Heider’s Balance Theory in psychology proposes that people strive for cognitive consistency in their attitudes and perceptions, particularly in triadic interpersonal relationships.
It suggests that when there’s an imbalance, such as liking someone who dislikes something you also like, people feel motivated to restore balance. They do this by changing an attitude or a perception.
Balance feels comfortable. Imbalance does not, and that discomfort is what drives the change.
The social psychologist Fritz Heider developed the theory, first in a 1946 paper and then fully in his 1958 book.
Heider originally applied it to relationships between people, but later extended it to attitudes toward objects, ideas, and events.
Key Takeaways
- P-O-X Triad: Heider modelled attitudes as three linked elements, a person (P), another person (O), and a third element (X), each relation either positive or negative.
- Sign Rule: A triad is balanced when the product of its three signs is positive, and unbalanced when the product is negative.
- Discomfort Drives Change: An unbalanced triad feels uncomfortable, motivating a change in attitude, perception, or relationship to restore balance.
- Wider Reach: Cartwright and Harary later reframed the theory using graph theory, extending it from three-person triads to whole social networks.
- Modern Evidence: Recent network-science studies support the theory at scale, though which version, strong or weak balance, fits real networks best remains debated.
What is Balance Theory?
Heider first set out this idea in a short 1946 paper, Attitudes and cognitive organization. He argued that a person’s likes and dislikes are not separate. They form one organised, coherent whole.
He developed the theory fully in his 1958 book, The Psychology of Interpersonal Relations, the same book that founded attribution theory. Heider called this project a “common-sense psychology.” It was an attempt to reconstruct how ordinary people already reason about their social world.
Heider explained that a balance must exist between interpersonal relationships or for something specific between two or more individuals so that psychological harmony can be achieved.
If two or more people share similar ideas about something, there is not likely to be any tension or complication surrounding this idea in the relationship.
The key notion of balance theory is that certain structures are balanced, whereas others are imbalanced. Balanced structures are usually preferred over imbalanced ones.
Imbalanced structures are associated with uncomfortable feelings, and this is what leads people to seek to achieve balance.
Heider suggested that ‘likes’ and ‘dislikes’ are related to balance and imbalance. Humans search for consistency between their attitudes and relations with others to make the balance neutral.
However, if there is an inconsistency between relations and a perceived imbalance, humans tend to seek modifications to reach consistency and thus cognitive harmony of the situation again.
In this way, balance theory describes how humans are motivated to change their attitudes.
How does balance theory work?
Heider explained how balance theory works with a simple model of interpersonal relationships: the P-O-X triad.
The triangle has three corners.
Each corner represents a different element.
P = the person to analyze.
O = Other, or the comparison person.
X = the third element for comparisons, such as a third person, a physical object, an idea, or an event.
Psychologists call relation patterns between three individuals or objects “triadic” relations.
The P-O-X model shows the positive and negative relationships between what each person perceives, both toward another person and toward an object.
Each relation is either positive (+) or negative (-). This combination of signs determines whether the relationship is balanced.
A triadic relationship is thought to be balanced when it includes either no or an even number of negative relations.
For instance, three positive relations are balanced, as well as two negative relations with one positive relationship. This would be a situation that is psychologically comfortable.
In contrast, a triad is imbalanced when it includes an odd number of negative relations, such as having one negative relation and two positives. This is a situation that would be psychologically uncomfortable, according to balance theory.
There are thought to be two types of relationship dynamics taking place in a triadic relationship:
-
Unit relationships – this is how much the different elements of the triangle belong together – the more similarities that exist, the more likely the psychological balance will occur.
-
Sentiment relationships – how one feels about something – are categorized as either ‘liking’ or ‘disliking.’
In most cases, if a positive unit relationship exists, a positive sentiment relationship will also exist. Likewise, a negative unit and negative sentiment relationship also tend to go together.
An example of balance theory
Heider (1958) used the following folk maxims to show how balance theory applies to everyday relationships.
‘My friend’s friend is my friend,
My friend’s enemy is my enemy,
My enemy’s friend is my enemy,
My enemy’s enemy is my friend.’
Balance theory does not only examine interpersonal relationships between three individuals. It can explore relationships between individuals and an object, activity, idea, or event.
For instance, take two individuals’ attitudes toward going to the gym. The examples below use George, Lily, and the gym as the three elements of the P-O-X triad.
- Balanced, shared liking: George likes Lily, and both like going to the gym (P+O, P+X, O+X), so the triad is comfortable and stable.
- Balanced, shared disliking: George likes Lily, but neither likes going to the gym (P+O, P-X, O-X), so the shared dislike keeps them in balance.
- Unbalanced, one disagreement: George likes Lily but dislikes the gym, while Lily likes it (P+O, P-X, O+X), so George may warm to the gym to restore balance.
- Unbalanced, shared liking with a disliked person: George dislikes Lily, yet both like the gym (P-O, P+X, O+X), so George may cool on the gym instead.
Across these examples, balance theory predicts that we tend to adopt similar attitudes to people we like and dissimilar attitudes to people we dislike.
What is the difference between balance theory and cognitive dissonance?
Cognitive dissonance is the mental discomfort that comes from holding two conflicting attitudes or beliefs.
For instance, taking the case of smoking cigarettes, two conflicting thoughts someone may have can be:
‘I enjoy smoking cigarettes.’
‘Smoking cigarettes is unhealthy.’
Cognitive dissonance theory suggests that people prefer cognitive consonance, where two attitudes, beliefs, or ideas are consistent with one another. While this seems similar to balance theory, they are defined in different ways.
Balance theory focuses on a triadic relation between the self, another person(s), and a third element.
