Powerpoint

2020-87078-001.pdf

Home >Business & Finance homework help >Management homework help >Powerpoint

Original Article

How Communication in Organizations Is Related to Foci of Commitment and Turnover Intentions Sebastian Holzwarth, George Gunnesch-Luca, Roman Soucek, and Klaus Moser

School of Business, Economics, and Society, University of Erlangen-Nürnberg, Germany

Abstract. The current study analyzes how two components of perceived organizational communication (vertical and horizontal) are related to employee turnover intentions via three types of affective commitment foci (organization, supervisor, and team). Using second-order confir- matory factor analysis and structural equation modeling techniques with a large cross-sectional dataset (n = 3,317), our results show that, in line with social cohesion theory, vertical communication (e.g., supportiveness from the organization) is strongly related to affective organizational commitment, whereas horizontal communication (e.g., supportiveness from colleagues) is primarily related to affective team commitment. Additionally, both communication dimensions are related to affective supervisory commitment. Finally, these three foci of affective com- mitment incrementally explain and differentially mediate the relationship between perceived organizational communication and turnover intention.

Keywords: organizational communication, vertical communication, horizontal communication, affective commitment, turnover intentions

A famous scholar in the organizational sciences, Luhmann (1984), stated that communication has a constitutive function for any organization. In fact, communication in organizations (organizational communication) has various important functions. For example, the coordination of members’ activities in organizations is a vital task, which is done through communication, such as written rules. Two further important clusters of important activities are the socialization of individuals toward understanding job tasks and the organizational culture, as well as the im- plementation of organizational objectives and strategy. Clearly, they can only be put into effect through com- munication. As another example, leadership is defined as influencing others by means of communication. Finally, organizations usually comprise groups, and communica- tion in groups is essential, for example, to ensure that group members follow common norms and goals. In sum, a functioning organization has a continuous stream of communication between its members, and communica- tion in organizations appears to be really important, if not necessary, for any organization. We use the term “organizational communication” in

order to describe communication processes within or- ganizations between employees who are related to

organizational membership issues. For example, talking about family, leisure activities, or politics at work would not belong to organizational communication. We consider organizational communication as being a descriptive term, which also includes that it might have more or less quality. For example, more accurate information should be a part of a high score on an organizational communication measure. However, although satisfaction with organiza- tional communication might be a variable worth investi- gating, it differs from organizational communication as a descriptive phenomenon. Finally, we consider organiza- tional communication as being an individual level vari- able, i.e., it describes communication contents and processes as enacted and perceived by the employees in the organization. This is why we prefer to not use the term “communication climate,” which suggests not only an additional affective component of communication but also that employees report how communication in general in the organization occurs. (Unfortunately, the terms “or- ganizational communication” and “communication cli- mate” were sometimes used interchangeably in previous research; cf. Bartels et al., 2010.) One interpretation of the assumption of a “constitutive

function” of organizational communication is that it is of

Journal of Personnel Psychology (2021), 20(1), 27–38 https://doi.org/10.1027/1866-5888/a000261

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

utmost importance. In fact, organizations are based on the coordination of cooperative behaviors, and communica- tion is the “tool” to achieve this. If it is even “constitutive” and thus an essential feature, organizations without communication could not exist. A second interpretation might be that organizations without communication ac- tivities might experience a lack of cooperation. For this reason, they will be faced with adversities and finally will not survive. One early sign of troubles in organizations would be strong withdrawal intentions of the organiza- tion’s members, that is, high turnover intentions. A third interpretation is that employees learn that the continuation of organizational membership seems to lose its rewarding value. They might assume that their career opportunities diminish, job insecurity increases, or their performance is not appropriately acknowledged. These three interpreta- tions have in common that a lack of communication ac- tivities might be detrimental for organizations. In the following, we want to add another interpretation to the assumption that communication is important: Organiza- tional communication strengthens affective commitment in employees, which in turn is negatively related to withdrawal.

Our paper is organized as follows: First, we will in- troduce the theory of relational cohesion (Lawler & Yoon, 1996), which predicts that repeated cooperation between entities creates an affective relationship, which in turn reduces withdrawal. We will extend this notion in such a way that we assume that, in the organizational context, the creation of such a cooperative relationship means the triggering of affective commitment. Coop- eration is organized through communication, and be- cause cooperation can happen between actors on the same hierarchical level (i.e., between colleagues) or actors from different hierarchical levels (i.e., between supervisor and employee), we introduce the distinction between horizontal and vertical communication and show how they are related to three types of affective commitment. These distinctions are important because (a) the various types of affective commitment are in- crementally related to turnover intentions and (b) be- cause the two types of organizational communication contribute to a reduction of turnover intentions via different mechanisms. In a nutshell, we assume that organizational communication has distinct relationships to different foci of affective commitment, which in turn are related to turnover intentions. The main contribu- tion of the current research is to demonstrate that dif- ferent forms of organizational communication matter because they contribute to the functioning of organi- zations (measured as low turnover intentions) via the strengthening of different forms of affective commit- ment in employees.

Cooperation, Organizational Communication, and Affective Commitment

Cooperation is not only a defining feature of organiza- tions, but it also has the potential to create bonds between individuals. For example, continuing reciprocity, which is an important part of cooperation, has long been known as a source for strengthening bonds between the persons involved (Gouldner, 1960). More recently, the theory of relational cohesion was introduced (Lawler & Yoon, 1996), which assumes that frequent exchanges be- tween actors create relational cohesion and finally be- havioral commitment. The actors involved can be individuals or other social entities, for example, work groups or organizations. Relational cohesion results from perceiving a relationship between the actors as “. . . an expressive object valuable in its own right” (Lawler & Yoon, 1996, p. 89). This relational cohesion is also de- scribed as “affective attachment” and can be called “affective commitment.” If the actors are individuals and the organization, relational cohesion can be called “af- fective organizational commitment”; if the actors are individuals and their work groups, it is “affective team commitment”; and if the actors are employees and their supervisor, it can be called “supervisory commitment.” According to the theory of relational cohesion, the final outcome of repeated cooperation is “behavioral com- mitment” (Lawler & Yoon, 1996), that is, the continua- tion of the relationship and thus the forgoing of alternative options, or in the organizational context, the abstaining from withdrawal.

