The Hot-Cold Empathy Gap: "I Would Never Do That"

Imagine asking a person who is well-fed, well-rested, and comfortable how they would behave if they were starving. They can articulate the answer intellectually — "I'd be very hungry, I'd want to eat urgently" — but they cannot actually feel the motivational force of real hunger. They underestimate how consuming it would be. Now imagine asking someone in the grip of rage what they would do if they were calm. They know, in some sense, that they'd be less reactive. But they cannot feel the detached composure of the calm state from inside the furious one. This bidirectional failure is the hot-cold empathy gap: an inability to accurately understand the influence of visceral, emotional, or drive states on behaviour when you are currently in the opposite state.

Loewenstein's Framework

The hot-cold empathy gap was developed and named by George Loewenstein in a series of papers spanning the late 1990s and early 2000s, with a landmark review in 2005 in Psychological Review. Loewenstein distinguished between hot states — conditions of strong visceral drive, including hunger, thirst, sexual arousal, pain, physical craving, and acute emotion — and cold states — the relatively calm, detached, deliberative states in which most planning, self-assessment, and policy-making occurs.

The core claim is not merely that behaviour differs between hot and cold states (that would be obvious). The claim is that people in cold states systematically underestimate how their behaviour will change when they enter a hot state, and that people in hot states systematically underestimate how different they will feel once the hot state passes. The gap is not just experiential — it produces predictable forecasting errors that lead to consequential mistakes in self-regulation, planning, and empathy for others.

The hot-cold empathy gap is closely related to projection bias — both involve projecting current states onto predicted future states — but they are conceptually distinct. Projection bias is primarily about preferences and utility; the empathy gap is about the motivational and behavioural influence of drive states. The empathy gap also has an interpersonal dimension (failing to understand others' hot states) that goes beyond projection bias.

Cold State, Wrong Predictions

The most extensively studied manifestation is the cold-to-hot prediction error: underestimating one's own future hot-state behaviour from a current cold state. Loewenstein's research group ran a now-classic study in which male participants in a neutral state were asked to rate the likelihood that they would engage in various sexually aggressive or unethical behaviours when sexually aroused. They later received sexually stimulating material and were asked the same questions. In the aroused state, they rated all the behaviours as significantly more appealing and more likely — including behaviours they had confidently predicted they would not engage in while calm.

The gap was large and the pattern was clear: cold-state predictions systematically underestimated aroused-state likelihood. This is not simply that arousal changed their preferences temporarily (though it did); it is that their cold-state selves could not adequately model what the hot state would feel like. They failed at empathy for their own future selves.

Similar patterns appear with hunger and food choices, with pain and pain tolerance (people underestimate how much pain will affect their decisions), with fear and risk-taking (people overestimate how brave they'll be under threat when predicting in safety), and with craving and substance use.

Addiction: The Empathy Gap's Most Consequential Domain

Addiction treatment is where the hot-cold empathy gap has perhaps its most significant practical implications. Relapse is predominantly a hot-state event: it occurs when craving is acute, social triggers are present, or emotional distress is high. Decisions about whether to seek treatment, stay in recovery, or avoid triggering environments are often made in cold states. The gap between these two states is vast — and people in recovery often cannot accurately simulate the motivational force of the craving state when planning their behaviour.

This is why "just say no" approaches and cold-state commitments ("I will never use again") consistently underperform in addiction treatment. The commitment was made by someone who could not feel what it would be like to be in the hot state. It wasn't insincere; it was epistemically impoverished. Treatment approaches that acknowledge the gap — by planning responses to hot states in advance, by building environmental barriers that operate independently of in-the-moment willpower, by practising responses to cravings rather than simply resolving to resist them — tend to work better precisely because they don't require the cold-state self to have accurately modelled the hot-state self.

Anger and the "I Would Never Do That" Fallacy

Anger is another domain where the empathy gap produces dangerous overconfidence. People who have never been in a rage cannot fully understand what it is like to be in one — and people who have been in rages, in a calm subsequent state, consistently underestimate the likelihood that they will enter them again. "I would never hit someone" is a cold-state statement about a hot-state event. Research on domestic violence, road rage incidents, and violent altercations consistently finds that perpetrators in cold states before and after incidents describe their behaviour with genuine confusion and often genuine remorse. The hot state was real; the cold-state self could not inhabit it.

