Research Article: 2021 Vol: 20 Issue: 4
Darko Tipuri?, University of Zagreb
Lana Cindri?, University of Zagreb
Domagoj Hruška, University of Zagreb
This paper aims to provide a comprehensive understanding of the identification split phenomenon in global voluntary organizations. Using a sample of 525 Rotary members in Croatia, the interlocking relationship between identification with the higher and lower-level organizational category was explored. Several important findings are presented. First, consistent with the prior studies, levels of subgroup, and superordinate identification proved to be significantly different. Further, contradictory to prior research, our research indicates that member engagement is a significant predictor of organizational identification and not vice versa.
Organizational Identification, Voluntary Organization, Organizational Engagement.
Modern organizations are complex systems who provide their members with the possibility of belonging to different organizational units, groups, and teams simultaneously. Organizational membership, therefore, implies the existence of a larger number of different organizational identities (Chen et al., 2013). The fact that members experience a sense of belonging to several categories within one organization indicates the complexity of the identification concept and stresses the need for a multilevel approach in its analysis. Although past studies provided empirical evidence that supports these claims, it remains unclear why such differences in the identification levels occur.
Global voluntary organizations, due to the specifics of their organizational structure, provide a particularly fruitful area for closer study of the multidimensionality in organizational identification phenomenon. The global voluntary organization provides members with the opportunity to adopt humanitarian values and commit to a global-level goal while smaller organizational units (workgroups or clubs) represent an arena where such values are nurtured. Members of such organizations, therefore identify with at least two organizational categories: macro-level category represented by a wider voluntary movement mission and a micro-level category represented by their workgroups.
The goal of this paper is to provide an in-depth understanding of the identification split phenomenon in voluntary organizations. For this study, it was of particular interest to investigate predictors causing the disparities in identification with different organizational levels. The paper represents the view that these disparities can be explained by two groups of predictors: (1) members' engagement in organizational activities and (2) club structural attributes. Several important findings emerged from the analysis. Our results provide answers to some of the most current questions on this topic as well as direct guidelines for further development of this domain in identification studies. Answering these questions is important because understanding of antecedents and consequences of organizational identification can aid leaders in their attempt to build more coherent organizational mental models, which is an especially important topic in voluntary organizations.
The paper is structured into four sections. The first section aims to provide an overview of the relevant literature on the organizational split phenomenon and to position our work in the ongoing debate on the topic. The second section, where methodological aspects of the research are presented, is followed by the discussion of the main findings and our concluding remarks.
Identification Split Phenomenon
Organizational identification is the extent to which members perceive oneness with the organization and integrate key components of organizational identities into theirs. Therefore, it is considered to be a process of cognitive linking between the definition of the organization and the definition of the self (Dutton et al., 1994; Ravasi & van Rekom 2003; He & Brown 2013). However, if there are multiple identifications to different entities within a complex organization, then identification cannot be observed as a unique and undisputed construct.
Organizations thus hold multiple identification relations with their members. It is an identification split phenomenon: the situation when a member simultaneously identifies with different organizational levels and when it is difficult to isolate clear and unbiased overall identification to the organization. Thus organizations provide their members with numerous “hats to wear” by allowing them to belong to more than one organizational category at the same time (Ashforth & Johnson 2001).
It is particularly interesting to observe the differences between the identification to the entire organization (macro-level) and the organizational entity (unit, team, etc.) in which the organization member operates (micro-level). When complex organizations have different key identities at macro and micro levels, then it can be assumed that the intensity of members' identification to those levels differs. Identification with a higher organizational category such as the organization as a whole is referred to as the superordinate identification, while identification with unites nested in such category is referred to as the subgroup identification (Lipponen et al., 2003).
