The empathy gap means we make incorrect predictions about our future behavior. It is related to the projection bias, which refers to our tendency to overestimate how much our future self will share the same tastes and preferences as our current self. Both biases mean that we make decisions only considering our narrow, short-term state, instead of considering the fact that emotions influence our current mental state to a great degree. Short-sighted decisions can lead to behavior that is not in our best interest. If we are in a hot mental state, we may make rash and impetuous decisions that cause us to act recklessly. For example, if we receive an email from our boss that makes us very angry, we may send a nasty email back, not considering that our anger would likely eventually subside, and this can have negative consequences on our job.

Alternatively, when we aren’t being influenced by visceral emotions, we may believe that we will have a higher degree of control over our own behavior in situations that these emotions are aroused. For example, imagine you have made a decision to quit drinking and one morning, a friend invites you to a party where other people will be drinking. When you make the decision to go, you are not in a highly emotional state. However, when you get to a party, your visceral state changes and you become very anxious and may be tempted to drink. If you’d been able to predict your behavior in a different mental state, you likely wouldn’t have made the decision to put yourself in a difficult situation.

As these examples demonstrate, the empathy gap acts as an obstacle for making the best decision for our long term goals, and it can occur in either direction, from hot-to-cold or cold-to-hot.

Although there are some exceptions, most economic models, like classical economics and game theory, that are used in order to predict human behavior, such as the expected utility theory, are predicated on the assumption that humans make rational decisions. Emotions are usually left out of the equation, but the empathy gap demonstrates that emotions can manipulate our decision-making process. Although we may strive to be rational decision-makers, we need to acknowledge the way in which our visceral states make us act in order to be able to better predict our future behavior when our emotional mindset has changed. Economic models need to be reconfigured to more aptly reflect the imperfect, often irrational beings that we actually are.

Moreover, while the empathy gap is often linked to our inability to understand how our emotional state affects our own behavior, it follows that we are also inaccurate in predicting other people’s behavior. An interpersonal empathy gap can occur when we fail to consider how other people may be affected by their emotions.² If we’re in a cold state and our friend acts in anger, we may deem their behavior ridiculous and unwarranted because we are not empathizing with the power of their emotions.

Behavioral economists George Loewenstein, Ted O'Donoghue, and Matthew Rabin have suggested that our current emotional states are used as an “anchoring point” for our tastes, behaviors, and beliefs.³ In other words, human cognition is state-dependent, and we rely too heavily on our current mental state to predict our future behaviors. Although we would like to be perfectly rational decision-makers, on a daily basis, we are constantly asked to make predictions about our future behavior, and it’s difficult not to be influenced by emotions.

Empathy gaps can occur on both a cognitive and affective level. A cognitive empathy gap may be the result of a lack of ability. Neurodivergent individuals, such as those with autism spectrum conditions, may experience difficulty inferring other people’s emotional and cognitive states—a skill often referred to as theory of mind.¹⁵ This doesn’t mean they lack empathy altogether, but rather that their way of processing social and emotional information may differ, which can contribute to misunderstandings or mismatches in emotional responses. Similarly, children can find it difficult to understand the perspectives of others.

In terms of affective empathy gaps, individuals may want to avoid empathizing with others’ emotions due to the emotional cost involved. For instance, according to social psychologist C. D. Batson’s model of empathy, empathizing with others can lead to either empathic concern—characterized by feelings of warmth and compassion toward another person—or personal distress, where witnessing someone else’s distress triggers discomfort or anxiety in oneself.¹⁴ Individuals who are more prone to experiencing personal distress rather than empathic concern may be more inclined to avoid situations that require empathy, as such encounters feel emotionally overwhelming. This tendency has been linked to a decrease in prosocial or helping behaviors.

Some studies have suggested that the empathy gap may have something to do with how similar other people appear to us. Jennifer Gutsell and Michael Inzlicht from the University of Toronto measured electroencephalographic (EEG) alpha oscillations while participants observed racial ingroup vs outgroup members (i.e., people from the same or different racial backgrounds).⁶ They found that participants showed similar activation patterns when feeling sad as when they observed ingroup members feeling sad. However, participants did not show the same patterns when watching outgroup members. The researchers concluded that empathy may be limited to those who appear close to us racially or whom we like, potentially intensifying prejudice and intolerance. This phenomenon is also referred to as the outgroup empathy gap.

