Content uploaded by Zhi Huang
Author content
All content in this area was uploaded by Zhi Huang on Jan 04, 2019
Content may be subject to copyright.
rAcademy of Management Discoveries
2018, Vol. 4, No. 4, 410–428.
Online only
https://doi.org/10.5465/amd.2016.0141
MAKING OLD FRIENDS: UNDERSTANDING THE CAUSES
AND CONSEQUENCES OF MAINTAINING FORMER
COWORKER RELATIONSHIPS
IAN J. WALSH
1
University of Massachusetts Amherst
DANIEL S. HALGIN
ZHI HUANG
University of Kentucky
In this article, we seek to recast scholarly thinking about workplace relationships by
exploring their lasting value. Drawing on a two-wave survey study, we find that ties
with former coworkers have a positive relationship with individuals’social in-
tegration into new workplaces. We further explore why individuals maintain such
ties, focusing on the effects of being hired and employed at particular points in time.
Age-at-hire has an inverted curvilinear relationship with the number of relationships
one maintains, with a peak that corresponds with the midlife transition. We find that
length of tenure, and the extent to which it involves the midlife transition, is positively
related to the maintenance of ties to former coworkers. Organizational growth at the
time of hire has a positive effect on relationship maintenance and reduces the influ-
ence of age-at-hire. Change in profit margin during one’s tenure has a positive effect
on tie maintenance. Lastly, the positive effect of employment during the midlife
transition is enhanced by employment growth during one’s tenure. These findings
draw attention to the temporal dynamics of workplace relationships, providing
support for the notion that when ties form and take root shapes the extent to which
they survive.
Organizational research has consistently docu-
mented the importance of coworker relationships
during an individual’s tenure in a workplace (Blau,
1964; Burt, 2002). But what happens to these
relationships when employees leave an organiza-
tion? Although scholars have tended to argue that the
loss of a shared context would create substantial
obstacles that lead such relationships to atrophy
(Burt, 2001; Dahlander & McFarland, 2013), they
have also recognized the merit of ties to individuals
who “move in different social circles”(Granovetter,
1Corresponding author.
410
Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express
written permission. Users may print, download, or email articles for individual use only.
1983: 205). Such relationships can foster job en-
gagement (cf. Halbesleben, 2006) and provide access to
novel opportunities and information (Kleinbaum,
2012; Levin, Walter, & Murnighan, 2011). These ideas
suggest that ties to individuals who are associated with
one’s past are both difficult and important to maintain.
Furthermore, the prospect for relationships with for-
mer colleagues is an ever more salient characteristic of
organizational life as a result of the increasing mobility
of the contemporary workforce and the exposure to
rotating casts of coworkers it affords. The rise of
corporate alumni associations is a testament to the
growing salience of such ties (Bardon, Josserand, &
Villes`
eche, 2015; Walsh & Bartunek, 2011).
Despite the evidence that ties to former coworkers
are both consequential and ever more prevalent, re-
search on workplace relationships has been largely
confined to explaining the nature of ties among
current coworkers. To the extent that scholars have
considered the effects of relationships among former
coworkers, they have tended to emphasize the ben-
efits that stayers attain by remaining in touch with
leavers (Bartunek, Huang, & Walsh, 2008; Klein-
baum, 2012). But what impact do such ties have on
those who have exited? There has been little direct
consideration of their longer term effects on in-
dividuals who transition to new workplaces. We thus
believe that scholarly understandings of workplace
relationships would be more thorough if theories
attended to the effects and motives associated with
maintaining ties to former coworkers. Do relation-
ships with prior coworkers make a difference in in-
dividuals’current workplaces? And why do they
maintain such ties? We drew on the results of a two-
wave survey study of former employees of a technol-
ogy firm to find answers to these research questions.