The fact that organizations provide the possibility of simultaneous belonging to different organizational categories raised the issue of determining the difference in the intensity of identification among those categories and comparing their impacts on certain work-related behaviors. Previous studies reported higher levels of subgroup identification compared to identification with the superordinate category. A meta-analysis of 40 independent samples by Riketta & van Dick (2005) provided evidence that the level of identification and commitment to the work-group on average is higher than to the entire organization. Using two samples, first being composed of the local government officials and second, of university faculty, van Knippenberg & van Schie (2000) demonstrated that employees identify more with their work-groups than with the entire company. Similar results were reported by Christ et al., (2003) on the sample of school teachers and by Barker & Tompkins (1994) who also provided evidence that long-term employees are more identified with both their teams and the company than short term employees. The intensity of work-group and organizational identification can also serve as a predictor of members' attitudes on upcoming organizational changes. As Jetten et al. (2002) demonstrated, both subgroup and superordinate identification are predictors of employees' emotions concerning organizational restructuring. Employees who are more strongly identified with their work-group express stronger negative feelings about organizational restructuring, while employees who are more strongly identified with a company as a whole show less negative feelings about organizational restructuring.
Accordingly, subgroup identification is considered to have greater power in influencing different work-related attitudes and behaviors. Van Knippenberg & van Schie (2000) were able to prove that work-group identification is more closely associated with several important work-related outcomes such as job satisfaction, job involvement, and job motivation. Further, employees' behaviors concerning the organization as a whole such as turnover intentions are proven to be more strongly linked to organizational identification while workplace behaviors, such as the perception of workplace climate, satisfaction with colleagues and superiors, and altruistic behavior are more closely associated with work-group identification (Riketta & van Dick, 2005). In cases of concurrent high-level superordinate and subgroup identification employees demonstrate a higher level of job satisfaction and organizational citizenship behavior than when they strongly identify with just one organizational level and poorly with the other.
Although numerous studies have focused on determining the antecedents of organizational identification, a surprisingly small number of authors examined how differences in subgroup and superordinate identification can be explained in terms of their different antecedents. One of the first to address this issue was Reade (2001) who identified different groups of antecedents to identification with the global multinational company and identification with its local subsidiary. He was able to show that the identification with the subsidiary is primarily influenced by factors related to the local context in which the subsidiary operates while identification with the multinational company is predicted by factors on a global level. Respectively, factors related to the external organizational environment such as perceived external prestige appear to be stronger predictors of organizational identification, while factors concerning organizational internal environment such as communication climate are more strongly related to identification with organizational sub-categories (Bartels et al., 2007). Lipponen et al. (2003) examined antecedents to four different identification profiles (1) having both strong superordinate and subgroup identification, (2) having strong superordinate and low subgroup identification, (3) having strong subgroup and low superordinate identification and (4) having week both superordinate and subgroup identification. In their study, the group prestige was the strongest predictor for the high levels of both work-group and organizational identification. Group size was also proven to be a strong predictor of identification profile, as it was shown that the respondents from smaller groups and who have frequent contacts with other group members have a profile characterized by a strong identification with the work-group and low identification with the whole organization.
Identification in Voluntary Organizations
Identifying to voluntary organizations is considered to be of great importance to organizations that rely on members’ voluntary work since it drives members' motivation to continue their engagement (Steimel, 2013). Unfortunately, voluntary associations have been relatively neglected in organizational identification studies given that most of the authors focused on the relationship between employee identification in profit organizations and numerous job-related behaviors and performance outcomes. Although identification within voluntary organizations has been confirmed as a predictor of engagement in the organization’s voluntary activities, financial contribution to the organization and members' satisfaction (Meisenbach & Kramer, 2014; Kang, 2016) it remains unclear how volunteers identify with different organizational levels of their organization and what causes the potential identification split.
For instance, global voluntary organizations like Rotary, Lions, and Kiwanis, have a have a pronounced impact in humanitarian and social activities in many communities worldwide. The key attribute of these organizations is a specific organizational structure: they are formed as associations of a large number of formally established entities (groups or clubs) and relevant activities are based on integrative organizational leadership, a set of common principles, values and members' commitment.
