Psychologist logo
BPS updates, Community, Social and behavioural

Our emotional neighbourhoods

Peter Totterdell, Karen Niven and David Holman look at how social networks can regulate what we feel.

20 June 2010

We are all embedded within various kinds of social network, such as friendship groups, work-based associates, team-mates at a sports club, and local community contacts. These networks form our personal neighbourhood. But do these networks affect how we feel? Does our happiness depend partly on the happiness of those to whom we are connected? In this article, we argue that the answer is yes. We review evidence for the transfer of feelings between people and for the mechanisms that enable such transfer, and explore some of the complexities involved in this process. But first we start off with some background to social networks, for those unfamiliar with this research tradition.

Social networks

A social network comprises the set of relations (also known as ties) among a set of entities (also known as actors or nodes). These entities can be individuals, groups or organisations. Pairs of actors can be joined in a social network by a variety of relationships, including interaction ties, affective ties (e.g. likes or dislikes), role-based ties (e.g. manager–employee), and influence ties. Relations can differ in intensity (e.g. strong or weak ties) and can be bi-directional (out-ties stem from actors, in-ties are received by actors). An actor’s position within a social network also varies and can be characterised by a number of attributes, including degree centrality (number of ties that emanate from or go to an actor), density (extent to which the people to whom an actor has ties also have ties to one another), and structural similarity (the extent to which an actor has the same set of ties as other actors).

The recent growth of online social networking communities (e.g. Facebook, Habbo, LinkedIn) has enhanced popular awareness of social networks. But research on the topic has a century-long tradition (Scott, 2000). Somewhat surprisingly, research on social networks continues
to attract much stronger interest from disciplines such as sociology, anthropology and epidemiology than from psychology, but psychologists – especially those researching community and organisational matters (Borgatti & Foster, 2003) – have nevertheless contributed to its development. We sketch some of the methods used by social network researchers later in this article, but for present purposes the critical message is that the building block of a social network is a relation between two individuals.

Evidence for transfer of feelings

Feelings can include transient affective states, such as emotions (e.g. anger, fear) and moods (e.g. gloomy, calm), as well as more enduring states, such as affective well-being and happiness. If we are to accept that feelings can be transmitted through social networks then, given that the basic unit of a social network is a relation between two individuals, there ought to be evidence that feelings can be transmitted from one person to another. But we don’t need to turn to social network research for that evidence, because there is plenty already available from experimental and field studies of social dyads and groups. The clearest evidence comes from a study showing that moods shift towards that of the most expressive person when people sit facing each other without verbal communication (Friedman & Riggio, 1981). Other support comes from studies showing that depressed persons can induce negative feelings in others (Joiner, 1994), and from studies showing that moods are temporally reciprocated between partners in close relationships (Levenson & Gottman, 1983).

The next set of evidence comes from research on teams, which can be conceived as small social networks. Field research on work and sport teams has found that individuals’ moods change in synchrony with the collective mood of their teammates (Totterdell, 2000) and in particular with the mood of a team leader (Sy et al., 2005). Importantly, these effects have been found not to depend on the influence of shared external events.

A carefully designed laboratory study has also shown that a trained confederate can manipulate the moods of members of experimental groups without their knowledge simply by expressing specific feelings, causing the moods within the groups to become congruent with the ones transmitted (Barsade, 2002).  

People’s connections usually extend beyond a single team though, so with our colleagues we conducted a social network study of employees in a large organisation (Totterdell et al., 2004). This study established that: affect is more similar when there is a work tie between two employees; groups of employees who frequently interact with one another have distinctive affect profiles; and an employee’s affect can be predicted from the affect of everyone else in the network if it is weighted by the similarity of their structural position. Further support for the transfer of feelings across social networks emerged from a study of over 4000 people who were followed across a 20-year period (Fowler & Christakis, 2008). A key conclusion was that people’s happiness depends on the happiness of others with whom they are connected. The effect was weaker but still held even when there was only an indirect connection between people (i.e. when they were only linked via their connection with another person).

So there are some converging lines of evidence for transfer of feelings within social networks. However, it is possible to find effects that resemble transmission through a social network even where there are no plausible mechanisms for such transmission (Cohen-Cole & Fletcher, 2008), such as the transfer of acne from person to person. It is therefore important to establish that such mechanisms exist.

Mechanisms for transferring feelings

There appear to be two main types of mechanism by which feelings are transferred between people: automatic and inferential. The automatic mechanism, also known as primitive emotional contagion (Hatfield et al., 1994), is activated when individuals involuntarily mimic the expressions and postures of those with whom they interact. Thereafter afferent feedback pertaining to personal emotional state – for example from the facial muscles involved in smiling or frowning – automatically brings the individual’s affect into line with that of their interaction partner. Contagion of this kind has been found to occur even in the absence of face-to-face interaction (Neumann & Strack, 2000). In contrast, transfer by inferential means involves conscious top-down cognitive processing. Such processing includes an individual’s appraisal of the significance of an interaction partner’s emotional display and consequent self-regulation of emotions to produce the appropriate response (Van Kleef, 2009).