In our initial examination of our data, we found
differences in relationship maintenance activity
between those individuals who reported being
employed in another organization at the time of
taking the survey and those who did not. This finding
suggested that relationships with former coworkers
might influence individuals’experiences in sub-
sequent work roles. It led us to investigate the re-
lationship between maintained ties and social
integration in new work contexts. While shedding
light on the impact of ties with former coworkers, our
findings make an important contribution to the lit-
erature by clarifying the theoretical intersection of
one’s relationships with past and present colleagues.
In addition, respondents were asked to explain in
their own words why they chose to take steps to
maintain ties with former colleagues. A thematic
analysis of the statements they made revealed
a number of motives that were broadly consistent
with the roles scholars have previously ascribed to
relationships, including social and instrumental
value (Blau, 1964), nostalgia (Bardon et al., 2015;
Iyer, Bamber, & Barefield, 1997), and legacy perpet-
uation (Walsh & Glynn, 2008). However, we were
struck by the frequency with which individuals
surfaced an unexpected theme. Specifically, re-
spondents often discussed their interest in former
coworkers as a function of the particular years of
their lives in which they worked at the organization,
which were likened to a formative period. For in-
stance, respondents referred to their employment as
“a time that represented many milestones for me,”“a
formative time in my career,”and as “an important
part of my life.”The formative character of their ten-
ures was often grounded in age-based explanations.
Respondents pointed to the “stage of life”at which
they had worked at the organization, likening it to
their “roots”and contrasting it with other periods,
such as high school or college. The lasting importance
of the ties was underscored through comments that
portrayed coworkers as “people who did so much to
shape me later in life.”Taken together, these com-
ments suggested that the timing of tie formation,
specifically the ages at which individuals form and
develop relationships, may play a role in how im-
portant they remain after exit. These ideas are con-
sistent with evidence that suggests an individual’sage
may play a role in fostering engagement in the work-
place (Avery, McKay, & Wilson, 2007). We thus set
out to specifically test the idea that age-at-hire, tenure,
and age during employment may impact one’sefforts
to maintain relationships with former coworkers.
In light of this preliminary evidence that an in-
dividual’s life stage during employment could be an
important determinant of tie maintenance, we also
wondered whether an organization’s life stage played
a role in an individual’s propensity to keep in touch with
former coworkers. Although this idea is broadly consis-
tent with the notion that the quality of relationships owes
in part to the nature of the context in which they are
formed (Kram, 1988), there has been little consideration
of the potentially formative influence of an organization’s
performance on the ties members form with one another.
Thus, we also sought to test the idea that being hired and
employed during periods of profitability and growth
would impact individuals’inclinations toward de-
veloping ties that become worthy of being maintained.
This study represents an initial step in developing
theory about maintenance of workplace relation-
ships in the absence of a shared context. Our research
Author’s voice:
What motivated you personally to
undertake this research? Why is it
important to you?
2018 411Walsh, Halgin, and Huang
provides insight regarding the reasons individuals
maintain tiesunder conditions, where scholars would
typically expect them to wither. We now turn to
explaining the ways in which research on workplace
relationships, employment and aging, and conser-
vation of resources shaped our ideas.
RELATIONSHIPS WITH FORMER COWORKERS
The Impact in Subsequent Workplaces
There is perhaps no greater challenge for an orga-
nizational newcomer than becoming socially in-
tegrated with one’s newly encountered colleagues. It
is thus not surprising that relationship-building ef-
forts are particularly prominent during socialization
and plateau after the first three months of employ-
ment (Song, Liu, Shi, & Wang, 2017). Although
socialization research has tended to explain re-
lationship formation in terms of characteristics of
a newcomer and/or the work environment (Morrison,
2002; Van Maanen & Schein, 1979), scholars have
found that individuals’behavior in newly joined
contexts may also be shaped in part by their prior
relational experiences (Kleinbaum, 2012). These
ideas suggest that the nature of an individual’s rela-
tionships with former coworkers may play a role in
their inclinations toward new coworkers.