Although global organizations of this type have a leadership structure, well-regulated set of rules and behaviors (obligations that each member has, such as membership fees, participation in projects, etc.), and often planning and monitoring systems - they are based primarily on clearly defined a set of common principles and values, a distinctive mission and joint integration activities that connect members, giving them a bigger or lesser sense of belonging or identification with the organization. On the other hand, many of the activities of such organizations happen at the micro-level - in the clubs. The clubs are groups of individuals that implement the values and goals of the organization in specific actions in local communities. They are a true example of organized teams: members meet and join regularly, plan and implement local community actions within the scope of a global nonprofit organization and participate in projects that initiate, and lead the leadership or superstructure of a global organization. Besides, clubs are characterized by socio-political relations and group dynamics. Members feel and behave based on friendship, joint goals, and interests, but also their relations depend on the level of group cohesion, structures of informal power and influence, cooperation, alliances, and conflicts. Members are voluntarily involved in the activities of such organizations, meaning that key organizational mechanisms are reduced to those who develop the identification, dedication, and motivation of members, and most of them are related to the core values of organization at the macro level. Many members of such organizations are strongly identified with such organizations. Members of such organizations are identified with at least two levels of organization at the same time: a macro level with a global voluntary organization and a micro level with the club. At the same time, members of these global organizations identify to the values, mission and organizational goals, global projects, history, influence, prestige and/or brand, and also to the club in which member relates with other members and establishes different sources of linkages (for instance, attractiveness to be part of the group, good interpersonal relations, and friendship, active participating in club humanitarian action, shared experience in serving community needs, etc.).
Research Objectives and Hypotheses
This paper aims to provide an in-depth understanding of the identification split phenomenon in voluntary organizations. First, the paper investigates how members relate to their organization at the superordinate and subgroup level. Further, the paper seeks to examine whether the identification split between these categories can be explained by a set of different antecedents. In doing so, we looked at two groups of antecedents: (1) member engagement, (2) attributes of their subgroup category (club).
We start by proposing that volunteers have a stronger sense of belonging to their clubs than to their global volunteering organization. Van Knippenberg & van Schie (2000) stated three reasons why the level of subgroup identification is expected to be higher than the superordinate identification: (1) drawing on Brewer (1991) they suggest that people identify more with smaller groups such as work-groups because large groups can pose a threat to individual distinctiveness, (2) members will have more common features with their work-groups than with the whole organization since they share a sense of common fate with the work-group, (3) group members more frequently meet with members of other organizational groups than with members of other organizations and therefore they emphasize their work- group affiliation more often. Also, work-group identities are more salient to members because they are characterized by teamwork and lateral communication.
Further, we propose that a higher level of member engagement in organizational activities is related to higher levels of superordinate identification and lower levels of subgroup identification. Namely, we propose that members with longer organizational tenure and those experienced in performing certain organizational formal duties will express a stronger sense of belonging to the global organization than to their club.
We further advocate that club structural attributes have the ability to influence both subgroup and superordinate identification levels. The following club attributes are proposed as important identification predictors: (1) club size, (2) club age, (3) member age diversity, and (4) member gender diversity.
The target population in the study was Rotary members in Croatia. An online questionnaire was used to collect the data and our final sample consisted of 525 Rotarians from 55 clubs, making the response rate of 44,68%. The male members made the majority in our sample, representing 77,71% of all respondents. Most of the respondents were in the age groups 41-45 years (18.10%) and 51-55 years (16.57%).
The largest club in the sample was composed of 58, and the smallest of 10 members. The average age of the club was 12.25 years. Of the total number of members in the observed clubs, women accounted for 22.81%. In as many as 36.36% of the clubs, no women were present, and in only three of the clubs the share of women was higher than 50%. The observed clubs were mostly composed of middle-aged or elderly members and less than 10% of the clubs had at least half of the members under the 40 years of age.
Mael & Ashforth's (1992) scale was used to measure identification with Rotary International and Rotary club. Sample items from the scale are: “I usually say "we" rather than "they" when I talk about Rotary International” and „I usually say "we" rather than "they" when I talk and my Rotary club”. Both scales had high Cronbach's alpha scores (0,75 for Rotary International, and 0,82 for Rotary club).
Member engagement was observed by focusing on two indicators. The first of them being tenure with the organization and the second performing a formal duty within the organization i.e. serving either as a club chair, club secretary, or district committee chair. Secondary sources were used to collect data on club attributes. Clubs' size was shown by the total number of club members and clubs' age was expressed in years. Three types of clubs were identified considering clubs' gender diversity measured by the proportion of women present. The same categorization was used for clubs' age diversity measured by the proportion of members under 40 years of age. Data processing and statistical analysis were carried using MS Excel and SPSS.
