Effects of Organizational Identifications on Supportive Behavior toward Former Organizations

Recent research has highlighted possible benefits to companies from former employees. This study draws on organizational identification research to explore the factors that affect former employees' supportive behavior toward their former organizations. The analysis of the data obtained from 302 Japanese employees indicated that supportive behavior toward a former organization positively correlates to identification with a former organization and the perceived prestige of the organization. Furthermore, the results also revealed that the relationship between supportive behavior toward a former organization and identification with the organization is moderated by the perceived prestige of the former organization and identification with the present organization. The theoretical contributions of the study are twofold. First, the findings suggest that drawing on organizational identification literature can advance research on the positive influence of employee mobility. Second, the study is the first attempt to address identifications with both the former organization and present organization simultaneously.


Introduction
In recent years, the increase in employee mobility has prompted re-examination of how turnover affects companies.Early studies view the consequences of turnover through the lens of human capital theory or cost analysis.More recent studies re-examine its effects from the viewpoint of social capital theory.Shaw, Duffy, Johnson, and Lockhart (2005) indicated that turnover diminished firms' internal social capital.Broschak (2004) associated managerial turnover with market tie dissolutions.Song, Almeida, and Wu (2003) showed that employee mobility damaged companies that lose employees via knowledge spillover.Thus, most studies have cited adverse effects of employee turnover (Hausknecht & Trevor, 2011).However, several recent studies have found that employee mobility can positively impact companies that lose employees.Somaya, Williamson, and Lorinkova (2008) argued that employee inflows from/outflows to customers may increase business with those customers.Carnahan and Somaya (2013) addressed this "alumni effect" by considering the effect of competitors' employee mobility.By analyzing patent data, Corredoira and Rosenkopf (2010) found that mobility-driven knowledge flows can be bi-directional between firms that lose employees and their new firms.
Such positive impacts are derived from former employees.Former employees may become customers, suppliers, or collaborators for their former companies (Hoffman, Casnocha, & Yeh, 2014).Such "alumni" can behave as goodwill ambassadors of their former companies at their present companies or bring valuable knowledge to the former companies (Carnahan & Somaya, 2015).
Although extant studies (Carnahan & Somaya, 2013;Corredoira & Rosenkopf, 2010;Somaya et al., 2008) shed light on the positive effects of employee mobility, they only indicate positive relationships between employee mobility and performance variables.The conditions under which job-switching employees benefit their former organizations have not been completely articulated.Some of the conditions might lie with former employees although situational factors can also be important.It is obvious that every former employee does not always bring positive effects to their former companies.Thus, it would be worthwhile to examine what kind of former employees bring positive effects to the former companies.
A relevant consideration is former employees' attitudes toward their former employers.Mael and Ashforth's (1992) classic work in organizational identification research suggested that former employees' attachment to their former employer may be associated with supportive behavior toward them.However, despite their suggestion, few studies explore how identification with former employers relates to supportive behavior toward them.
The aim of this study is to investigate characteristics of former employees who bring positive effects to former companies.Drawing on organizational identification research, the current paper extends the literature in two ways.First, it discusses the factors that are associated with support for former organizations.Identifying attitudinal antecedents of supportive behavior toward former companies will promote research on turnover effects.Second, this study will extend organizational identification literature by simultaneously addressing employees' identification with former organizations (IDFO) and identification with present organizations (IDPO).Previous research investigates them separately.

