to simulated events (Bechara & Damasio, 2005). All this suggests that when we construct mental simulations or anticipatory glimpses of future occurrences, we inevitably trigger our affective responses too. These emotional reactions serve as a foundation for our predictions regarding the probable emotional outcomes the events will have on us, influencing our perceptions and guiding our behaviour (Damasio, 1997; Gilbert, 2006).

Why does affective forecasting go awry?

The most common factor impacting affective forecast accuracy is our tendency to overestimate the intensity and duration of the impact future events will have on us, known as the impact bias (Wilson & Gilbert, 2005). The overestimation of emotional intensity and the underestimation of the rate at which feelings dissipate leads us to simulate inaccurate versions of our future selves. For example, during the COVID-19 pandemic, people who predicted their feelings regarding a future infection reported feeling significantly more negatively than participants who recalled their feelings during an actual past infection (Dillard & Meier, 2023). This means participants who had not had the virus could only simulate the experience and imagined it more negatively than it actually was. Moreover, perpetrators of harm have been shown to overestimate the intensity and duration of negative feelings surrounding apologizing to those harmed, compared to the intensity and duration of negative feelings perpetrators who apologized actually experienced (Leunissen et al., 2014). Similarly, researchers found people’s expectations of positive events, such as visiting a desired holiday destination, were more positive than their feelings during the actual vacation (Mitchell et al., 1997). This indicates that our emotional future is usually less intense and durable than we imagine it to be. We feel less positively when something positive happens and less negatively when something negative happens than we predict we will. But why is that so?

One of the most common sources of impact bias is focalism; the tendency to focus all attentional resources on a particular event while failing to consider other contextual factors that influence emotions (Morewedge & Buechel, 2013). This occurs when attention is overly focused on essential features of, for example, a birthday party, while incidental features, such as getting ready and buying a present, are disregarded. This is problematic because when incidental features are compiled, they have an important bearing on our feelings. An elegant illustration of this is a study with college football fans and predicted happiness (Wilson et al., 2000). Two months in advance, two groups of fans were interviewed. One group of fans was asked to predict how happy they would feel immediately after their team won a game. A day after their team actually won, they were asked to report how happy they felt. Simultaneously, another group was first asked to imagine a future day and predict how many hours they would spend on different everyday activities, and then asked to predict their happiness levels if their team won.. After their team won, they were asked to report levels of experienced happiness. Researchers found people who were not asked to think about the contextual setting of the event exhibited an impact bias, predicting they would feel happier than they actually felt after their team won, compared to people who considered the contextual embeddedness of other everyday activities and correctly predicted the game outcome would be less influential on their affective state. Similarly, a study on unethical behavioural tendencies showed focalism and consequently unethical behaviour can be reduced by prompting individuals to think about different pursuits that bring them happiness other than a specific desired goal, e.g. obtaining a sum of money. When focalism was reduced, people were less focused on the one goal, did not overestimate their emotional future and were consequently less likely to engage in unethical behaviour to achieve that goal (Noval, 2016). Despite this being a concrete and widespread finding (Lench et al., 2011; Wilson & Gilbert, 2013), some literature has shown focalism is not as universal as perhaps believed. Compared to the generally linear thinking tendencies of Western culture, East Asia has an intellectual tradition that focuses more on holistic thinking and understanding, which has been shown in relation to affective forecasting as well (Lam et al., 2005; Markus & Kitayama, 1991). In a study investigating cultural differences in focalism, participants from East Asia were less prone to focalism effects and exhibited less impact bias than Western participants (Lam et al., 2005). Interestingly, when Western participants were guided not to focus only on one aspect, they showed equally accurate affective forecasts as East Asian participants did generally. These findings suggest that when one’s thinking is naturally or with guidance less focused on one particularity, affective forecasts are more indicative of future affective experiences and aid both ethical and considerate decision-making. In contrast, if thinking is too narrowly focused on one aspect, overestimation, inaccurate predictions and selfishness occur. This nuance suggests that individual differences and ways of thinking contribute to the accuracy of affective forecasts and consequently behaviour.

Another way in which affective forecasts go awry is through misconstrual of imagined events, namely when the content of the simulated event is dissimilar to the content of the actual event. How a simulated event is construed has radical implications for the accuracy of predictions – the more detailed and accurate a simulation of an event is, the more likely predictions will be accurate (Schacter, 2012). For example, people have been shown to make overconfident predictions about time and money expenditures for themselves and others in uncertain situations compared to specified situations (Griffin et al., 1991). This suggests a failure to account for the option that one’s momentary conceptualization is only one of the possible conceptualizations, rather than the only possible one. When participants had specificity regarding the event, there was less room to simulate their own images, prompting them to act on the specified ones, which were more accurate for forming the basis of affective predictions. There is no limit to how inaccurate construal can be, and the more inaccurate they are, the less likely they are to yield accurate affective predictions. To demonstrate this further, a study investigating Virtual Reality (VR) exposure and predicted happiness showed people predicted higher levels of future happiness and decisions to visit a holiday destination after viewing a preview in VR with vivid 3D images, compared to people who viewed regular 2D images (Skard et al., 2021). This indicates that those exposed to a more detailed, immersive, and accurate representation of the holiday destination anticipated stronger positive future feelings, resulting in them committing to visit the destination. Because simulation is essential to affective forecasting, the finding gives credence to the assumption that increased detailedness of simulated events leads to better construction, enabling people to imagine and predict from forecasts more accurately. However, this was the first study examining VR and affective forecasting, and its primary focus was influencing consumer behaviour, so further research is needed to compare affective predictions of future happiness with people experiencing events in real-life versus VR settings.