Our analysis started by examining the relationship between levels of identification with the Rotary International and Rotary club. Table 1 presents descriptive statistics for our variables.
| Table 1 Identification with Rotary International and Rotary Club | ||||
| Mean | N | Std. Deviation | Std. Error Mean | |
| OID-Rotary International | 3.96 | 525 | 0.67 | 0.02 |
| OID-Rotary Club | 4.18 | 525 | 0.72 | 0.03 |
Pearson's correlation coefficient showed us that the identification levels were strongly positively correlated (r=0.868; p<0.01). Further, results indicated that the observed difference in identification is statistically significant (t(524)=-10,226, p=0.000).
The next step in our analysis was to investigate whether the identification split can be explained by member engagement and club attributes. Tenure with Rotary as an indicator of member engagement was examined first. Identification levels considering members' tenure are presented in Table 2.
| Table 2 Organizational Identification and Tenure | ||||
| Tenure | Mean | Std. Deviation | N | |
| OID-Rotary International | = 2 | 3.80 | 0.76 | 148 |
| 3 - 6 | 4.03 | 0.61 | 160 | |
| 7 - 10 | 4.06 | 0.63 | 90 | |
| 11 - 14 | 4.12 | 0.64 | 63 | |
| = 15 | 3.89 | 0.65 | 64 | |
| Total | 3.96 | 0.67 | 525 | |
| OID-Rotary club | = 2 | 4.02 | 0.82 | 148 |
| 3 - 6 | 4.21 | 0.69 | 160 | |
| 7 - 10 | 4.31 | 0.62 | 90 | |
| 11 - 14 | 4.39 | 0.59 | 63 | |
| = 15 | 4.03 | 0.72 | 64 | |
| Total | 4.17 | 0.72 | 525 | |
The highest intensity of identification with both Rotary International and Rotary club was recorded for members whose tenure was from 11 to 14 years while members with tenure of 15 and more years had lower identification compared to other tenure categories. The observed differences proved to be statistically significant [F (4.520) =4.06, p=0.003]; [F (4.520) =4.75, p=0.001].
The two-way multivariate analysis of variance (two-way MANOVA) was used to capture the relationship between our outcome variables (OID-Rotary International and OID-Rotary club). MANOVA enabled us to detect differences in outcome variables between the different levels of the independent variables (Table 3).
| Table 3 Manova (Tenure) | ||||||
| Effect | Value | F | Hypothesis df | Error df | Sig. | |
| Intercept | Pillai's Trace | 0.972 | 8970.675b | 2.000 | 519.00 | 0.000 |
| Wilks' Lambda | 0.028 | 8970.675b | 2.000 | 519.00 | 0.000 | |
| Hotelling's Trace | 34.569 | 8970.675b | 2.000 | 519.00 | 0.000 | |
| Roy's Largest Root | 34.569 | 8970.675b | 2.000 | 519.00 | 0.000 | |
| Tenure | Pillai's Trace | 0.043 | 2.839 | 8.000 | 1040.00 | 0.004 |
| Wilks' Lambda | 0.957 | 2.850b | 8.000 | 1038.00 | 0.004 | |
| Hotelling's Trace | 0.044 | 2.861 | 8.000 | 1036.00 | 0.004 | |
| Roy's Largest Root | 0.038 | 4.956c | 4.000 | 520.00 | 0.001 | |
| a. Design: Intercept + Tenure | ||||||
| b. Exact statistic | ||||||
| c. The statistic is an upper bound on F that yields a lower bound on the significance level. | ||||||
The results indicated that there is a statistically significant difference between identification with Rotary International and Rotary club based on members' tenure, F (8. 1036=2.85, p<0.0005; Wilk's Λ=0.96).