Identification with Former Organizations
Organizational identification research has developed significantly over the last two decades since Ashforth and Mael (1989) introduced the social identity approach (cf.Hornsey, 2008).They defined organizational identification as a sense of oneness or belongingness with one's organization.A majority of later research has adopted their definition.
Based on their definition, Mael and Ashforth (1992) later conducted an empirical study of identification with an alma mater, using data collected from the alumni of a religious college in the U.S. Their results showed that identification with an alma mater had a positive association with supportive behavior toward the college, such as making financial contributions and having a willingness to advise one's offspring and others to attend.They also indicated that both organizational factors, such as organizational prestige, and personal factors, such as tenure, were antecedents of organizational identification.Furthermore, Mael and Ashforth (1992) suggested implications of the concept of "alumni" for a variety of organizations, stating, "It should be determined whether identification with the former employer fosters continued proactive behavior on behalf of that employer (Mael & Ashforth, 1992, p. 119)."Although Rousseau (1998) noted that employees can have an attachment to the organizations they have left, much of the research after Mael and Ashforth (1992) have focused only on employees' identification with current organizations.
The only notable exception is Iyer, Bamber, and Barefield (1997), who developed a revised model based on Mael and Ashforth (1992).They added some antecedents to identification with former firms, such as alumni relations and personnel policies.Using data obtained from the alumni of three large public accounting firms (three of the Big Six accounting firms), they showed that IDFO was positively associated with the inclination to support the former firms.In addition, they indicated that several factors, such as a former firms' prestige, personnel counseling, and alumni relations had significant and positive associations with IDFO.
However, their study has limited application to a range of professions and organizations.Three limitations are especially noteworthy.First, the Big Six firms were prominent in the accounting industry.Former employees of the big accounting firms generally continued to work as accounting professionals after their departure from these firms.Work at the present organizations often reminded them of their former organizations because these big firms hold prominent positions in the accounting industry.That is, the salience of former firms was extremely high for the participants, which is not necessarily the case for everyone who changes jobs.Second, Iyer et al. (1997) conducted their survey by using alumni relations directories of big public accounting firms.Unlike these firms, it does not appear to be a norm for organizations to maintain alumni relations directories.Third, people who agree to be listed in alumni directories inherently feel more charitable toward their former firms than typical departed employees.Therefore, as Iyer et al. (1997) partially acknowledged, self-selection bias might affect their results.
In addition to the limitations, Ashforth and Mael (1992) and Iyer et al. (1997) have a common critical limitation, although they raise meaningful issues.These studies shed light only on IDFO and do not address identification with the present organizations (IDPO) where former employees currently work.From the viewpoint of multiple identifications, both IDFO and IDPO can be referred to simultaneously.
Multiple identifications have been eagerly investigated in the literature over the past ten years (Ashforth & Johnson, 2001;Ramarajan, 2014;van Knippenberg & van Schie, 2000).In short, studies on multiple identifications in organizational contexts have addressed multiple foci of identification.It is understandable that the two earlier studies did not tackle multiple identifications because Mael and Ashforth's (1992) is one of the earliest empirical studies and Iyer et al. (1997) followed Mael and Ashforth's (1992) model.
The effects of IDFO and IDPO should be explored simultaneously because IDPO, which is more easily evoked than IDFO, can affect various behaviors, including supportive behavior toward former organizations.Therefore, the present study addresses both IDFO and IDPO from the viewpoint of multiple identifications and examines how both affect supportive behavior toward former organizations.

Effects of Identification with Former Organizations
A vast body of literature has addressed identification with organizations to which people currently belong.However, people can identify with their former organizations despite lacking current membership in the organizations.Pratt (1998) argued that identification with an organization does not require belonging to it because social identity derives from a cognitive mechanism (Ashforth & Mael, 1989).Tosti-Kharas (2012), for example, investigated continued organizational identification after an involuntary job loss.
Extensive literature demonstrates that identification with an organization greatly affects attitudes and behavior of its members.Many studies show that organizational identification engenders supportive behavior toward the organization, such as organizational citizenship behavior and cooperative behavior (e.g., Dukerich, Golden, & Shortell, 2002;van Dick, Grojean, Christ, & Wieseke, 2006).
These findings can apply to IDFO.That is, a person with high IDFO is apt to behave supportively toward the former organization as Mael and Ashforth (1992) and Iyer et al. (1997) indicated.Therefore, the following hypothesis is posited.
H1: Former employees' inclination to support their former companies is positively related to the employees' identification with the companies.

Effects of Perceived Prestige of Former Organizations
When former employees rate products or services of former organizations as outstanding, it is easy for them to adopt supportive behavior toward the organizations.Similarly, the belief that former organizations are distinguished can elicit supportive behavior toward them.In organizational identification literature, such evaluation and belief generally correspond to organizational prestige (Fuller, Hester, Barnett, Frey, Relyea, & Beu, 2006).
Previous studies show a positive association between the perceived prestige of former organizations and supportive behavior toward them.Before their empirical analyses Mael and Ashforth (1992) and Iyer et al. (1997) posited that organizational prestige was an antecedent to organizational identification.However, both studies indicated that organizational prestige was significantly and directly associated with the supportive behavior.Therefore, high prestige of former organizations leads to supportive behavior toward them, leading to the second hypothesis: H2: Former employees' inclination to support their former company is positively related to their perceived prestige of the company.
Organizational prestige can also moderate the relationship between former employees' inclination to support their former companies and their identification with them.Even though former employees have strong identification with former firms, they might hesitate to support their former firms when they are not sure that the products or services of their former firms are outstanding.On the other hand, the high prestige of former firms can encourage former employees who have strong identification with former firms to support the firms.