Unlike cognitive dissonance theory, balance theory emphasizes inconsistencies between interpersonal relations, while cognitive dissonance can occur without any interpersonal inconsistencies.
Balance Theory in Network Science
Heider’s theory was built around three-person triads. Its biggest modern impact, though, has come from extending it to networks of any size.
The Graph-Theoretic Reformulation
Cartwright and Harary (1956) restated Heider’s theory using graph theory. People and relationships become points and lines. Each line carries a positive or negative sign.
A signed graph is balanced when every closed loop within it, called a cycle, has an even number of negative edges. For a three-point triad, this is exactly Heider’s rule.
From this, Cartwright and Harary proved a structure theorem. A balanced network can always be split into two subsets. Every positive tie sits within a subset, and every negative tie sits between the two.
In social terms, this is neat. A fully balanced network splits cleanly into two mutually hostile camps.
They also introduced a degree of balance: the proportion of a network’s cycles that are positive. This turns balance from an all-or-nothing verdict into a measurable quantity.
Researchers can now compare networks. One network can be more balanced than another, instead of only balanced or not.
Weak Balance
Cartwright and Harary’s rule treats three mutual enemies as unbalanced, since they cannot split into only two camps. Davis (1967) thought this was too strict.
His concept of weak balance allows a network to split into any number of hostile clusters, not just two. It is a looser standard.
It forbids only the pattern of two friends who share a common enemy. Three mutual enemies, by contrast, count as stable under weak balance. They simply form three separate factions instead of two.
Whether real networks show strong balance (Heider’s version) or weak balance (Davis’s version) is a live empirical question. It is where the field’s newest evidence has focused, and it remains genuinely unsettled.
Which version fits best is not a matter of taste. It depends on the statistical benchmark a researcher chooses, as the most recent test below shows.
Contemporary Research
Early large-scale studies were encouraging. Leskovec, Huttenlocher, and Kleinberg (2010) analysed millions of trust and distrust ties on sites such as Epinions, Slashdot, and Wikipedia.
They found balanced triads over-represented, just as the theory predicts. Facchetti, Iacono, and Altafini (2011) then measured whole-network balance directly.
Their conclusion was striking. Large online social networks sit extremely close to a perfect two-camp split.
Marvel, Kleinberg, Kleinberg, and Strogatz (2011) modelled how balance emerges over time. They built a mathematical model. Local tension-reduction, it showed, can drive a whole network toward two warring blocs.
They used the model to illuminate real history: the alliances that hardened before the First World War. Rawlings and Friedkin (2017) took a different approach.
They measured interpersonal tension directly, inside one real community. Structural balance appeared, but only under specific conditions, notably strong community commitment.
- Aim: Gallo, Garlaschelli, Lambiotte, Saracco, and Squartini (2024) tested whether real signed networks support strong balance (Heider’s version) or weak balance (Davis’s version).
- Method: They extended statistical network models to signed ties. Then they compared observed triad patterns against benchmarks that did or did not account for each person’s number of positive and negative ties.
- Results: Simple benchmarks favoured weak balance, but once individual differences in tie-counts were controlled for, the data favoured strong balance instead. Non-social networks showed no balance at all.
- Conclusion: Structural balance is a genuine feature of human social networks, and the long-running strong-versus-weak dispute partly reflects which statistical benchmark is used.
This study carries particular weight. It supplies the statistical benchmark earlier triad-counting studies lacked.
It tests only the pattern of ties, though, not the psychological tension-reduction process Heider originally described.
Strengths and Limitations of Balance Theory
Balance theory has aged well in some respects and shown clear weaknesses in others. Weighing both helps explain why the theory survived in one form while fading in another.
Strengths
Three qualities explain why the theory remains a fixture in social psychology.
- Elegance: A single sign-multiplication rule predicts which relationships feel comfortable. This simplicity is what let later researchers generalise it far beyond Heider’s original three-person triad.
- Foundational Status: It was the first cognitive-consistency theory, published in 1946. It directly shaped congruity theory, cognitive dissonance theory, and symbolic psycho-logic, the theories that followed it.
- Modern Support: Unlike most 1940s social-psychology theories, its graph-theoretic form has been tested successfully on real, large-scale social networks. Few ideas from that era have proved this testable at scale.
These three qualities are why balance theory outlived several other 1940s ideas. Most never made the jump from the psychology lab into modern network science at all.
Limitations
The theory also has real gaps, several of which the field has never fully closed.
- Ignores Strength: The theory treats every relation as simply positive or negative. It cannot represent a mildly positive tie versus a passionately positive one, so it misses differences in intensity.
- No Predicted Outcome: Balance theory says an unbalanced relationship will change, but not which relation will change. This makes it difficult to test against any single case.
- Positivity Bias: People tend to rate positive relations as pleasant almost regardless of balance. This positivity bias can outweigh the effect of balance itself, especially when the self is not involved.
- Superseded on Mechanism: For inconsistency within one person, cognitive dissonance theory offers more precise mechanisms and a far larger experimental literature. Attention shifted away from balance theory at the individual level, even as its structural form later flourished in network science.
Taken together, balance theory’s person-level influence has waned in favour of cognitive dissonance theory. Its structural core, however, has become one of the organising ideas of modern network science.
References
Heider, F. (1946). Attitudes and cognitive organization. The Journal of psychology, 21 (1), 107-112.
Heider, F. (1958). The psychology of interpersonal relations. Wiley. https://doi.org/10.1037/10628-000
Hummon, N. P., & Doreian, P. (2003). Some dynamics of social balance processes: Bringing Heider back into balance theory. Social Networks, 25(1), 17-49. https://doi.org/10.1016/S0378-8733(02)00019-9