The empathy gap is not just an individual issue; it has profound implications for how we understand, relate to, and interact with others. As we’ve seen above, the empathy gap may contribute to racism and separation between groups of people from different racial backgrounds. Interpersonal empathy gaps also play a major role in social and political divisions. People in positions of privilege or stability often struggle to understand the experiences of those facing hardship, leading to poorly informed policy decisions.

The empathy gap can also impact ethical decision-making. In a study of judgments concerning end-of-life care, the empathy gap was found to influence college students’ willingness to trade a healthy lifespan for better death scenarios.⁷ The results showed that college students were more inclined to prioritize a longer life over a merciful (painless) death for cancer patients their own age, but were significantly less likely to make the same choice for elderly patients. While this study was hypothetical, young people will likely need to make difficult decisions about an elderly relative’s care during their lives. Understanding the feelings and emotions of the other person is vital in this situation.

On an individual level, the empathy gap is important for understanding our decision-making. On a daily basis, we are asked to make predictions about our future behavior, from deciding what to eat later to making major life choices like financial planning, career moves, and relationships. However, the empathy gap highlights that we are not very accurate in these predictions, leading to decisions that may not align with our long-term goals or best interests.

This disconnect can become problematic in a range of situations. For example, someone intending to save for a house may impulsively purchase an expensive car, believing they will compensate for the expense later. Because we struggle to anticipate how our future selves will feel, we often make choices that prioritize immediate gratification over long-term well-being.

Emotions are incredibly powerful influences that often supersede rationality and logic. This makes it difficult for us to avoid their impact on our decision-making. However, the empathy gap is a problem of not being able to correctly identify the power of emotions. Therefore, while we may not be able to turn off our emotions, we can improve our ability to anticipate and account for empathy gaps. We need to acknowledge the way emotions distort our actions instead of pretending we are always rational decision-makers.

One technique for becoming better at making predictions is to ensure that we consider not only our future actions but our future mental states as well. Before being able to predict what we may do, we need to be able to predict how we may feel.² To do this, we need to use actual data from our past behavior because past actions are the best predictors for the future. For example, if you’re worried that you'll feel awkward at a social event, think back: The last time you felt this way before going out, did the discomfort last, or did you end up enjoying yourself once you arrived?

Visualizing mental states that differ from the one you’re currently in is also an effective way to bridge the empathy gap. This can be applied to understanding someone else’s emotions and perspective or projecting how your future self will react or feel in a situation.

When it comes to reacting to other people’s emotions or mental states, take a moment to ask yourself: Would I want someone to say or do this to me? This reflection can help us to regulate the way we respond to other people’s ‘hot states’ and maybe even make us reflect on how we might feel in the same situation.

The hot-cold empathy gap was coined by George Loewenstein, a well-known and influential figure in behavioral economics.⁴ Loewenstein stated that “affect has the capacity to transform us, profoundly, as human beings; in different affective states, it’s almost as if we are different people”.⁴ Through a series of studies that he has either led or been a part of, Lowenstein has demonstrated the empathy gap in response to pain, addiction, thirst, and fear.⁴

Loewenstein first observed the empathy gap in relation to the powerful emotion of pain in 1999 with Daniel Read, a professor at the University of Leeds.⁴ In the experiment, participants were asked to rate their willingness to endure pain for a monetary reward, a scenario that would occur one week after the question had been posed. The pain they had to endure was putting their hands in ice-cold water. Participants had to indicate how much money they’d have to receive to be willing to endure the pain in three different conditions. The first group was given a sample of what the pain would feel like right before making their decision, the second group had experienced a sample of the pain a week before, and the last group had no experience of the pain.

Read and Loewenstein found that participants in the first group, who had just experienced pain before making their decision, indicated that they would need the largest monetary reward to endure the pain, followed by the participants who had endured it a week later.⁴ Participants who had never experienced a sample of the pain indicated the lowest compensation. The researchers concluded that the pain the participants in the first group were experiencing caused them to make decisions based on their emotional state. Alternatively, the participants who hadn’t experienced the pain were not as likely to have their decision influenced by emotions, which caused them to accept lower monetary compensation. They may have overestimated their willingness to endure pain because they were currently in a rational, calm mindset. Moreover, as the authors also point out, if a person cannot remember their own pain, how likely are they to empathize with the pain of others?

This self-forecasting error was later coined the empathy gap.