Whereas the theory of relational cohesion assumes that the core of a cooperation relationship is frequent ex- changes, it is hard to imagine that cooperation is feasible without communication. In other words, communication enabling and constituting cooperation, i.e., organizational communication, should exert a comparable effect on af- fective commitment and withdrawal intentions. The the- ory of relational cohesion (Lawler & Yoon, 1996) can therefore be adapted as follows: Organizational commu- nication creates affective commitment, which in turn re- duces withdrawal intentions.

Two Dimensions of Organizational Communication

Based on the assumption that organizations have a hierarchical structure and that the mean size of the

28 S. Holzwarth et al., Communication in Organizations

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

control span of supervisors is bigger than 1, the primary challenge for any organization is to maintain the functioning of the hierarchy, the elements being the hierarchical levels, and the work groups, the elements here being the work group members. For this to func- tion, organizational communication is needed. In other words, organizational communication comprises two directions of communication, horizontal communica- tion and vertical communication (Bartels et al., 2010; Smidts et al., 2001). More generally, cooperation in organizations is the process of working together for a common benefit, either on the same level or in a hi- erarchical relationship. The respective communication activities are horizontal and vertical communication. Horizontal communication occurs between employees on the same hierarchical level, whereas vertical com- munication travels top-down or bottom-up within the organizational hierarchy. The distinction of two types of organizational com-

munication leads to the question whether it is appro- priate to stick to the general assumption of the organizational communication creates affective com- mitment hypothesis. For example, whereas frequent communication in a work group (= horizontal com- munication) might strengthen the affective bonds in the team, meaning affective team commitment of the team members, it is less clear whether horizontal communication is also related to commitment to the organization. As another example, vertical communi- cation might contribute to affective commitment to the organization although it is less clear how it should be related to affective commitment to other team mem- bers. In fact, affective commitment is more than a general affective state in individuals, it has a focus, i.e., organizational, work group, and supervisory affective commitment can be distinguished. More specifically and consistent with the target similarity model (Lavelle et al., 2007), a fit between target of communication and target of commitment is to be expected. First, there should be a particularly strong relationship between vertical communication and affective organizational commitment.

Hypothesis 1 (H1): Vertical communication in the or- ganization is related to affective organizational commitment.

Second, also consistent with the target similarity model (Lavelle et al., 2007), horizontal communication should be primarily related to team commitment. This assumption is consistent with the results of studies in small-group re- search that have shown that communication with proxi- mate others, a component of horizontal communication,

increases attachment to and cohesion with other team members (Levine & Moreland, 2012).

Hypothesis 2 (H2): Horizontal communication in the organization is related to affective team commitment.

Finally, supervisory affective commitment, as a form of attachment to a target somewhere in between the distal organization and proximate colleagues, should be affected by both horizontal communication and vertical communication. On the one hand, the supervisor rep- resents the organization and explains policies or de- cides on the distribution of rewards. On the other hand, the supervisor often closely interacts with team members and might sometimes even be a primus inter pares. In fact, the supervisor effect hypothesis has been put forward assuming that the supervisor is even the primary target of commitment in organizations (Askew et al., 2013). In sum, we consider it useful to include the supervisor as a third target of commitment and hypothesize:

Hypothesis 3 (H3): Both vertical communication and horizontal communication are related to affective supervisory commitment.

Three Affective Commitment Foci and Turnover Intentions

Previous research has found that affective organizational commitment has a strong negative relationship with turnover intentions (Meyer et al., 2002). Comparable ef- fects have been found concerning the relationship be- tween affective team commitment and turnover intentions (for example, Felfe & Yan, 2009). Finally, similar effects were found concerning the relationship between affective supervisory commitment and turnover intentions. For example, Stinglhamber and Vandenberghe (2003) found that affective commitment to the supervisor was related to actual turnover, whereas Vandenberghe and Bentein (2009) found affective commitment to the supervisor to exert negative effects on turnover intentions. In sum, all three foci of affective commitment are related to turnover intentions. Moreover, the differentiation between the three affective commitment foci is promising because it might be a way to improve the prediction of turnover intentions by combining the three affective commitment foci. In fact, organizational, team, and supervisory com- mitment incrementally predict turnover intentions (Askew et al., 2013; Wombacher & Felfe, 2017).

S. Holzwarth et al., Communication in Organizations 29

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Hypothesis 4 (H4): Affective organizational, team, and supervisory commitments are incrementally related to turnover intentions.

The incremental effects of the three foci of affective commitment on turnover intentions are important because of the relevance of the outcome variable for any organi- zation. However, their distinction might be even more important if it can help to explain why the distinction of two types of organizational communication is justified. More specifically, we expect that the effects of the two dimensions of organizational communication on turnover intentions are differentially mediated by affective orga- nizational, supervisory, and team commitment.

Hypothesis 5 (H5): The negative effects of vertical and horizontal communication on turnover intentions are mediated by affective organizational, team, and su- pervisory commitment, such that horizontal com- munication is mediated by affective team and supervisory commitment, while vertical communi- cation is mediated by affective organizational and supervisory commitment.

The five hypotheses postulated above can be summa- rized in a process model which describes how we assume that communication in organizations exerts its “consti- tutive function” (see Figure 1, model H).