This has uncomfortable legal implications. Legal systems are largely designed by cold-state reasoning, applied by cold-state juries and judges, to hot-state behaviour. The law of provocation — which recognises diminished responsibility when someone acts under extreme emotional disturbance — is an institutional acknowledgment of the empathy gap. So is the concept of "heat of the moment" decisions. But the gap also means that juries, evaluating a defendant's hot-state behaviour from a cold deliberative context, may systematically underestimate the motivational force of the state the defendant was in. This is interpersonal empathy gap operating in the courtroom.

The Interpersonal Dimension

The empathy gap is not only intrapersonal. Cold-state observers systematically underestimate the motivational force of hot states in other people. Someone who has never experienced severe clinical depression cannot fully appreciate why a depressed person cannot "just decide to feel better." Someone who has never experienced acute poverty cannot fully understand why someone in desperate financial straits makes decisions that seem obviously self-defeating in calm analysis. Someone who has never been addicted underestimates the compulsion — and designs treatment programs, policies, and social responses accordingly.

This interpersonal empathy gap shapes policy in ways that compound suffering. Drug laws designed by legislators in cold states, applied to people in hot craving states. Poverty reduction programs designed by comfortable policy-makers, for people whose decision-making is dominated by acute scarcity stress. Anger management programs that assume cold-state rationality can be retrieved on demand in hot-state situations. The gap between designer and experiencer is, in many cases, the gap between cold and hot states.

Hot State, Wrong Memories

The reverse direction of the empathy gap — hot-to-cold — also operates, though it is less studied. People in acute pain consistently overestimate how long the pain will last and underestimate the adaptability that cold-state recovery brings. People in the grip of grief cannot feel that the grief will eventually diminish; the hot state projects itself forward. People in states of intense happiness overestimate how persistently happy they will remain (related to projection bias).

This has a direct clinical implication: decisions made in acute distress — including decisions about medical treatment, end-of-life care, relationship dissolution, and financial restructuring — are made by people who cannot accurately model what the non-distressed state will feel like. Advance directives and cooling-off periods in contracts are institutional responses to the hot-to-cold empathy gap: they force a time delay that allows the hot state to pass before the decision is irrevocable.

What Can Be Done

Because the empathy gap operates at a level below deliberate reasoning, cognitive awareness of it offers only partial protection. More effective strategies tend to be structural:

• Pre-commitment devices: Commit in cold states to actions that will constrain hot-state behaviour — locking credit cards in ice, removing substances from the house, setting automatic savings transfers. The cold self makes rules; the hot self operates within them.
• Implementation intentions: Rather than making a general resolution ("I will not lose my temper"), plan specific if-then responses ("If I feel my anger rising in a meeting, I will say I need a moment and leave the room"). This partial simulation of the hot state improves cold-state preparation for it.
• Experiential simulation: Exposure therapies in addiction treatment deliberately put people in partial hot states (exposure to cues) while practising responses, rather than relying on cold-state commitment. This directly addresses the empathy gap by giving the cold self some experiential data about the hot state.
• Institutional design: Policies affecting people in chronic hot states (poverty, addiction, acute stress) should be designed with input from people who have experienced those states, and should assume hot-state behaviour rather than cold-state rationality.

The empathy gap is a reminder that we are multiple selves, visiting different states, unable to fully remember or anticipate the others. "I would never do that" is not a moral statement about character. It is a forecast, made from a state that has no access to the state in which the action will occur. Forecasts, unlike character, can be wrong.

Sources & Further Reading

• Loewenstein, G. "Hot-Cold Empathy Gaps and Medical Decision Making." Health Psychology 24, no. 4 Suppl (2005): S49–S56.
• Ariely, D., & Loewenstein, G. "The Heat of the Moment: The Effect of Sexual Arousal on Sexual Decision Making." Journal of Behavioral Decision Making 19, no. 2 (2006): 87–98.
• Van Boven, L., & Loewenstein, G. "Social Projection of Transient Drive States." Personality and Social Psychology Bulletin 29, no. 9 (2003): 1159–1168.
• Nordgren, L. F., van der Pligt, J., & van Harreveld, F. "The Instability of Health Cognitions: Visceral States Influence Self-assessments of Smoking Dependence." Health Psychology 27, no. 2 (2008): 239–244.
• Wikipedia: Hot-cold empathy gap