As newcomers navigate their way through the
uncertainty of socialization (Van Maanen & Schein,
1979) maintained relationships may provide valu-
able social support (cf. Halbesleben, 2006), reservoirs
of mutual trust, and sources of novel information
(cf. Granovetter, 1973). Relationships with former
coworkers offer value even after extended periods
in which individuals have had no contact because of
the access they provide to different social circles and
novel insights (Levin et al.,2011). However, investing
time and staying in touch with former coworkers re-
duces one’s availability to build relationships with
new coworkers. Furthermore, devoting extensive
time to former coworkers may lead one’s newly en-
countered colleagues to conclude an individual is
disinterested or unavailable to develop relationships
with them. Thus, we sought to test whether ties to
former coworkers heighten or dampen one’ssocial
integration into a subsequent workplace.
The Impact of Age and Tenure
Individuals enter organizations at particular
points in their lives, and their work-related motives
can vary with age (Kanfer & Ackerman, 2004; Kooij,
de Lange, Jansen, Kanfer, & Dikkers, 2011; Ng &
Feldman, 2010). In the first years of their working
lives, individuals are likely to be relatively ambiva-
lent about developing close relationships in a newly
joined workplace. Young people experience great
uncertainty and often flounder, changing jobs fre-
quently without much clarity or direction (Arnett,
2015; Mortimer & Simmons, 1978). As they move
through the “trial period”of early adulthood (roughly
corresponding with one’s twenties and thirties)
and are presented with a greater number of career
opportunities, individuals arelikelytoexperience
an uptick in their work-based motives (Boumans,
de Jong, & Janssen, 2012), becoming increasingly
focused on demonstrating to those around them
that they have professionally established them-
selves (Mortimer & Simmons, 1978). Thus, as they
join new workplaces later into early adulthood, in-
dividuals could be expected to see greater merit in
forging strong relationships with their co\workers
(cf. Kooij et al., 2011). As a result, individuals pro-
gressing through early adulthood are increasingly
likely to embrace opportunities for socializing them-
selves in new workplaces they join (Arnett, 2015).
These ideas suggest that, all else equal, individuals
who are hired at later points in early adulthood are
more inclined to accrue more workplace ties than
their younger counterparts.
Individuals undergo a midlife transition as they
approach their forties that entails a “renegotiation”
of their life goals (Nakai, Chang, Snell, & Fluckinger,
2011) that leads them to gradually recast their de-
velopmental concerns (Kanfer & Ackerman, 2004).
As they progress through middle adulthood, in-
dividuals realize a more solid footing in their work
lives. Their concerns about establishing themselves
likely taper off as their aspirations converge around
a particular set of ideals (Levinson, 1986, 1996).
Middle-era adults increasingly come to focus on
exploiting resources and capabilities they have al-
ready developed and become steadily less con-
cerned with developing new ones (Kooij et al, 2011;
Ng & Feldman, 2012). As individuals age, they are
more inclined to focus on meaningful relationships
they have already developed (Ng & Feldman, 2010).
Thus, we expect interest in forming workplace re-
lationships of adults hired later in middle adulthood
to mirror that of those hired earlier in early adult-
hood, albeit for different reasons. In sum, the closer
newcomers are to the midlife transition, the more
likely they are to develop valuable workplace ties.
The aging and socialization literatures provide
less guidance regarding individuals’motives to
preserve and sustain such ties. Conservation of re-
sources theory (Hobfoll, 2001) suggests that in-
dividuals will strive to preserve coveted social
resources in situations that place them at risk and
doing so can foster effective adaptation (Dutton,
Roberts, & Bednar, 2010). As individuals exit orga-
nizations, their relationships with their erstwhile
412 DecemberAcademy of Management Discoveries
coworkers are at an elevated risk of dissolving due to
the loss of the shared social context that fostered
their development (Dahlander & McFarland, 2013;
Feld, 1997). The importance of such relationships is
amplified by the uncertainty individuals experience
after exit (Ibarra, 1999; Kleinbaum, 2012). The degree
to which individuals have cultivated valuable re-
lationships will likely shape their diligence about
maintaining them. By contrast, those whose ties are
largely inconsequential or served strictly instru-
mental purposes in the workplace may simply let
them dwindle away (cf. Burt, 2001).