Method

Participants

Bilingual cross-sectional data from n = 3,317 employees in 54 locations of a multinational industrial company from the automotive sector were collected with an online survey in German and English. During the time the data were gathered, the organization had around 20,000 employees. Therefore, the final response rate was around 17%. Overall, 52% of the respondents answered in German and 48% in English. Considerably more men (87%) than women (13%) answered the questionnaire, which ap- proximately reflects the worldwide female:male ratio of the workforce in that organization. With regard to age, 26% of the participants were between 26 and 35 years old and 35% were between 36 and 50 years old. The average length of work experience was 17.50 years (SD = 11.14), 70% of the respondents had a university-level education with a degree, and 62% were working in an engineering or technical area. Twenty-seven percent of the respondents had managerial responsibilities with at least one direct report. Note that these results are based on the dataset that resulted after several data screening steps. For example, respondents who did not complete at least 80% of the survey or who finished the survey “too quickly” or “too slowly” (i.e., respondents who spent an extremely short or long period of time filling out the questionnaire) were excluded. Due to the considerable sample size, no missing data imputation techniques were used.

Analytic Procedure

We followed a three-step approach. First, with the means of confirmatory factor analysis (CFA), competing mea- surement models for organizational communication and commitment were tested and compared. The main question addressed here was whether the data justify separate measurements of horizontal and vertical com- munication, as well as organizational, supervisor, and team commitment. Second, we were interested in the differential relationships between horizontal and vertical communication and the three foci of affective commit- ment, as well as their contributions to turnover intentions. These issues were addressed by means of structural equation modeling (SEM). Third, we investigated whether the model can give the best explanation for the relation- ships. To address this question, the hypothesized model (H) was modified and the fit of different alternative models was tested. Figure 1 presents an overview of the models that were considered. First, model H was examined. As a next step, the hypothesized model was modified by testing

Figure 1. Hypothesized and alternative relationships between or- ganizational communication, affective commitment, and turnover intentions. HC = horizontal communication; OC = affective orga- nizational commitment; SC = affective supervisory commitment; TC = affective team commitment; TI = turnover intentions; VC = vertical communication.

30 S. Holzwarth et al., Communication in Organizations

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

its exact opposite. Here, vertical communication was connected with supervisory and team commitment, and horizontal communication was connected with supervi- sory and organizational commitment only (alternative model B1). In alternative model 1 (A1), horizontal com- munication and vertical communication were connected with all three commitment foci. In alternative model 2 (A2), the path between vertical communication and team commitment was eliminated, and in alternative model 3 (A3), a model in which horizontal communication is not connected to organizational commitment was investi- gated. Because the main research interest was to compare the impact of two forms of communication on three af- fective commitment foci in parallel, stable, and compa- rable conditions for the models tested needed to be created. Accordingly, in the five structural models inves- tigated, the same relationship between commitment and turnover intentions was assumed, which means that or- ganizational, supervisory, and team commitment all re- duced turnover intentions. First- and second-order CFA and SEM were conducted

in R using the lavaan package (Rosseel, 2012). As rec- ommended by Bandalos (2014), the robust weighted least squares algorithm for estimation was used because the data did not meet the assumption of multivariate normality, nor was it continuous, and the sample size was large. Root mean square error of approximation (RMSEA) as an absolute measure of fit and standardized root mean square residual (SRMR) are reported to assess model fit; 90% confidence intervals (CIs) for RMSEA are also reported. As incremental measures of fit, the robust versions of the Tucker–Lewis index (TLI) and the

comparative fit index (CFI) as well as RMSEA were used (Savalei, 2018).

Measurement Instruments

Vertical communication was assessed by means of 12 items divided into four dimensions (Bartels et al., 2010). The four dimensions measuring vertical communication are

1. participation in decision-making; 2. trust, confidence, and credibility of top management; 3. information adequacy; and 4. supportiveness from the organization.

Sample items from these four dimensions are “My opinions are relevant in day-to-day decisions affecting my job”; “The information I receive from top management is credible”; “It is easy to obtain information that I need to accomplish my job”; and “Suggestions I submit to man- agement are seriously considered.” Although the primary interest was in vertical communication as a summary construct, we were also interested in the construct validity of the respective measure. Therefore, a CFA was com- puted, which showed that a four-dimensional model of vertical communication fits the data (see Table 1: χ2diff = 2,025.77, dfdiff = 6) better than the one-dimensional model. Horizontal communication was assessed by using nine

items divided into three dimensions: (1) level of contact with colleagues, (2) openness and candor, and (3) sup- portiveness from colleagues. Sample items from these three dimensions are “My colleagues and I regularly talk to each other about our daily work”; “In general, people in

Table 1. Fit comparisons between competing measurement models

CFA model Latent

variables χ2 df χ2/df CFI TLI SRMR RMSEA 90% CI RMSEA

Baseline RMSEA Δχ2 Δdf

Affective commitment, one-dimensional

1 6,704.41 90 74.49 0.915 0.901 0.109 0.102 .100–.104 0.323

Affective commitment, three-dimensional

3 1,345.49 87 15.47 0.986 0.983 0.044 0.042 .040–.044 0.323 5,358.92* 3

VC, one-dimensional 1 2,957.19 54 54.76 0.975 0.969 0.071 0.067 .065–.070 0.384

VC, four-dimensional 4 931.42 48 19.40 0.993 0.991 0.037 0.036 .034–.038 0.384 2,025.77* 6

HC, one-dimensional 1 1,291.48 27 47.83 0.974 0.965 0.066 0.061 .058–.063 0.323

HC, three-dimensional 3 188.88 24 7.87 0.997 0.995 0.023 0.022 .019–.025 0.323 1,102.60* 3

Communication (VC and HC), one-dimensional

1 16,539.39 189 87.51 0.885 0.872 0.123 0.105 .103–.106 0.292

Communication (VC and HC), two-dimensional

2 5,455.23 188 29.01 0.963 0.959 0.068 0.059 .058–.061 0.292 11,084.17* 1

Communication (VC and HC), second-order model

4 + 3 and 2 3,212.23 181 39.65 0.980 0.977 0.050 0.045 .043–.046 0.292 2,243* 7

Note. CFA = confirmatory factor analysis; CFI = comparative fit index; HC = horizontal communication; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; TLI = Tucker–Lewis index; VC = vertical communication. *p < .001.