H3:
The relationship between former employees' inclination to support their former companies and their identification with the companies is more positive when their perceived prestige of the companies is high.

Moderating Effect of Identification with Present Organizations
Starting to work for a new company engenders multiple identifications in an employee.Joining a new company brings about identification with the company and membership in the new company can potentially change an employee's attitudes and behavior toward a former organization.However, this does not mean that IDPO replaces IDFO completely (cf.Conroy & O' Leary-Kelly, 2014) because people desire continuity in self-concepts over time (Steele, 1988).Therefore, both present organizations and former organizations can be simultaneous targets of identification, although their degrees may vary.IDPO  Data were self-reported because this method is appropriate for self-referential constructs, such as organizational identification (cf.Conway & Lance, 2010).Participants who had changed jobs multiple times were asked to base their response on their most recent previous employer.It was found that 40.4% of the participants had changed jobs only once.
The data sample was restricted to 302 individuals who were certain that their former company was still operating.The average age was 36.28 years (S.D. = 6.79 years), and 35.4% were female.The average tenure at the former company was 4.73 years (SD=3.71).Respondents had worked at their present company for an average of 5.67 years (SD = 3.98).Of the respondents, 27.5% held a managerial/supervisory position at their present companies and 20.2% held a managerial/ supervisory position at their former companies at the time of their departure.
The sample was diverse in terms of industries.A dummy variable was created indicating whether participants' current company and former company were in a similar industry.When both companies were in a similar industry, the value of 1 was assigned, and the value of zero was assigned for other situations.Data show 58.6% of the participants remained in similar industries.

Measures
All measures were translated into Japanese using a standard translation-back-translation procedure (cf.Brislin, 1970).All responses were measured on five-point Likert scales (1 = strongly disagree, 5 = strongly agree).Supportive behavior toward former organization was assessed by three items adapted from Mael and Ashforth (1992), Iyer et al. (1997), and Ahearne and Bhattacharya (2005).The three items were as follows: "When I am asked for my opinion on my former company's products or services, I will provide a positive recommendation," "If I hear unfounded negative word-of-mouth about my former company, I will strongly deny it," and "When someone expresses their interest in working at my former company, I will recommend them to work for the company" (α= .82).
IDFO and IDPO were measured using five items derived from Mael and Ashforth (1992).A sample item is "When someone criticizes the company, it feels like a personal insult."One item-"When I talk about the organization, I usually say 'we' rather than 'they' "-in Mael and Ashforth's (1992) scale was not adopted.Most Japanese workers do not say "them" when they talk about the company they work for, but tend to refer to that company as "my company" (Takao, 2013).Reliability exceeded 0.80 for both scales (IDFO= .89,IDPO= .84).
Perceived prestige of the former company was measured using a three-item scale developed by Iyer et al. (1997) and Wan-Huggins, Riordan, and Griffeth (1998).A sample item is "Generally, I think the company has a good reputation in the industry."(α= .90).As a control variable, voluntariness of job change was measured using one item ("Leaving the company was voluntary.")
Descriptive statistics and zero-order correlations for all variables appear in Table 1.The correlation between IDFO and IDPO was significantly positive.It corresponds to previous studies of multiple identifications.The mean for IDPO (3.15) exceeded that for IDFO (2.90).This result seems convincing because this indicates that present organizations tend to have more salience than former organizations.

Hypothesis Testing
A hierarchical regression analysis on supportive behavior toward the former company was conducted to test the hypotheses.Following Aiken and West (1991), the predictor variables were mean-centered to minimize multicollinearity.Table 2 presents the results.IDFO and the former company's perceived prestige related positively to supportive behavior toward the former company in Models 2 and 3.These results provide support for H1 and H2.