The value of relationships develops through
repeated interactions over time (cf. Dahlander &
McFarland, 2013). Thus, we suspected that ten-
ure in an organization will be positively related to
the maintenance of relationships. However, indi-
viduals’access to valuable relationships is not uni-
form throughout one’s career. For instance, Burt
(2016) found that managers have the greatest access
to advantageous relationships in the middle years
of their careers.2This finding led us to expect that
the extent to which an individual’s employment en-
compasses the midlife transition will be positively
related to one’s maintenance of relationships with
former coworkers.
The Impact of Organizational Performance
and Growth
Given the formative influence of contextual char-
acteristics on relationships (Kram, 1988), we further
considered how organizational conditions at the
time of hire and during employment could play a role
in individuals’maintenance of ties with former co-
workers. Amidst evidence that an organization is
highly profitable and/or experiencing high growth,
members are particularly likely to hold it in high
regard. To the extent that they attribute an organi-
zation’s success to their newly encountered co-
workers, individuals may be prone to see ties with
them as important to develop and sustain (cf. Schein,
1985). Thus, we expect that the extent to which in-
dividuals’tenures are marked by organizational
profitability and growth will positively impact their
maintenance of ties with former coworkers by con-
ditioning the perceived value of their ties. More
specifically, our awareness of the heightened sus-
ceptibility of newcomers to imprinting (Marquis &
Tilcsik, 2013) suggests that the level of organiza-
tional performance and growth at the time of hire
would condition their relational inclinations. Thus,
we also sought to test the idea that profitability and
growth at the time of hire has a positive effect on tie
maintenance.
We expect that concerns about relationship main-
tenance are likely to stem from a confluence of
the individual attributes and organizational char-
acteristicswehaveidentified.Thevisibletrappings
of performance and growth, such as the expansion of
facilities or public acclamations by leaders, are likely
to induce a consensus among members that an or-
ganization is successful. Such widespread agree-
ment is broadly consistent with a “strong situation”
(Mischel, 1977) that could induce newcomers to
more consistently invest themselvesin ties regardless
of their ages. These ideas are consistent with the no-
tion that organizations are capable of providing
a context that affords a sense of belonging and being
an “insider”to older members (Armstrong-Stassen &
Schlosser, 2011), who may otherwise show relatively
low levels of interest in forging and eventually
maintaining ties with their coworkers. Thus, we ex-
pected that the effect of age-at-hire will be reduced by
higher levels of profitability and growth at the time of
hire. Strong situations also provide more coherent
shared work climates (Haggerty & Wright, 2009),
which we expect would enhance the quality of the
interactions that give rise to workplace relationships.
Thus, we further expected that experiences of suc-
cess in organizational contexts would amplify the
effects of tenure and age during employment on tie
maintenance.
METHODS
Data
Our analysis is based on two rounds of surveys of
former employees of Digital Equipment Corporation
(DEC), a technology firm that was dissolved follow-
ing its acquisition in 1998. Given that it is defunct,
DEC represents a useful setting for exploring individ-
uals’propensities toward relationships with former
coworkers and their value in subsequent workplaces.
When an organization is active, relationships with
former coworkers could owe to other causes, such as
its explicit attempts to keep in contact with former
members (Iyer et al., 1997) or individuals’aspirations
for re-employment. Moreover, our theoretical interest
in the maintenance of relationships without the sup-
port of a shared social context can be more credibly
studied when the context has dissolved.
Author’s voice:
How did you get access to your
data or site?
2Burt (2016) also finds that the actual returns to network
advantage vary over the life course and the timing and
trajectory of the “peak”is influenced by the broader social
norms of the organization.