S. Holzwarth et al., Communication in Organizations 31

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

my department are honest with each other”; and “When I talk to my direct colleagues, I feel taken seriously.” Again, the primary interest was in a global measure of horizontal communication, although the corroboration of the con- struct validity of the respective scale was also searched for. Again, a three-dimensional model fits the data (see Table 1: χ2diff = 1,102.60, dfdiff = 3) better than a one- dimensional model.

Vertical communication and horizontal communication were conceptualized as two higher-order constructs, consisting of four and three dimensions, respectively. A higher-order construct is a multidimensional latent model in which the dimensions serve as indicators of the con- struct. Higher-order models are applicable when the lower-order factors are substantially correlated with each other, and there exists a higher-order factor that is hy- pothesized to account for the relationship among the lower-order factors. In fact, a chi-squared difference test indicated a significant difference between the two-factor, first-order communications model and an alternative model with seven first-order and two second-order factors (see Table 1: χ2diff = 2,243, dfdiff = 7). In addition, incre- mental and absolute-fit indices all indicated an excellent fit for the second-order organizational communication model. In sum, the two second-order factors represent organizational communication. However, because the 2 one-dimensional models of horizontal and vertical com- munication still had acceptable fit indices, and because no further assumptions about differential relationships of the subdimensions with the other outcome variables were made, we decided to use the summary score of horizontal and vertical organizational communication.

The measurement of affective organizational, supervi- sory, and team commitment was conducted with scales develop by Allen and Meyer (1990), as well as a German adaption of these scales (Felfe & Franke, 2012). Each commitment focus was measured with five items. Sample items are “I would be very happy to spend the rest of my working life in this organization”; “My supervisor is very important to me personally”; and “I’m proud to be a member of this team.” First-order CFA reveals that a three-dimensional model better fits the data than a one- dimensional model (see Table 1: χ2diff = 5,358.92, dfdiff = 3). Absolute (RMSEA and SRMR) and incremental (CFI and TLI) fit measures of the three-dimensional model were satisfactory (see Table 1; Feinian et al., 2008). In sum, affective commitment toward the organization, the su- pervisor, and the team are three distinct constructs.

Finally, respondents’ turnover intentions were measured with three items adapted from the Michigan Organizational Assessment Questionnaire (Cammann et al., 1983). A sample item is “I often think about leaving the company.” All items were answered on 5-point Likert scales.

Results

Means, SDs, correlations, and reliability coefficients of the study variables are reported in Table 2. The reliabilities of all measures are good (.74 ≤ α ≤ .89).

Test and Modification of the Structural Model

Using SEM, the tenability of the core model (H) and modifications were tested. Table 3 shows the results for seven alternative models. The mediation framework (Hypothesis 5) was tested with a SEM approach. Table 3 indicates that model H, in which vertical communi- cation has a direct effect on organizational and su- pervisory commitment, horizontal communication has a direct effect on supervisory and team commitment, and all three commitment foci have direct effects on turnover intentions, fits the data. In order to test the first four hypotheses, the relationships between the three types of affective commitment and turnover in- tentions were not varied. This was driven by the need to decrease model complexity and to create stable condi- tions for investigating the communication–commitment relationships.

As expected, for alternative model B1, in which ver- tical communication has a direct effect on supervisory and team commitment and horizontal communication has a direct effect on organizational and supervisory commitment, all measures of fit were worse. Model H can be further modified by adding two direct relation- ships between vertical communication and team com- mitment as well as horizontal communication and organizational commitment. The resulting alternative model A1 fits the data well. In order to further examine which relationship is exactly responsible for the sig- nificant increase in model fit, the path between vertical communication and team commitment (alternative model A2) and the path between horizontal communi- cation and organizational commitment (alternative model A3) were eliminated. Comparing model A2 to model A1 shows a significant decrease in all measures of fit (Δχ2 (1) = 1,382.99, p < .001), whereas model A3 shows measures of fit for CFI, TLI, SRMR, and RMSEA com- parable to model A1. Thus, the more parsimonious al- ternative model A3 is the best model so far (see also Figure 2).

Figure 2 displays the relationships and standardized parameter values for model A3. Vertical communication shows the strongest impact on organizational commitment (β = .74, p < .001), which confirms Hypothesis 1. However, an effect of vertical communication on team commitment

32 S. Holzwarth et al., Communication in Organizations

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

(β = .41, p < .001) was also found. With regard to horizontal communication, a significant effect on team commitment was obtained, which is in line with Hypothesis 2 (β = .38, p < .001). Finally, both vertical communication and hor- izontal communication have strong effects on affective

supervisory commitment (β = .58 and β = .16, ps < .001), which confirms Hypothesis 3. Organizational commitment had the strongest negative

effect on turnover intentions (β = �.48, p < .001) by far, followed by team (β = �.11, p < .01) and supervisory

Table 3. Fit comparisons between competing structural models

Structural model

Latent variables χ2 df χ2/df CFI TLI SRMR RMSEA

90% CI RMSEA

Baseline RMSEA Δχ2 Δdf

Hypothesized model H

4 + 3 + 3 + 1 and 2

9,640.99 687 14.03 .969 .967 .056 .048 .047–.049 .264

Alternative model B1

4 + 3 + 3 + 1 and 2

14,203.63 687 20.67 .953 .949 .070 .060 .059–.061 .264

Alternative model A1

4 + 3 + 3 + 1 and 2

8,107.41 685 11.83 .975 .973 .050 .043 .042–.044 .264 1,533.58* [H vs. A1]

2 [H vs. A1]

Alternative model A2

4 + 3 + 3 + 1 and 2

9,490.40 686 11.81 .970 .968 .056 .047 .046–.048 .264 1,382.99* [A2 vs. A1]

1 [A2 vs. A1]

Alternative model A3

4 + 3 + 3 + 1 and 2

7,971.58 686 11.62 .975 .973 .050 .043 .042–.044 .264 135.83* [A3 vs. A1]

1 [A3 vs. A1]

Alternative model C1

4 + 3 + 3 + 1 and 2

9,759.10 685 14.24 .970 .967 .056 .048 .047–.049 .264

Alternative model C2

4 + 3 + 3 + 1 and 2

8,234.85 683 12.05 .975 .973 .050 .043 .042–.044 .264 1,524.25* [C2 vs. C1]

2 [C2 vs. C1]

Alternative model C3

4 + 3 + 3 + 1 and 2

8,108.53 684 11.85 .975 .973 .050 .043 .042–.044 .264 1,650.57* [C3 vs. C1]

1 [C3 vs. C1]

Note. CFI = comparative fit index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; TLI = Tucker–Lewis index. *p < .001.