This account of how feelings are transferred between people fits well with recent findings from social neuroscience concerning the mechanisms of empathy (Decety & Jackson, 2006). The same neural architecture appears to be engaged when individuals produce emotional states in themselves as when they try to understand the emotions of others. This architecture involves automatic bottom-up processes (e.g. mimicry), top-down inferential processes (e.g. perspective-taking), and self-regulatory processes that enable the person to distinguish between their own and other people’s feelings (e.g. distancing).

Does transfer of feelings serve any function?

So if you have a communication tie to your neighbour – at home or work – and if that tie isn’t independent of other ties in the locale (for example, your neighbour has ties to other people), then the social network structure makes it is likely you will be more happy if your neighbourhood contains happy people. This logic can apply to other types of feeling too, for instance anxiety, sadness and excitement. Might this transfer of feelings serve any function?

Emotions and moods are known to have a wide range of effects on cognition and behaviour, and a dizzying array of theories have been proposed to account for those effects, but one effect deserving mention in this context is that synchrony of feelings among a group of people seems to enhance cooperation on tasks (Barsade, 2002). It is not yet resolved whether it is the valence of the group’s affect or the synchrony that is the critical factor, but it may be that spreading affect through a network via contagion can help mobilise coordinated action.

With respect to inferential processes, research on deliberate social sharing of emotion indicates that the impulse to share emotions is strong but does not seem to reduce the impact of the emotion and can even heighten it by reactivating memory of the associated events. Sharing emotion does, however, appear to strengthen social bonds and distribute knowledge about important events across the neighbourhood (Rimé, 2007). So the benefit of social sharing of emotion may primarily be that it helps maintain the social network itself.

The idea that emotions are best understood through their social interpersonal functions, rather than as private personal phenomenon, has gained support in recent years. For example, emotions have been defined as ‘ways of aligning and realigning interpersonal and intergroup relations’ (Parkinson et al., 2005, p.235). Similarly, the recent emotions-as-social-information (EASI) model explains that emotional expressions regulate social interaction by triggering inferences and automatic reactions in observers (Van Kleef, 2009). The reciprocal nature of this social influence process is captured in social interaction theory, which proposes that feedback from the observer impinges on the person expressing the emotion (Côté, 2005). 

Resistance is not futile

Lest it should seem from this account that the influence of the social network on people’s feelings is straightforward and inevitable, we should put the record straight and say that this is unlikely. There are many factors at play, and the dynamics are most probably complex. Moreover, the network will influence but not fully determine what people feel; and in general the sizes of the effects appear to be small to medium. People also develop ways of resisting emotional influence, such as using dark humour to blunt its impact.

In our own research, we have identified a wide range of strategies that people use to deliberately improve or worsen other people’s affect (Niven et al., 2009). People often use these strategies to align other people’s feelings with their own, but sometimes they use them to regulate people’s feelings away from their own. By conceptualising these affect regulation strategies as relational (i.e. as something that passes along a tie), we have begun to investigate what effects they have on well-being and relationships within social networks. This forms part of a larger ESRC-funded project examining the neurophysiological and social processes that govern the emotion regulation of others and self (EROS). 

Another factor that has to be taken into account is the structure of the social network. How central people are within a network can influence their feelings.

We found that employees with many work ties were more enthusiastic but also more anxious, perhaps because of the effort required to sustain those ties (Totterdell et al., 2004). Equally employees with dense networks were less anxious, probably because of the availability of social support. Such findings imply that the formation and breaking of ties will have an influence on affect that reverberates through the social network.

Propagation of affect through the network also depends on characteristics of the individual carriers. Extant research indicates that individuals who score high on scales measuring affective communication (also known as charisma) or extraversion or who are perceived by others as displaying positive affect (Totterdell et al., 2008) or energy (Cross et al., 2002) are not only more likely to have a greater influence on how others feel but are also more likely to become more central within their social networks. People even seem to prefer individuals who they like over those who are competent at their job when it comes to choosing who they seek advice from (Casciaro & Lobo, 2005). Equally, individuals who readily make connections with others – social networkers – will exert greater influence on the affect of their social network (Totterdell et al., 2008). Some individuals have also been identified as fulfilling a role within social networks in which they mitigate the toxic emotions of others (Frost & Robinson, 1999). So the affective dynamics of the social network will depend on the makeup of the individuals within it.

A virtual unknown

Earlier we noted the increasing popularity of online social networking communities. It is not yet known, however, whether the phenomena we have spoken of here apply in the same way when the network is online and the communication ties are virtual. Transfer of feelings is not restricted to a face-to-face modality, but research on virtual interaction indicates that differences in online social cues and lack of local understanding can lead to greater disinhibition and misunderstandings and make it harder to establish strong relationships online (Cramton et al., 2007). Research has also not yet established whether feelings are compartmentalised into online and offline networks, or whether they transfer from one to the other. In this and related issues concerning affect in social networks, there exists great potential for psychologists to contribute to what has been termed the ‘new’ science of networks (Watts, 2004).