An identical procedure was used to examine the effect of members' experience in performing a formal duty within the organization (Table 4). More than half of the respondents (52.58%) performed at least one formal duty, 38.67%, of them being club chairs, 34.67% club secretaries, and 12.19% district committee chairs.
| Table 4 Organizational Identification and Formal Duties | ||||
| Formal duty | Mean | Std. Deviation | N | |
| OID-Rotary International | Yes | 4.06 | 0.63 | 276 |
| No | 3.85 | 0.90 | 249 | |
| Total | 3.96 | 0.67 | 525 | |
| OID-Rotary club | Yes | 4.28 | 0.66 | 276 |
| No | 4.06 | 0.77 | 249 | |
| Total | 4.18 | 0.72 | 525 | |
Members experienced in performing formal duties within the organization had both higher levels of identification with Rotary International [F (1.523) =12.83, p=0.000] and Rotary club [F (1.523) =13.46, p=0.000].
Further, as results is Table 5 indicate the difference between identification with Rotary International and Rotary club based on members' experience in performing formal duties within the organization was proven to be statistical y significant F (2.522) =7.28, p<0.0005; Wilk's Λ=0.001).
| Table 5 Manova (Formal Duty) | ||||||
| Multivariate Testsa | ||||||
| Effect | Value | F | Hypothesis df | Error df | Sig. | |
| Intercept | Pillai's Trace | 0.975 | 10300.858b | 2.000 | 522.00 | 0.000 |
| Wilks' Lambda | 0.025 | 10300.858b | 2.000 | 522.00 | 0.000 | |
| Hotelling's Trace | 39.467 | 10300.858b | 2.000 | 522.00 | 0.000 | |
| Roy's Largest Root | 39.467 | 10300.858b | 2.000 | 522.00 | 0.000 | |
| Formal duty | Pillai's Trace | 0.027 | 7.275b | 2.000 | 522.00 | 0.001 |
| Wilks' Lambda | 0.973 | 7.275b | 2.000 | 522.00 | 0.001 | |
| Hotelling's Trace | 0.028 | 7.275b | 2.000 | 522.00 | 0.001 | |
| Roy's Largest Root | 0.028 | 7.275b | 2.000 | 522.00 | 0.001 | |
| a. Design: Intercept + Formal duty | ||||||
| b. Exact statistic | ||||||
The next step was to examine identification split considering club attributes. Identification levels in clubs of different sizes, ages, and diversity are shown in the following tables (Table 6-9).
| Table 6 Organizational Identification and Club Size | ||||
| Club size | Mean | Std. Deviation | N | |
| OID-Rotary International | < 20 | 4.04 | 0.76 | 70 |
| 20-30 | 3.97 | 0.68 | 272 | |
| > 30 | 3.92 | 0.62 | 183 | |
| Total | 3.96 | 0.67 | 525 | |
| OID-Rotary club | < 20 | 4.24 | 0.70 | 70 |
| 20-30 | 4.16 | 0.77 | 272 | |
| > 30 | 4.18 | 0.66 | 183 | |
| Total | 4.18 | 0.72 | 525 | |
| Table 7 Organizational Identification and Club Longevity | ||||
| Club age | Mean | Std. Deviation | N | |
| OID-Rotary International | <10 | 4.00 | 0.65 | 140 |
| 10-20 | 3.99 | 0.69 | 283 | |
| > 20 | 3.83 | 0.64 | 102 | |
| Total | 3.96 | 0.67 | 525 | |
| OID-Rotary club | Up to 10 years | 4.23 | 0.70 | 140 |
| 10-20 years | 4.21 | 0.74 | 283 | |
| More than 20 years | 4.03 | 0.70 | 102 | |
| Total | 4.18 | 0.72 | 525 | |
| Table 8 Organizational Identification and Gender Diversity | ||||
| the proportion of women | Mean | Std. Deviation | N | |
| OID-Rotary International | > 20% | 3.96 | 0.68 | 260 |
| 20-50% | 3.93 | 0.68 | 172 | |
| > 50% | 4.05 | 0.64 | 93 | |
| Total | 3.96 | 0.67 | 525 | |
| OID-Rotary club | > 20% | 4.14 | 0.74 | 260 |
| 20-50% | 4.17 | 0.75 | 172 | |
| > 50% | 4.31 | 0.61 | 93 | |
| Total | 4.18 | 0.72 | 525 | |
| Table 9 Organizational Identification and Age Diversity | ||||
| members under 40 years of age | Mean | Std. Deviation | N | |
| OID-Rotary International | > 20% | 3.96 | 0.70 | 300 |
| 20-50% | 3.96 | 0.65 | 201 | |
| > 50% | 4.06 | 0.50 | 24 | |
| Total | 3.96 | 0.67 | 525 | |
| OID-Rotary club | > 20% | 4.17 | 0.74 | 300 |
| 20-50% | 4.17 | 0.70 | 201 | |
| > 50% | 4.31 | 0.63 | 24 | |
| Total | 4.18 | 0.72 | 525 | |
Although at the descriptive level, some differences in observed cases were apparent, none of the club's attributes proved to be a statistically significant predictor of the identification split (for all of the cases, the value of Wilk's Λ was greater than 0.05).