Table 2. Means, standard deviations, and correlations of study variables

n M SD 1 2 3 4 5 6 7 8 9 10 11 12 13

1. Affective organizational commitment

3,072 3.47 .72 (.84)

2. Affective supervisory commitment

2,926 3.26 .76 .46* (.86)

3. Affective team commitment

2,817 3.32 .72 .50* .48* (.85)

4. Vertical communication 3,314 3.32 .63 .59* .52* .44* (.91)

5. Participation in decision- making

3,300 3.37 .76 .48* .48* .42* .83* (.74)

6. Trust, confidence, and credibility of top management

3,300 3.17 .84 .53* .39* .33* .85* .57* (.89)

7. Information adequacy 3,311 3.44 .72 .44* .40* .34* .79* .54* .54* (.81)

8. Supportiveness from organization

3,289 3.37 .71 .51* .47* .38* .88* .67* .69* .60* (.74)

9. Horizontal communication 3,314 3.91 .52 .31* .38* .48* .46* .45* .26* .43* .40* (.87)

10. Level of contact with colleagues

3,312 3.81 .63 .25* .27* .36* .33* .35* .17* .30* .29* .81* (.75)

11. Openness and candor 3,312 3.89 .65 .27* .36* .45* .45* .43* .28* .41* .39* .86* .48* (.81)

12. Supportiveness from colleagues

3,311 4.05 .57 .26* .33* .41* .38* .36* .20* .38* .34* .87* .56* .68* (.79)

13. Turnover intentions 2,628 2.23 .93 �.51* �.32* �.31* �.36* �.31* �.29* �.30* �.33* �.24* �.15* �.21* �.26* (.87) Note. Internal consistency (Cronbach’s α) in parentheses on the diagonal. 5-point scales (1 = disagree, 5 = agree) were used. *p < .001, 2,608 < N < 3,310.

S. Holzwarth et al., Communication in Organizations 33

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

commitment (β = �.09, p < .05). Because all three effects are significant, our results confirm Hypothesis 4 in such a way that the combination of employees’ strong emotional commitment to the organization, to the team, and to the respective supervisor leads to particularly low turnover intentions.

An explicit multiple mediation framework was further considered. First, model C1 assumes that the effect of vertical communication on turnover intentions is mediated by organizational and supervisory commitment, whereas the effect of horizontal communication on turnover in- tentions is mediated by supervisory and team commitment (model H, Figure 1). Second, model C2 assumes that the effect of vertical communication on turnover intentions is mediated by organizational, supervisory, and team com- mitment, and the effect of horizontal communication on turnover intentions is also mediated by organizational, supervisory, and team commitment (cf. model A1, Figure 1). Model C2 improves model fit compared to model C1, Δχ2 (2) = 1,524.25, p < .001. Finally, model C3 assumes that the effect of vertical communication on turnover intentions is mediated by organizational, supervisory, and team commitment, and the effect of horizontal commu- nication on turnover intentions is only mediated by su- pervisory and team commitment (cf. model A3, Figure 1). This model has a better fit than model C2, and it is more parsimonious.

Regarding the indirect effects of vertical communication on turnover intentions, effects were obtained for team (a1b1 = �.08, p < .001), supervisory (a2b2 = �.11, p < .01), and organizational commitment (a3b3 = �.62, p < .001). Thus, a strong total indirect effect of vertical communication on turnover intentions via commitment (a1b1 + a2b2 + a3b3 = �.81) can be found. The remaining direct effect of vertical communication on turnover intentions becomes positive (c9 = .23, p < .05), thus reducing the total effect of vertical communication on turnover intentions (c = �.58, p < .001), although these results still support the

assumption of a (partial) mediation by commitment of the vertical communication turnover intentions’ relationship. Finally, with respect to the indirect effects of horizontal communication on turnover intentions, small, indirect ef- fects for team (a4b1 = �.13, p < .01) and supervisory commitment (a5b2 = �.05, p < .01) were obtained. This resulted in a weak, negative total indirect effect of hori- zontal communication on turnover intentions (a4b1 + a5b2 + a6b3 = �.18), whereas the direct effect of horizontal communication on turnover intentions was not significant (d9 = .00, n.s.). In other words, the total effect of horizontal communication on turnover intentions (d = �.18, p < .05) is the result of the indirect effects of team and supervisory commitment on turnover intentions, indicating full mediation.

Finally, data from our study came from employees at more than 50 locations. However, the locations did not seem to constitute either a theoretically important or an empirically relevant nesting factor (see the Discussion section). For example, we found small ICC(1) scores for the core research variables (Bliese, 1998, p. 360). In addition, multilevel SEM did not improve the fit of model C3. For details, see the Electronic Supplementary Material (ESM 1).

Discussion

We started with Luhmann’s (1984) notion that com- munication has a constitutive function for any organi- zation, and as a result, communication in organizations (organizational communication) should be related to important outcomes—from proximal variables, such as affective commitment, through more distal variables, such as turnover intentions. The assumption of a rela- tionship between organizational communication, af- fective commitment, and withdrawal is based on the

Figure 2. Structural equation model illus- trating the relationships between organiza- tional communication, affective commitment, and turnover intentions (model A3). Stan- dardized parameter values. *p < .05, **p < .01, ***p < .001.