Lastly, another content-based error in affective forecasting occurs when prior expectations about an event change one’s current emotional experience (Gilbert & Wilson, 2003). Expectations serve an important role in preparing us for possible outcomes and shielding us from the pain and discomfort of receiving bad or unfavourable news without warning (Kahneman et al., 1999). Anticipating hearing something undesirable supposedly lessens the negative impact of receiving such information, thus shielding from disappointment and lowered self-esteem. However, studies examining prior expectations and post-event affect have found that people with low self-esteem almost exclusively expect greater chances of failure on tasks, but after failing, they report feeling even worse than they expected and even worse than people with higher self-esteem who expected to succeed but failed (Brown & Marshall, 2001; Marshall & Brown, 2006). The suggested reason for this occurrence is linked to general personal tendencies. Namely, people who expect failure tend to be more pessimistic and view the world through a negative lens, preventing overly positive feelings from arising. Therefore, if one possesses a negative state of mind or tendency, the baseline from which their affective forecasts are generated is negatively skewed from the onset. Illustratively, people with generalized anxiety or major depressive disorders exhibited heightened experiences of negative affect, and overestimated negative future affect even more than control groups, with this overestimation being apparent and significant in both affective forecasts and past memories (Dev et al., 2023). Therefore, for people with certain mental health conditions or predispositions, such negative tendencies form the basis for expectations, which is another way emotional predictions are influenced and varyingly accurate. Moreover, if we approach an event with strong expectations, those expectations tint the emotional lens through which we react. For example, students who were led to anticipate being more bored during a lecture reported higher levels of boredom afterwards than those who did not have such expectations (Tam et al., 2023). This process of assimilation occurs when experiences come to align with their expectations—in this case, when participants were primed to be bored, their experience reflected their anticipation. However, Marshall and Brown (2006) found that these assimilation instances are measured and reported immediately after the event in question, making the expectation still salient and active in memory. When the time interval between the event, such as getting a grade for an examination, and one’s expectation of it increased, people’s expectations faded from memory, and their emotional reports were based on the memory of the event itself and not their expectations (Golub et al., 2009). This means that expectations, while important for our feelings, might not be as important as quality of memories is for affective forecasting accuracy, which is what literature has attested to before (Schacter & Addis, 2007). Nonetheless, this phenomenon raises critical concerns about the potential for expectations to create self-fulfilling prophecies, particularly in individuals with predisposed negative outlooks or mental health conditions (Mathersul & Ruscio, 2020). It suggests that the power of expectations may lie not just in their ability to protect us, but also in their capacity to create negative emotional cycles, and that those with stronger symptoms of mental health disorders like depression and anxiety show biased memory recollections which form the basis of affective forecasts (Rizeq, 2024). Instead of protecting, they end up perpetuating a maladaptive, negative view of the future, how events are perceived, and which decisions are made.

Conclusion

To conclude, our ability to predict future emotional states is crucial for decision-making but often inaccurate. These inaccuracies stem from memory distortions, impact bias, and factors like focalism, misconstrual of events, and prior expectations. Understanding these errors is essential for improving our predictions and decision-making processes. By adopting holistic thinking, considering broader contexts, refining expectations, and using technologies like virtual reality for detailed simulations, we can enhance the accuracy of our emotional forecasts. Such an understanding helps to improve self-awareness and cognitive strategies, ultimately benefiting our psychological well-being, ethical behaviour and informed decision-making. As research on affective forecasting and VR is scarce, future studies could examine how VR-enhanced affective forecasting influences decision-making and behaviour. For example, if VR simulations lead to more accurate predictions of future emotional states, does this translate into better decision-making, such as avoiding impulsive choices or reducing regret after a decision? Studies could track participants’ decisions and outcomes over time to determine whether VR can be a valuable tool in improving overall life satisfaction and reducing decision-related stress. Such an undertaking would bring clinically relevant insight which would help us better understand how affective forecasting works and how we can be helped in lessening its errors or making it an adaptive instead of a maladaptive mechanism.