Several important findings emerged from our analysis. Firstly, our results regarding the relationship between superordinate and subgroup identification are supporting those reported in previous studies (Lipponen et al., 2003). Members with a strong sense of belonging to a global organization, have a strong sense of belonging to their clubs as well. Such a result can be explained, at least to some extent, by the fact that all levels of voluntary organizations are committed to a common goal. Also in line with the previous studies, identification levels proved to be significantly different, as members identified more intense with their clubs than with the global organization.
Such a split in organizational identification can be explained, to some extent, by different levels of member engagement. Namely, both tenure and member experience in performing a formal duty were found to be significant predictors of identification differences. Some findings regarding organizational identification and member engagement were noted in previous studies as well. Using a 13-item engagement scale with three dimensions: affective commitment, positive affectivity and empowerment, Kang (2016) was able to prove that organizational identification is related to the volunteer engagement in a faith-based charity organization as well as to have a moderating effect on the relationship between member satisfaction and affective commitment. However, Kang’s research examined the effects of volunteer-organization identification and satisfaction on volunteer engagement, while we argue how actually the engagement itself triggers the organizational identification. That notion is also supported by our second finding – the relation between identification and tenure. If we look at the identification levels by tenure categories, it is evident that identification increases as tenure increases but only up to the certain level when it starts to decline again. This finding is hard to elaborate from the point of view in which organizational identification triggers engagement since that would mean that the organizational identification is diminishing in time. In the case that the organizational members would cease to correspond to the organizational purpose, in volunteer type organizations, they would most likely leave the organizations altogether. That line of thinking is also congruent with our third finding, that the organizational members who are more experienced in performing formal organizational duties identified more strongly to both organizational categories.
Also, our study tries to explain the identification differences by structural attributes of the lower-level organizational units. Prior findings regarding member demographics as organizational identification predictors are at the very least contradictory. Our study did not demonstrate existence of relation between such attributes and organizational identification. Although the observed club attributes (size, age, and diversity) didn't found to be statistically significant in explaining the identification split, the results outlined avenues for further research.
Several limitations should be acknowledged when interpreting the results of this study. First, there is a lack of prior studies on the sources of identification split which has made it difficult to identify all of the key predictors to this phenomenon. Further, the study was carried over self-reported data which makes it eligible for the response bias. Lastly, since the study is carried in a rather specific form of voluntary organization, one should be careful in applying these findings to voluntary organizations in general.
The presented findings confirm that volunteers demonstrate different levels of organizational identification to their micro and macro environment. Such differences can be explained to some extent by the levels of members' engagement in their club activities. This is an important finding in the understanding of the complex relationship volunteers hold with their organizations. The broad implication of the present research is that organizations should consider designing more engagement opportunities for their members and thus strengthen their sense of belonging to organization.
Our main findings indicate need for different approach in observing relation between volunteer-organization identification and engagement. Contradictory to prior research, we argue that the it is actually the engagement itself that triggers the organizational identification and not vice versa. Firstly, both tenure and member experience in performing a formal duty were found to be significant predictors of identification differences. Secondly, relation between tenure categories and identification level shows non-linear trend indicating more complex association. Lastly, we have found that the organizational members who are more experienced in performing formal organizational duties also identify more strongly to both organizational categories.
Further research should examine which engagement activities have more potential to influence member identification. In addition, an important issue for future research is to explore in more detail how a group (club) attributes contribute to identification levels. Apart from club structural attributes, further studies should pay more attention to group dynamics processes such as group cohesiveness, trust and conflict.