34 S. Holzwarth et al., Communication in Organizations

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

theory of relational cohesion (Lawler & Yoon, 1996). According to this theory, repeated cooperation creates a “special” relationship, which in turn reduces the prob- ability of withdrawal. This theory can also be applied to the intraorganizational context with some specifica- tions: for example, analyzing communication given its close relationship with cooperation, using affective commitment as a measure of relationship, and consid- ering turnover intentions being an appropriate measure of withdrawal. Of note here is that the “communication creates af-

fective commitment” assumption implies that simply working in a cooperative organizational context is re- warding. For example, it is not necessary that there are additional organizational values, i.e., there exists no necessity of a strong organizational culture representing specific organizational values. It is also not necessary that organizational success is publicly portrayed as a basis to develop the employees’ self-worth. In brief, processes that are related to organizational identification (e.g., Marique & Stinglhamber, 2011) are not necessary in order to create affective commitment. Note that we do not assume that there is a direct (negative) effect of coop- eration on withdrawal. Rather, cooperation (and com- munication) is experienced as rewarding (Tabibnia & Lieberman, 2007), which contributes to the willingness to continue the relationship with the respective exchange partner. We proposed that organizational communication would

have an effect on affective commitment, which in turn should be related to a reduction of turnover intentions. In fact, both dimensions of organizational communication, horizontal and vertical, are related to turnover intentions, although driven by different mechanisms. Commitment foci represent particular entities, such as the employing organization, the respective team, or the individual su- pervisor to whom an employee can be committed (cf. Becker, 1992). We hypothesized distinct effects of hori- zontal communication on team commitment and of ver- tical communication on organizational commitment (= the target–similarity hypothesis). Additionally, we proposed that supervisory commitment is affected by both dimen- sions of communication because supervisors can be in- volved in both horizontal communication, for example, giving informal feedback, and vertical communication, for example, being responsible for the implementation and execution of organizational strategy. Our results are by and large consistent with both assumptions. More importantly, these distinct relationships between organizational com- munication and foci of commitment can better explain how both vertical and horizontal communication pro- cesses contribute to the effectiveness of organizations in an incremental manner.

Previous research had already suggested that organiza- tional communication is related to organizational com- mitment. According to a meta-analysis (Mathieu & Zajac, 1990), communication was the strongest of all contextual antecedents of organizational commitment. However, this result was based on a measure of the evaluation of com- munication rather than on how communication takes place. Thus, the respective correlation can be merely the result of an evaluative consistency in answering questions on com- munication and the organization. In addition, it was a rather narrow measure (leader communication). It is remarkable that direct effects of both forms of

communication on turnover intentions either are no longer significant, or positive, when taking affective commitment into account. For example, after controlling for affective commitment, vertical communication increases turnover intentions. One explanation might be that vertical com- munication, for example, supportive communication, is also an encouraging sign of recognition of the employees’ superior competencies that, in turn, foster career aspira- tions in employees and finally increase efforts to consider job opportunities outside of the deploying organization in order to speed up one’s career development.

Limitations

It should be noted that the data were collected in a Germany-based but globally operating multinational com- pany in the automotive industry. Hence, the data might represent a variety of economic environments and orga- nizational subcultures. Although some researchers have emphasized the need for studies that further contextualize workplace commitment, especially focusing on cultural characteristics and more local research (cf. Wasti et al., 2016), we decided to investigate the individual differences between members of the organization in total, rather than to focus on differences between locations. This decision was driven by our interest in investigating the effects of two dimensions of organizational communication in general. Note that location is not a “culture” or a “country” proxy because a number of locations are in the same country (Germany). In addition, location is also no proxy for a local culture or a group. In fact, employees often work in project teams across locations or in virtual project groups. There- fore, we were not surprised that no evidence for an influ- ence of nested data was found (see ESM 1). Nevertheless, we believe that we were successful in gaining a truly global and diverse sample, representing various cultures, hierar- chical levels, and functions within one single organization. We succeeded in having respondents from 54 locations worldwide who participated in the study, representing different regions (Europe, North and South America, Asia-

S. Holzwarth et al., Communication in Organizations 35

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Pacific), job foci (technological, commercial, and admin- istrative), and different hierarchical levels in terms of managerial, functional, and personal responsibilities.

Further issues are the cross-sectional nature of the study, the self-report measures employed, and restrictions asso- ciated with online survey research. As our data are cross- sectional in nature, one should be cautious with drawing final causal interpretations. Concerning turnover intentions, this could come into effect as employees who are consid- ering leaving the organization might, in turn, also lose their affective bonding to the organization, the supervisor, or the team and these employees might also experience a decline in the quality of communication activities directed toward them. Note, however, that previous research has consis- tently treated affective commitment as an antecedent of turnover intentions (cf. Meyer et al., 2002). In addition, all constructs analyzed, like communication, affective com- mitment, and turnover intentions, represent individual at- titudes and subjective evaluations of employees. Hence, we had to rely on respondents’ self-reports, which in turn can contribute to common method variance. The research follows Spector and Brannick (2009) who stated that problems occurring from measuring variables with the same method are overstated, casting doubt that the method itself produces systematic variance in observations, which inflates correlations to any significant degree. Moreover, we systematically compared competing models, further in- vestigating the most theoretically and empirically plausible model. Fortunately, the sample size of our study was large, allowing for the test of ambitious and complex structural models. Consistent with previous research (e.g., Bartels et al., 2010), we measured organizational communication by means of self-report scales, which, strictly speaking, represents communication as perceived by the respondents. Thus, future research might consider alternative measures of organizational communication (e.g., log file analyses or observations of communication activities).

Finally, online questionnaires to collect the data were used. As the online survey was distributed on company time, this approach systematically excludes employees without access to the internal IT infrastructure, resulting in a probably considerable exclusion of blue-collar workers. Our concentration on white-collar workers might limit the organization-wide generalizability of our findings.