Have you ever tried to buy someone a birthday present but struggled to know what they will like and how they will react? The same problem occurs when developing products. A study by consulting firm PwC found that only 38% of customers believe brands truly understand what they need.¹²

The empathy gap can be a real threat to product design teams. It’s impossible to create an excellent and successful product if you can’t empathize with the many stakeholders involved in your product: the people who build, sell, support, and use the product. That’s a lot of people to empathize with.

According to the interpersonal empathy gap model, we struggle to understand the perspective or predict the actions of people who are in mental states different from our own. This could be a problem when designing a mental health app, for instance. A designer who hasn’t got technical support from a professional psychologist may struggle to anticipate the needs, emotions, and decision-making processes of users who are experiencing distress, anxiety, or depression—especially if they are not in that state themself. What seems intuitive or helpful from a neutral perspective may feel overwhelming, dismissive, or unhelpful to someone in crisis.

To bridge this empathy gap, designers can incorporate empathy mapping, user research, real-world testing, and feedback from individuals with lived experiences. These approaches can help designers to put themselves in the shoes of their users.

Bridging the empathy gap is also crucial to successful marketing campaigns. According to IPSOS, campaigns that go the extra mile to truly understand their audience are the most effective.¹¹ Tesco’s ‘Together this Ramadan’ campaign, in partnership with BBH, increased footfall amongst four million British Muslims during the annual holiday because they appreciated the UK supermarket’s commitment to them and their culture. However, as marketing expert Andy Nairn points out, empathy is bidirectional. While it’s important for marketers to put themselves in the shoes of their audience, it’s also crucial for them to invite their audience in to understand the brand. In other words, the brand should understand the audience, and the audience should understand the brand.¹¹ A strong and convincing brand identity can help customers empathize with a product and generate loyalty.

Brands and companies often infuse empathy into their digital assistants and chatbots in order to reduce the empathy gap when customers aren’t happy with a product or service. When registering a complaint online, the first response from a chatbot is usually ‘I’m terribly sorry’ or ‘I understand how you feel’, giving the customer a sense, albeit false, that the brand understands their frustration. Empathy, therefore, is key to customer experience and a way to keep customers, even the most unhappy ones, loyal to a brand.¹³

If you thought that AI doesn’t perceive emotions, maybe think again. In a recent article, TDL’s Content Lead Gabrielle Wasco explores why over half of us are courteous and polite to ChatGPT, even though we’re fully aware that we’re interacting with an algorithm and not a human.

ChatGPT and other machine learning models have what’s called computational empathy, or the ability to detect feelings and generate sincere emotional responses. This concept shouldn’t be confused with real empathy, which is the sharing of emotions. However, recent studies have shown that AI can infer a user’s tonality through word choice and make it appear as though it’s matching your emotional state through its response. This phenomenon has earned LLMs the nickname ‘stochastic parrots’, referring to the fact that they use statistical probability to mimic language patterns without actually understanding them.¹⁰

So is there an empathy gap in AI? Research from the University of Cambridge says yes.⁸ According to a study led by Dr Nomisha Kurian, increasing incidents of children seeing chatbots as quasi-human and trustworthy have sparked concern within the AI community about the potential risk of distress or harm. In one incident from 2021, Amazon’s AI voice assistant, Alexa, challenged a 10-year-old girl to touch a live electrical plug with a coin after the girl asked the chatbot for a “challenge to do.”⁹ Fortunately, the girl’s mother intervened by shouting ‘No, Alexa, no’ before her daughter could act on the suggestion.

Building on these concerning stories, Kurian’s research brought together key cases of human-chatbot interaction and research findings from education, computer science, and human-computer interaction studies and provided practical recommendations for ensuring child-safe AI. AI’s empathy gap lies in the fact that chatbots handle the abstract, emotional, and unpredictable aspects of conversation quite poorly, despite having incredible language skills.

Children also have a hard time understanding the empathy gap between themselves and the chatbots with whom they interact. Research shows that children are more likely than adults to treat AI like humans and that they will disclose more about their own mental health to a friendly-looking chatbot than to an adult. However, children don’t understand AI’s empathy gap and the fact that chatbots are only mimicking human emotion, not actually feeling empathy.

Louis Giordano, a professional counselor, worked with Loewenstein in order to examine whether the empathy gap was involved in the predictive behavior of drug addicts.⁴

Giordano asked drug addicts to predict how much money they would choose over a maintenance drug that helps with withdrawal symptoms, five days from when they were being asked. For example, they were asked to pick whether they’d prefer $10 or another dose of the maintenance drug, $20 or another dose, etc.