References

Barrash, J., Tranel, D., & Anderson, S. W. (2000). Acquired Personality Disturbances Associated With Bilateral Damage to the Ventromedial Prefrontal Region. Developmental Neuropsychology, 18(3), 355–381. https://doi.org/10.1207/S1532694205Barrash

Bechara, A., Damasio, A. R., Damasio, H., & Anderson, S. W. (1994). Insensitivity to future consequences following damage to human prefrontal cortex. Cognition, 50(1), 7–15. https://doi.org/10.1016/0010-0277(94)90018-3    

Bechara, A., & Damasio, A. R. (2005). The somatic marker hypothesis: A neural theory of economic decision. Games and Economic Behavior, 52(2), 336–372. https://doi.org/10.1016/j.geb.2004.06.010

Damasio, A. R. (1997). The somatic marker hypothesis and the possible functions of the prefrontal cortex. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 351(1346), 1413–1420. https://doi.org/10.1098/rstb.1996.0125

Damasio, A. R. (1998). Emotion in the perspective of an integrated nervous system1. Brain Research Reviews, 26(2), 83–86. https://doi.org/10.1016/S0165-0173(97)00064-7

Golub, S. A., Gilbert, D. T., & Wilson, T. D. (2009). Anticipating one’s troubles: The costs and benefits of negative expectations. Emotion, 9(2), 277–281. https://doi.org/10.1037/a0014716

Griffin, D., Dunning, D., & Ross, L. (1991). The Role of Construal Processes in Overconfident Predictions About the Self and Others. Journal of Personality and Social Psychology, 59, 1128–1139. https://doi.org/10.1037//0022-3514.59.6.1128

Kahneman, D., Diener, E., & Schwarz, N. (1999). Well-Being: Foundations of Hedonic Psychology. Russell Sage Foundation.

Lench, H. C., Safer, M. A., & Levine, L. J. (2011). Focalism and the underestimation of future emotion: When it’s worse than imagined. Emotion, 11(2), 278–285. https://doi.org/10.1037/a0022792

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224–253. https://doi.org/10.1037/0033-295X.98.2.224

Marshall, M., & Brown, J. D. (2006). Emotional reactions to achievement outcomes: Is it really best to expect the worst? Cognition & Emotion – COGNITION EMOTION, 20, 43–63. https://doi.org/10.1080/02699930500215116

Mitchell, T. R., Thompson, L., Peterson, E., & Cronk, R. (1997). Temporal Adjustments in the Evaluation of Events: The “Rosy View”. Journal of Experimental Social Psychology, 33(4), 421–448. https://doi.org/10.1006/jesp.1997.1333

Morewedge, C. K., & Buechel, E. C. (2013). Motivated underpinnings of the impact bias in affective forecasts. Emotion, 13(6), 1023–1029. https://doi.org/10.1037/a0033797

Noval, L. J. (2016). On the misguided pursuit of happiness and ethical decision making: The roles of focalism and the impact bias in unethical and selfish behavior. Organizational Behavior and Human Decision Processes, 133, 1–16. https://doi.org/10.1016/j.obhdp.2015.12.004

Poppa, T., & Bechara, A. (2018). The somatic marker hypothesis: Revisiting the role of the ‘body-loop’ in decision-making. Current Opinion in Behavioral Sciences, 19, 61–66. https://doi.org/10.1016/j.cobeha.2017.10.007

Schacter, D. L., & Addis, D. R. (2007). The cognitive neuroscience of constructive memory: Remembering the past and imagining the future. Philosophical Transactions of the Royal Society B: Biological Sciences, 362(1481), 773–786. https://doi.org/10.1098/rstb.2007.2087

Schacter, D. L., Addis, D. R., & Buckner, R. L. (2008). Episodic Simulation of Future Events. Annals of the New York Academy of Sciences, 1124(1), 39–60. https://doi.org/10.1196/annals.1440.001

Schacter, D. L. (2012). Adaptive constructive processes and the future of memory. American Psychologist, 67(8), 603–613. https://doi.org/10.1037/a0029869

Skard, S., Knudsen, E. S., Sjåstad, H., & Thorbjørnsen, H. (2021). How virtual reality influences travel intentions: The role of mental imagery and happiness forecasting. Tourism Management, 87, 104360. https://doi.org/10.1016/j.tourman.2021.104360

Tam, K. Y. Y., Van Tilburg, W. A. P., & Chan, C. S. (2023). Whatever will bore, will bore: The mere anticipation of boredom exacerbates its occurrence in lectures. British Journal of Educational Psychology, 93(1), 198–210. https://doi.org/10.1111/bjep.12549

Voss, J. L., Bridge, D. J., Cohen, N. J., & Walker, J. A. (2017). A closer look at the hippocampus and memory. Trends in Cognitive Sciences, 21(8), 577–588. https://doi.org/10.1016/j.tics.2017.05.008

Wilson, T., Wheatley, T., Meyers, J., Gilbert, D., & Axsom, D. (2000). Focalism: A Source of Durability Bias in Affective Forecasting. Journal of Personality and Social Psychology, 78, 821–836. https://doi.org/10.1037/0022-3514.78.5.821

Wilson, T. D., & Gilbert, D. T. (2003). Affective forecasting. In M. P. Zanna (Ed.), Advances in experimental social psychology, Vol. 35, pp. (345–411). Elsevier Academic Press. https://doi.org/10.1016/S0065-2601(03)01006-2

Wilson, T. D., & Gilbert, D. T. (2005). Affective Forecasting: Knowing What to Want. Current Directions in Psychological Science, 14(3), 131–134.

Wilson, T. D., & Gilbert, D. T. (2013). The impact bias is alive and well. Journal of Personality and Social Psychology, 105(5), 740–748. https://doi.org/10.1037/a0032662