Practical Implications and Future Research

Given the general and robust effects of affective com- mitment on various work-related attitudes and behaviors (e.g., Meyer et al., 2002, 2012) as well as its importance for leveraging firm performance (Chun et al., 2013), individual and organizational advantages of a highly committed

workforce are evident. Two questions for practitioners are (1) how to foster commitment and (2) whether it is ben- eficial to differentiate between commitment foci when measuring commitment. With respect to the first question, we hypothesized and found that organizational commu- nication is related to affective commitment, but the re- lationship depends on both the type of communication and the type of affective commitment. To achieve a high level of affective organizational commitment, it seems to be more important that employees feel informed about or- ganizational strategy and goals, are encouraged to par- ticipate in decision-making, and experience honest and direct interactions between themselves and management. If practitioners are more interested in fostering supervi- sory and team commitment, they should consider en- couraging interactions between employees on the same hierarchical levels, for example, by means of team de- velopment activities. Referring to our second question, the results clearly indicate the usefulness of distinguishing between three different commitment foci. First, high levels of affective organizational, supervisory, and team com- mitment incrementally contribute to reduced turnover intentions. Second, the research shows that the three commitment foci differentially mediate the relationship between organizational communication and turnover in- tentions. Although practitioners are interested in designing employee surveys to be as short as possible, the results underline the usefulness of separate measures of affective organizational, supervisory, and team commitment.

High commitment has been called a central criterion of good internal communications and internal public relations (Wright & Kehoe, 2009) as well as for evalu- ating team building activities (Paillé, 2009). Our results suggest that evaluations should be done with specific scales targeting the respective affective commitment focus. More importantly, improving both vertical com- munication by, for example, means of participatory processes and horizontal communication by, for exam- ple, means of team building training can be worth the effort if one takes into account how they specifically contribute to foci of affective commitment and finally decrease withdrawal intentions.

An important issue for future research is the replication of our findings with longitudinal data. This seems to be an important step to conclusively demonstrate a causal effect of organizational communication on affective commit- ment. It is still possible that the relationship also works the other way around, meaning that highly committed em- ployees are practicing, encouraging, or perceiving a special kind of communication, as already mentioned in the Limitations section. Longitudinal data could also be used to rule out another alternative explanation of the link between communication and commitment, which can be

36 S. Holzwarth et al., Communication in Organizations

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

found in an account on the evolution of human sociality (Watanabe & Smuts, 2004). Cooperation is a risky en- deavor for an individual because the cooperation partner can act in a self-interested way or because the cooperation partner might just be afraid of self-interested behavior on the side of the individual. Hence, signs of commitment, which are of course delivered, among other ways, through communication, can become important (Watanabe & Smuts, 2004). We also encourage future researchers to conduct re-

search with a special focus on under-investigated blue- collar employees. It seems plausible that the impact of vertical communication might be impaired in a classic manufacturing environment because top management is often not visible and organizational participation is more difficult, whereas horizontal communication processes between workers on the same organizational level could have a greater impact on affective team commitment and even turnover intentions. Finally, turnover intentions are only one important

withdrawal measure. Future research should also consider other outcomes, for example, higher motivation of em- ployees, or more organizational citizenship behavior, as well as whether the effects of organizational communi- cation on these outcomes are also mediated by various foci of affective commitment.

Electronic Supplementary Material

The electronic supplementary material is available with the online version of the article at https://doi.org/ 10.1027/1866-5888/a000261 ESM 1. Additional data analyses: ICC(1) scores and a test of a multilevel structural equation model.

References

Allen, N. J., & Meyer, J. P. (1990). The measurement and ante- cedents of affective, continuance and normative commitment to the organization. Journal of Occupational Psychology, 63(1), 1–18. https://doi.org/10.1111/j.2044-8325.1990.tb00506.x

Askew, K., Taing, M. U., & Johnson, R. E. (2013). The effects of commitment to multiple foci: An analysis of relative influence and interactions. Human Performance, 26(3), 171–190. https:// doi.org/10.1080/08959285.2013.795571

Bandalos, D. L. (2014). Relative performance of categorical diagonally weighted least squares and robust maximum likelihood estima- tion. Structural Equation Modeling: A Multidisciplinary Journal, 21(1), 102–116. https://doi.org/10.1080/10705511.2014.859510

Bartels, J., Peters, O., de Jong, M, Pruyn, A., & van der Molen, M. (2010). Horizontal and vertical communication as determinants of

professional and organisational identification. Personnel Review, 39(2), 210–226. https://doi.org/10.1108/00483481011017426

Becker, T. E. (1992). Foci and bases of commitment: Are they distinctions worth making? Academy of Management Journal, 35(1), 232–244. https://doi.org/10.2307/256481

Bliese, P. D. (1998). Group size, ICC values, and group-level cor- relations: A simulation. Organizational Research Methods, 1(4), 355–373. https://doi.org/10.1177/109442819814001

Cammann, C., Fichman, M., Jenkins, G. D., & Klesh, J. (1983). Michigan Organizational Assessment Questionnaire. In S. E. Seashore, E. E. Lawler, P. H. Mirvis, & C. Cammann (Eds.), As- sessing organizational change: A guide to methods, measures and practices (pp. 71–138). Wiley.

Chun, J. S., Shin, Y., Choi, J. N., & Kim, M. S. (2013). How does corporate ethics contribute to firm financial performance? The mediating role of collective organizational commitment and organizational citizenship behavior. Journal of Management, 39(4), 853–877. https://doi.org/10.1177/0149206311419662

Feinian, C., Curran, P. J., Bollen, K. A., Kirby, J., & Paxton, P. (2008). An empirical evaluation of the use of fixed cutoff points in RMSEA test statistic in structural equation models. Sociological Methods & Research, 36(4), 462–494. https://doi.org/10.1177/0049124108314720

Felfe, J., & Franke, F. (2012). COMMIT Commitment Skalen – Fragebogen zur Erfassung von Commitment gegenüber Organisation, Beruf/ Tätigkeit, Team, Führungskraft und Beschäftigungsform [COMMIT commitment scales – An instrument to assess commitment con- cerning one’s organization, profession/job, team, leader and em- ployment form]. Huber.