The addicts were asked either before receiving a dose of the maintenance drug (opioid deprived) or after. Giordano and Loewenstein argued that individuals experience a significant shift in their state before and after receiving a dose. As a result, the timing of the question—whether it was asked before or after the dose—would influence how much money they were willing to accept instead of receiving the maintenance drug.⁴

The experiment revealed that participants who were asked before receiving a dose of the maintenance drug on average said they’d accept $60 instead of another next dose five days from then, whereas those that were asked the question after receiving a dose on average said they’d choose just $35 over another dose.⁵ From these results, Giordano and Loewenstein concluded that participants who had been asked after receiving the maintenance drug were in a cold visceral state and did not accurately predict how they would feel in a hot state five days later when they had not just received a dose.⁴

As has been mentioned, the empathy gap is a bias that also affects our attitudes towards others. Loewenstein worked with Leaf Van Bowen, professor of psychology and neuroscience, to examine the interpersonal empathy gap.⁴

For their study, Loewenstein and Bowen hypothesized that people would project their own visceral state onto others, thereby allowing their emotions to influence their prediction of other people's behavior.⁴ In the experiment, participants were asked to predict the feelings of hikers who had gotten lost in the woods without food or water, and how they themselves would feel if they were the hikers. After being asked to describe how they believed the hikers would feel, the participants were also asked questions to make predictions about the hikers’ hunger and thirst. One example of a question was “Which would be more unpleasant for the hikers, hunger or thirst?”. Participants were then asked about their own visceral state.

Participants answered these questions in one of two groups. The first group answered these questions immediately before partaking in 20 minutes of cardiovascular exercise, while the other group answered the questions immediately after engaging in 20 minutes of cardiovascular exercise. Loewenstein and Bowen found that participants who answered the questions after exercising were more likely to indicate thirst as the more unpleasant feeling and that not bringing water warranted more regret than not bringing food, both for their predictions of how the hikers would feel and how they would feel if they were in the hikers’ position.⁴

From these results, the researchers concluded that participants who had just exercised before answering the questions were in a hot visceral state and were letting their own thirst influence their predictions both about how they would behave in a certain scenario and how other people would feel. Loewenstein and Bowen concluded that the empathy gap is a bias that affects both self-forecasting predictions and our ability to predict others’ behavior.⁴

What it is

The empathy gap describes our inability to correctly identify how our emotions impact our behavior.

Why it happens

The empathy gap occurs because we underestimate how much emotions impact the decisions that we make, causing us to leave emotions out of the equation when making predictions. We look to our current feelings and attitudes to give us an idea of how we might behave in the future, but depending on whether we are in a cold or hot visceral state, our actions may be very different. There are three types of empathy gaps, categorized by time and person: Intrapersonal Prospective, Intrapersonal Retrospective, and Interpersonal Empathy Gap.

Example 1 - The empathy gap and addiction

Addiction leads to very high emotional states in withdrawal that greatly impact behavior. However, when an addict has just received a maintenance drug and is currently not experiencing withdrawal, they are likely to overestimate how much withdrawal would impact their need for the drug at a later date. Hence, they may believe they’d make decisions similar to actions they take in a cold state.

Example 2 - Projecting our emotions onto others

The empathy gap is not only a bias that makes us unable to predict our own behavior but also makes us less likely to understand the behavior of others who are in a different visceral state. We are likely to project our current emotions onto other people and fail to acknowledge that they might be in a different mental state. 

How to avoid it

It’s almost impossible for us to avoid the influence of emotions on our behavior, so instead, it’s important that we acknowledge their impact. The empathy gap is mostly an issue that causes us to incorrectly predict our future behavior.  Understanding the way we feel, not just rational logic, impacts how we act and enables us to take emotions into account in our predictions.

Empathy Mapping

Empathy mapping is a tool that can help designers understand users’ needs and improve their end product. This article explores what empathy mapping involves and how it relates to user experience. 

Nurturing Children’s Empathy: Opportunities for Behavioral Science

Children can lack empathy. It’s not their fault; they’re passing through a crucial stage in developing socio-emotional skills, including understanding the emotions and preferences of others. However, the COVID-19 pandemic had a huge effect on children’s development and set many back in acquiring key skills and prosocial behaviors. Mariel Guevara explores how behavioral science can reverse this impact.