Peter Totterdell is at the University of Sheffield
[email protected]

Karen Niven is at the University of Sheffield
[email protected]

David Holman is at the University of Sheffield
[email protected]

Box - Collecting social network data

Social network research offers many possibilities for investigating fundamental and applied research questions about social phenomena, such as how psychological characteristics shape and respond to changes in social structure. There are a number of methods that researchers can use for collecting social network data, such as the roster method in which respondents are presented with the names of all people in the relevant network and asked to identify those with whom they have ties. The resulting data is usually coded in matrix format that locates a different actor in each row and the actors with whom they could have ties in each column.

The data requires special analysis procedures, such as quadratic assignment procedure (QAP), which is essentially regression for matrix data. Fortunately many of these procedures can be found in UCINET (Borgatti et al., 2002), which is a widely used statistical package for social network analysis (the social network researcher’s equivalent of SPSS). An advantage of social network data is that it includes not only the actor’s view of his or her relationships with others, but also includes the other person’s view of the same relationship. A disadvantage is that it commonly requires people to record judgements about specific others, which can raise privacy concerns.


Barsade, S.G. (2002). The ripple effect: Emotional contagion and its influence on group behavior. Administrative Science Quarterly, 47, 644–675.
Borgatti, S.P., Everett, M.G. & Freeman, L.C. (2002). UCINET 6 for Windows: Software for social network analysis. Harvard, MA: Analytic Technologies.
Borgatti, S.P. & Foster, P.C. (2003). The network paradigm in organizational research: A review and typology. Journal of Management, 29, 991–1013.
Casciaro, T. & Lobo, M.S. (2005). Competent jerks, lovable fools, and the formation of social networks. Harvard Business Review, 83, 92–100.
Cohen-Cole, E. & Fletcher, J.M. (2008). Detecting implausible network effects in acne, height, and headaches: Longitudinal analysis. British Medical Journal, 337, a2533.
Côté, S. (2005). A social interaction model of the effects of emotion regulation on work strain. Academy of Management Review, 30, 509–530.
Cramton, C.D., Orvis, K.L. & Wilson, J.M. (2007). Situation invisibility and attribution in distributed collaborations. Journal of Management, 33, 525–546.
Cross, R., Baker, W. & Parker, A. (2002). Charged up: The creation and depletion of energy in social networks. Cambridge, MA: IBM Institute for Knowledge-Based Organizations.
Decety, J. & Jackson, P.L. (2006). A social-neuroscience perspective on empathy. Current Directions in Psychological Science, 15, 54–58.
Fowler, J. & Christakis, N.A. (2008). The dynamic spread of happiness in a large social network: Longitudinal analysis over 20 years in the Framingham heart study. British Medical Journal, 337: a2338.
Friedman, H.S. & Riggio, R.E. (1981). Effect of individual differences in nonverbal expressiveness on transmission of emotion. Journal of Nonverbal Behavior, 6, 96–104.
Frost, P. & Robinson, S.L. (1999). The toxic handler: Organizational hero and casualty. Harvard Business Review, 77, 96–106.
Hatfield, E., Cacioppo, J.T. & Rapson, R.L. (1994). Emotional contagion. Cambridge: Cambridge University Press.
Joiner, T.E., Jr. (1994). Contagious depression: Existence, specificity to depressed symptoms, and the role of reassurance seeking. Journal of Personality and Social Psychology, 67, 287–296.
Levenson, R.W. & Gottman, J.M. (1983). Marital interaction: Physiological linkage and affective exchange. Journal of Personality and Social Psychology, 45, 587–597.
Neumann, R. & Strack, F. (2000). ‘Mood contagion’: The automatic transfer of mood between persons. Journal of Personality & Social Psychology, 79, 211–223.
Niven, K., Totterdell, P. & Holman, D. (2009). A classification of controlled interpersonal affect regulation strategies. Emotion, 9, 498–509.
Parkinson, B., Fischer, A. & Manstead, A.S.R. (2005). Emotion in social relations: Cultural, group, and interpersonal processes. Philadelphia, PA: Psychology Press.
Rimé, B. (2007). Interpersonal emotion regulation. In J. Gross (Ed.) Handbook of emotion regulation (pp.466–485). New York: Guilford.
Scott, J. (2000). Social network analysis: A handbook (2nd edn). London: Sage.
Sy, T., Côté, S. & Saavedra, R. (2005). The contagious leader: Impact of the leader’s mood on the mood of group members, group affective tone, and group processes. Journal of Applied Psychology, 90, 295–305.
Totterdell, P. (2000). Catching moods and hitting runs: Mood linkage and subjective performance in professional sport teams. Journal of Applied Psychology, 85, 848–859.
Totterdell, P., Holman, D. & Hukin, A. (2008). Social networkers: Measuring and examining individual differences in propensity to connect with others. Social Networks, 30, 283–296.
Totterdell, P., Wall, T., Holman, H. et al. (2004). Affect networks: A structural analysis of the relationship between work ties and job-related affect. Journal of Applied Psychology, 89, 854–867.
Van Kleef, G.A. (2009). How emotions regulate social life: The emotions as social information (EASI) model. Current Directions in Psychological Science, 18, 184–188.
Watts, D.J. (2004). The ‘new’ science of networks. Annual Review of Sociology, 30, 243–270.