Felfe, J., & Yan, W. H. (2009). The impact of workgroup commitment on organizational citizenship behaviour, absenteeism and turnover in- tention: The case of Germany and China. Asia Pacific Business Review, 15(4), 433–450. https://doi.org/10.1080/13602380802667411

Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25(2), 161–178. https:// doi.org/10.2307/2092623

Lavelle, J. J., Rupp, D. E., & Brockner, J. (2007). Taking a multifoci approach to the study of justice, social exchange, and citizenship behavior: The target similarity model. Journal of Management, 33(6), 841–866. https://doi.org/10.1177/0149206307307635

Lawler, E. J., & Yoon, J. (1996). Commitment in exchange relations: Test of a theory of relational cohesion. American Sociological Review, 61(1), 89–108. https://doi.org/10.2307/2096408

Levine, J. M., & Moreland, R. L. (2012). A history of small group research. In A. W. Kruglanski, W. Stroebe, A. W. Kruglanski, & W. Stroebe (Eds.), Handbook of the history of social psychology (pp. 383–405). Psychology Press.

Luhmann, N. (1984). Soziale Systeme - Grundriß einer allgemeinen Theorie [Social Systems, Stanford: Stanford University Press, 1995]. Suhrkamp.

Marique, G., & Stinglhamber, F. (2011). Identification to proximal tar- gets and affective organizational commitment: The mediating role of organizational identification. Journal of Personnel Psychology, 10(3), 107–117. https://doi.org/10.1027/1866-5888/a000040

Mathieu, J. E., & Zajac, D. M. (1990). A review and meta-analysis of the antecedents, correlates, and consequences of organiza- tional commitment. Psychological Bulletin, 108(2), 171–194. https://doi.org/10.1037/0033-2909.108.2.171

Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continuance, and normative commitment to the or- ganization: A meta-analysis of antecedents, correlates, and consequences. Journal of Vocational Behavior, 61(1), 20–52. https://doi.org/10.1006/jvbe.2001.1842

Meyer, J. P., Stanley, D. J., Jackson, T. A., McInnis, K. J., Maltin, E. R., & Sheppard, L. (2012). Affective, normative, and continuance commit- ment levels across cultures: A meta-analysis. Journal of Vocational Behavior, 80(2), 225–245. https://doi.org/10.1016/j.jvb.2011.09.005

S. Holzwarth et al., Communication in Organizations 37

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Paillé, P. (2009). The relationship between support, commitment and intent to leave team: A social exchange perspective. Team Performance Management: An International Journal, 15(1/2), 49–62. https://doi.org/10.1108/13527590910937711

Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. https:// doi.org/10.18637/jss.v048.i02

Savalei, V. (2018). On the computation of the RMSEA and CFI from the mean-and-variance corrected test statistic with nonnormal data in SEM. Multivariate Behavioral Research, 53(3), 419–429. https://doi.org/10.1080/00273171.2018.1455142

Smidts, A., Pruyn, A. T. H., & Van Riel, C. B. M. (2001). The impact of employee communication and perceived external prestige on organizational identification. Academy of Management Journal, 44(5), 1051–1062. https://doi.org/10.2307/3069448

Spector, P. E., & Brannick, M. T. (2009). Common method variance or measurement bias? The problem and possible solutions. In D. A. Buchanan & A. Bryman (Eds.), The Sage handbook of orga- nizational research methods (pp. 346–362). Sage.

Stinglhamber, F., & Vandenberghe, C. (2003). Organizations and supervisors as sources of support and targets of commitment: A longitudinal study. Journal of Organizational Behavior, 24(3), 251–270. https://doi.org/10.1002/job.192

Tabibnia, G., & Lieberman, M. D. (2007). Fairness and cooperation are rewarding: Evidence from social cognitive neuroscience. Annals of the New York Academy of Sciences, 1118(1), 90–101. https://doi.org/10.1196/annals.1412.001

Vandenberghe, C., & Bentein, K. (2009). A closer look at the re- lationship between affective commitment to supervisors and organizations and turnover. Journal of Occupational and Orga- nizational Psychology, 82(2), 331–348. https://doi.org/10.1348/ 096317908X312641

Wasti, S. A., Peterson, M. F., Breitsohl, H., Cohen, A., Jørgensen, F., de Aguiar Rodrigues, A. C., Weng, Q., Xu, X. (2016). Location, location,

location: Contextualizing workplace commitment. Journal of Orga- nizational Behavior, 37(4), 613–632. https://doi.org/10.1002/job.2094

Watanabe, J. M., & Smuts, B. B. (2004). Cooperation, commitment, and communication in the evolution of human sociality. In R. W. Sussman (Ed.), The evolution and nature of sociality among human and nonhuman primates (pp. 288–309). de Grutyer.

Wombacher, J. C., & Felfe, J. (2017). Dual commitment in the or- ganization: Effects of the interplay of team and organizational commitment on employee citizenship behavior, efficacy beliefs, and turnover intentions. Journal of Vocational Behavior, 102, 1–14. https://doi.org/10.1016/j.jvb.2017.05.004

Wright, P. M., & Kehoe, R. R. (2009). Organizational-level ante- cedents and consequences of commitment. In H. J. Klein, T. E. Becker, & J. P. Meyer (Eds.), Commitment in organizations: Accumulated wisdom and new directions (pp. 285–307). Routledge/Taylor & Francis Group.

History Received June 9, 2018 Revision received July 17, 2020 Accepted July 23, 2020 Published online November 12, 2020

Acknowledgments We thank Florian Offergelt for comments on previous versions of the manuscript.

Klaus Moser School of Business, Economics, and Society University of Erlangen-Nürnberg Lange Gasse 20 90403 Nürnberg Germany [email protected]

38 S. Holzwarth et al., Communication in Organizations

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .