Affect Networks: A Structural Analysis of the Relationship Between Work Ties and Job-Related Affect Peter Totterdell, Toby Wall, David Holman, and Holly Diamond University of Sheffield Olga Epitropaki Athens Laboratory of Business Administration The relationship between organizational networks and employees’ affect was examined in 2 organiza- tions. In Study 1, social network analysis of work ties and job-related affect for 259 employees showed that affect converged within work interaction groups. Similarity of affect between employees depended on the presence of work ties and structural equivalence. Affect was also related to the size and density of employees’ work networks. Study 2 used a 10-week diary study of 31 employees to examine a merger of 2 organizational divisions and found that negative changes in employees’ affect were related to having fewer cross-divisional ties and to experiencing greater reductions in network density. The findings suggest that affect permeates through and is shaped by organizational networks. Do employees’ feelings depend on the network of people with whom they work? Increasing recognition of the importance of affect in organizational life has, according to Barsade, Brief, and Spataro (2003), spawned an “affective revolution,” which is lead- ing to the emergence of a new research paradigm in organizational behavior. In this paradigm, affect is understood to have an influ- ential role in a wide range of organizational processes. Part of the affective revolution involves a movement away from understand- ing feelings only as personal experiences and toward studying how people socially share and influence each other’s affect at work and how this affects work life. In this investigation, we contribute to this movement by examining how employees’ feelings are related to the structural relationships present in organizational networks. There is already evidence that employees’ feelings are influ- enced by the structural relationships present in work teams. Recent studies, for example, have demonstrated that individuals’ moods are congruently related to the ongoing collective feelings of their work team, independent of shared events (Totterdell, 2000; Tot- terdell, Kellett, Teuchmann, & Briner, 1998). The transfer of moods and emotions within work teams is thought to be brought about by a variety of implicit and explicit processes, such as primitive emotional contagion and expression management (Kelly & Barsade, 2001). In most workplaces, however, employees are embedded within organizational networks that extend beyond the confines of their immediate work group. For example, employees usually have work ties that they do not share with other teammates and they are normally indirectly linked to other nonteam members via their teammates’ work ties. It is therefore credible to imagine that affect can permeate through the whole of an organization via the total network of ties among people. Consequently, in the two studies described here, we adopted a social network perspective to examine whether the arrangement of ties among employees in organizational networks was related to their affect. A social network perspective is concerned with the structure and patterning of relationships between people, groups, or organiza- tions (Tichy, Tushman, & Fombrun, 1979). In the last few decades, increased interest in organizational networks has created a sub- stantial body of social network research concerning both intra- and interorganizational networks (Ahuja, 2000; Fombrun, 1982; Ibarra, 1993; Lincoln & Miller, 1979; Mehra, Kilduff, & Brass, 2001). Research on intraorganizational networks views employees as objects embedded in social networks. The structural properties of these social networks are typically used to explain outcomes within the organization, such as employees’ influence, perfor- mance, or innovation behavior (Brass, 1984; Sparrowe, Liden, Wayne, & Kraimer, 2001; Tushman, 1977). Relating demonstrable structural properties of organizations to perceived outcomes has the added advantage that it avoids the problem of using percept- to-percept relations that besets much organizational research. Al- though some research on social networks has focused on the subjective reactions of employees, it has not addressed the issue of employee affect per se. Affect refers to a range of feeling states, including different moods and emotions. Concerted efforts to understand the role of affect at work are relatively recent, but there is a longer history of research concerning job-related feelings in relation to outcomes such as job satisfaction and strain (Briner & Totterdell, 2002; Warr, 2002). Many studies that have investigated the structure of emotions, moods, and affective well-being, both in and out of work settings, have converged on a circumplex model in which different feeling states are arranged around the circumference of a circle (e.g., Remington, Fabrigar, & Visser, 2000; Russell, 1980; Warr, 2002; Watson & Tellegen, 1985). The principal axes of the circle Peter Totterdell, Toby Wall, David Holman, and Holly Diamond, Insti- tute of Work Psychology, University of Sheffield, Sheffield, England; Olga Epitropaki, Athens Laboratory of Business Administration, Athens, Greece. We gratefully acknowledge the support of the Economic and Social Research Council (ESRC) UK. The work was part of the program of the ESRC Centre for Organisation and Innovation. Correspondence concerning this article should be addressed to Peter Totterdell, Institute of Work Psychology, University of Sheffield, Sheffield S10 2TN, England. E-mail: p.totterdell@sheffield.ac.uk Journal of Applied Psychology Copyright 2004 by the American Psychological Association 2004, Vol. 89, No. 5, 854 – 867 0021-9010/04/$12.00 DOI: 10.1037/0021-9010.89.5.854 854