Discuss
The high.These results suggest that conflicts between identities may arise in the behavioral dimension assumed in constructing the hypotheses.Moreover, the results confirm that the perceived prestige of the former company positively affects supportive behavior toward it.This relationship is especially true among people who strongly identified with former organizations.
Theoretical contributions of this study are threefold.First, this study demonstrates several antecedents of former employees' supportive behavior.The rise in employee mobility creates opportunities for former employees' supportive behavior to have notable values for former companies (Somaya & Williamson, 2008).However, recent research on the positive influence of employee mobility lacked investigation into the antecedents of such positive influence.This study illustrates the characteristics of former employees who tend to adopt supportive behavior.It also shows that companies that have good reputations can expect former employees to be supportive.
Second, the findings confirm Mael and Ashforth's (1992) prediction that IDFO fosters supportive behavior toward former employers.They also support the supposition that the concept of "alumni" is appropriate in business settings, as Mael and Ashforth (1992) suggested (Carnahan & Somaya, 2015;Somaya & Williamson, 2008).
Third, this study makes a unique contribution to the work-related identification literature.This study is the first attempt to address both IDFO and IDPO and examine their effects simultaneously.Focusing on not only identification with present organizations or groups but also identification with former organizations or former groups will be useful for deeper understandings of work-related identity and identifications.
Specifically, the findings dispute extant multiple identifications research that attributes positive correlations among identifications with different foci to identity integration.This study finds that IDFO and IDPO can exert opposite effects on supportive behavior toward former companies despite a positive correlation between IDFO and IDPO.These results do not appear congruent with identity integration and require other interpretations for the positive correlation.A possible explanation might be that positive correlation is partly derived from an individual's need for organizational identification (Glynn, 1998).Thus, the temporal extension of identification foci can bring novel research schemes for unraveling work-related identity and identification.After presenting the results, you are in a position to evaluate and interpret their implications, especially with respect to your original hypotheses.Here you will examine, interpret, and qualify the results and draw inferences and conclusions from them.Emphasize any theoretical or practical consequences of the results.(When the discussion is relatively brief and straightforward, some authors prefer to combine it with the Results section, creating a section called Results and Discussion).
The findings have several practical implications.First, this study raises a benefit of retention policies from a novel perspective.That is, an effective retention policy (cf.Mitchell, Holtom, & Lee, 2001) may be helpful for eliciting supportive behavior from employees who depart in the future.Many retention management practices can boost employees' identification with the organization, which will be strongly related to their identification with it after their departure (Carnahan & Somaya, 2015).On the other hand, it may be rather difficult for former employers to enhance former employees' identification with them after their departure.Therefore, measures that build employees' identification with their employers will not only decrease turnover in the short run (van Dick et al., 2004) but also contribute to invoking departed employees' supportive behavior in the long run.
Second, the findings can encourage operations of alumni relations programs for departed employees (Sertoglu & Berkowitch, 2002;Rousseau, 1998).Although rapid developments of information and communication technology have decreased costs for maintaining relationships with former employees, companies do not always adopt alumni relations programs (Hoffman et al., 2014).However, such programs remind former employees of their former companies, which might contribute to maintaining their identification with former employers.
Moreover, communicating with former employees and delivering the latest news to them may also help uphold their perceived reputation.The news can help former employees renew their knowledge of products or services that their former companies provide.Such a renewal might give them more confidence in their perceived prestige of the former companies.On the other hand, former employees who get less information on their former firm may be uncertain whether their perceived prestige of the former companies is still valid.
This study has limitations that suggest directions for future research.First, the data are cross-sectional.As with all cross-sectional survey data, conclusions about causality must be suspended.Future studies could employ a longitudinal research design to examine causality.
Second, this study is based on self-reported measures, which may introduce common method variance.It is appropriate that IDFO and IFPO are measured by a self-report method because organizational identification is a self-referential construct.However, with respect to supportive behavior, other methods could be explored in future research.
Third, other antecedents of supportive behavior toward former companies should be explored.For example, personal contacts or social networks with former colleagues who keep working for the same company may be promising ones (cf.Hoffman et al., 2014).Another possible variable is whether former and present companies are competitors.Severe competitive rivalry could discourage supportive behavior toward former companies.Furthermore, some variables may affect degrees of IDFO directly.An integrative model that includes both antecedents of IDFO and those of supportive behavior toward former companies should be explored.
Fourth, the effects that former employees may have on former companies should be examined more precisely.This study addressed general intentions to support former companies.Future studies that specify former employees' possible functions based on features of former companies' business can generate more practical implications.For example, in knowledge-intensive industries, focusing on knowledge exchange function of former employees (Corredoira & Rosenkopf, 2010) may lead to useful possibilities.In contrast, in service sectors where employee turnover is high, former employees' intentions toward re-employment may be more important (cf.Shipp, Furst-Holloway, Harris, & Rosen, 2014).
Finally, this study addressed only positive functions derived from former employees.However, they may have negative influences on former companies as well (cf.Aquino & Douglas, 2003).Future studies need to examine negative repercussions and their antecedents.
Figure 2 i between I relationshi company w company w Figure 3 p toward the more stron

Fi
and INFO can be addressed from the viewpoint of multiple identifications.Studies on multiple