Teaching Empathy
Empathy is partly dispositional and partly learned. What developmental psychology tells us about when it emerges, how it can be cultivated, and what neurodiversity considerations change about the picture.
How Empathy Develops in Children
Empathy is not present from birth in its full form, but its precursors are. Newborns respond to other infants' crying by crying themselves, a phenomenon documented by Sagi and Hoffman (1976) and often cited as the earliest evidence of emotional contagion. By 6-12 months, infants show clear signs of social referencing: looking to caregivers' emotional expressions to calibrate their own responses to ambiguous situations.
The more complex forms of empathy develop in stages, linked to broader cognitive development:
| Age Range | Empathic Development Stage | What This Looks Like |
|---|---|---|
| 0-12 months | Global empathy / emotional contagion | Crying when another baby cries; mirroring caregiver's emotional tone |
| 1-2 years | Egocentric empathy | Recognises another's distress and tries to help, but help is based on what would comfort themselves (offering their own toy or comfort object) |
| 2-3 years | Theory of mind begins | Starts to understand that others can have different thoughts and feelings; false belief tasks show mixed capability |
| 4-6 years | Perspective-taking emerges | Able to understand that others' feelings may differ from their own; begins tailoring comfort to the other person's preferences, not their own |
| 7-10 years | Empathy for others' life circumstances | Able to empathise with situations they have not directly experienced; understanding of chronic or invisible difficulties begins |
| Adolescence | Social identity and in-group/out-group empathy | Strong empathy within peer group; potential for reduced empathy toward out-groups; development of moral reasoning |
Research by Martin Hoffman, whose developmental model spans from the 1970s through to his 2000 book Empathy and Moral Development, is the standard reference here. Hoffman argues that empathy is the motivational foundation for moral behaviour, and that its development depends critically on both biological maturation (cognitive capacity for perspective-taking) and environmental input (modelling by caregivers and educators).
What Research Says About Teaching Empathy
A widely cited meta-analysis by Zins et al. (2004) and subsequent work on social-emotional learning (SEL) programmes found that structured SEL programmes, which typically include empathy-building components, produce measurable improvements in prosocial behaviour and reductions in aggression and bullying.
Durlak et al.'s 2011 meta-analysis of 213 SEL programmes involving 270,000 students found that SEL participants showed an 11% increase in academic achievement, significant reductions in conduct problems, and improved social skills, compared to control groups. Empathy skills were a core component of these programmes.
The research on what specifically works within empathy teaching includes:
Labelling emotions explicitly
Children who can name emotions accurately are better able to recognise them in others. Research by Brackett et al. (2011) and others shows that emotional vocabulary (the ability to distinguish between disappointed, frustrated, embarrassed, and sad) is a teachable skill that underpins empathic accuracy.
Perspective-taking exercises
Activities that require children to consider a situation from another person's point of view, including role play, reading fiction about characters different from themselves, and structured discussion of disagreements, all show positive effects on perspective-taking ability.
Reading fiction
Multiple studies, including work by Mar, Oatley, and Peterson (2006) and Kidd and Castano (2013), find that reading literary fiction is associated with improved theory of mind and empathic accuracy. The mechanism is exposure to the inner lives of characters whose experiences differ from the reader's own.
Modelling by adults
Children learn empathy partly by observing it. Caregivers and teachers who name their own emotions, acknowledge others' feelings explicitly, and respond empathically to children's distress (rather than dismissing or punishing emotional expression) provide the developmental scaffold for empathy development.
Neurodiversity Considerations
The relationship between neurodivergent conditions and empathy is significantly more complex than popular characterisations suggest. The persistent myth that autistic people "lack empathy" has been challenged by researchers and autistic advocates for more than a decade and does not reflect the current scientific consensus.
Autism and the Double Empathy Problem
Damian Milton, an autistic researcher, proposed the "double empathy problem" in a 2012 paper in Disability and Society. Milton argued that empathy failures in autism research have been measured almost exclusively in terms of autistic people's ability to understand neurotypical social norms and signals. When the measurement is reversed, neurotypical people show equally poor ability to understand autistic social cues and communication.
Subsequent research has supported this framing. Studies by Crompton et al. (2020) found that autistic people communicate more effectively with other autistic people than with neurotypical people, and that information transfer in autistic-autistic dyads is comparable to neurotypical-neurotypical dyads. The communication difficulty is mutual, not unidirectional.
Many autistic people report high levels of affective empathy (strong emotional resonance with others' feelings), sometimes to an overwhelming degree, while experiencing difficulty with the cognitive empathy component (accurately reading others' emotions from typical social cues). This is essentially the inverse of the stereotype.
ADHD, Alexithymia, and Empathy
Children and adults with ADHD often experience empathy through a different pattern: strong emotional reactivity in the moment, combined with difficulty regulating and communicating that reaction, can make them appear less empathic even when they are intensely affected by others' emotions.
Alexithymia (difficulty identifying and describing one's own emotions) affects approximately 10% of the general population and is more prevalent in neurodivergent individuals. Research by Bird and Cook (2013) found that when alexithymia is controlled for in autism research, many of the apparent differences in emotional empathy between autistic and neurotypical groups reduce significantly.
The practical implication for teaching empathy with neurodivergent children is that standard approaches (which rely heavily on reading facial expressions, inferring emotional states from behaviour, and producing socially conventional empathic responses) may not work and should not be the benchmark. What matters is whether the child cares about others' wellbeing and acts to support it, not whether they express that care in neurotypically legible ways.