2018 413Walsh, Halgin, and Huang
All former employees of DEC were eligible to
complete our study. All respondents stopped being
members of DEC at least 10 years before our study
began, providing a particularly long time horizon for
exploring individuals’attunement to relationships
with former coworkers. Unlike members of intact
settings, former employees cannot be readily iden-
tified and enumerated within the broader social
context, and constructing a true sampling frame in
such cases is impractical. In studying “hidden pop-
ulations,”Salganik and Heckathorn (2004) recom-
mended the use of respondent-driven sampling
procedures, which rely on “seeds,”or known mem-
bers to complete and distribute a survey to others not
known to the researchers. We initiated our contact
with former members through alumni associations
for former employees of DEC. Participants in the
survey were encouraged to share the survey an-
nouncement, which was also posted on the websites
of alumni groups. Approximately 35 percent of re-
spondents reported being alumni group members,
suggesting that the surveys were broadly distributed
beyond the initial seeds. We collected complete re-
sponses from 1,678 individuals through the first
survey. For our second survey, which was conducted
6 years later, we contacted all original respondents
for whom we had current contact information and
received 704 responses, of which 539 were complete.
The individuals whose responses were complete did
not differ significantly from those with incomplete
responses in terms of demographics (e.g., age, gen-
der, education) or work-related characteristics (e.g.,
hierarchical position, functional disciplines).
Three possible concerns can be raised about our
data collection method. First, reliance on chain-
referral methods may result in a sample that over-
represents individuals with a large number of
relationships and underrepresents individuals with
few or no relationships (Salganik & Heckathorn,
2004). However, the distribution of the number of
reported relationships does not seem to suggest such
bias. When these data are plotted in a histogram, the
highest density area pertains to respondents with no
relationships with former coworkers (27 percent of
the sample). This finding likely owes to the posting of
the survey announcement on a number of industry
message boards and the ease with which individuals
can join alumni associations without interacting
with any former colleagues. Thus, it does not appear
that our sample underrepresents unconnected in-
dividuals. In addition, less than 14 percent of re-
spondents reported more than nine ties, suggesting
that our sample does not overrepresent those with
many ties. Second, our sampling approach could
raise questions about bias toward alumni group
members, who might be expected to systematically
have more relationships with former coworkers.
Therefore, we compared the mean number of
friendships that alumni group members, former
group members, and nonmembers reported they had
originally formed at DEC and were now maintaining
(3.91, 3.81, and 3.75, respectively). This analysis
suggestedthere were no significant differences among
the three subgroups. Although our reliance on mem-
bers and former members of alumni groups as seeds
could introduce some bias, it would be more likely to
result in a regression line having a higher intercept
than in one with a materially different slope. Third, it
could also be argued that our reliance on online sur-
veys may introduce a bias against older individuals,
who may be less inclined to have email or Internet
access. However, the distribution of ages in our sam-
ple ranged from 30 to 81 and follows an almost per-
fectly normal distribution. Given that there is no
reason to expect former employees of DEC would
follow an atypical distribution, we believe age-based
bias is unlikely to have unduly impacted our analysis.
Measures
Dependent variables. To measure social in-
tegration in new workplaces, we asked individuals to
consider the workplaces they joined immediately
after leaving DEC and presented them with an
established seven-item scale that measures social
integration (Morrison, 2002). All items for this scale
have loadings greater than 0.7 on a single factor (a5
0.89).3To measure maintained relationships with
former coworkers, we used a conventional name
generator (Marsden, 1990), asking respondents to list
up to 12 former members with whom they were
presently in contact.4We asked them to indicate the
frequency of contact and eliminated anyone with
whom they were not in contact on at least an annual
basis to focus our inquiry on active relationships and
to exclude ties that are dormant (Levin et al., 2011).
We also asked respondents to note whether the re-
lationship with each contact was formed prior, dur-
ing, or after their employment at DEC, and to classify
3We measure relationship maintenance in the present
and social integration at an earlier point in time. This
seems against the basic rule of causality. However, it is
likely the reported relationships are the same or less than
those that existed at the time of exit (cf. Burt, 2002). Model 2
is thus likely to be a conservative test.
4This cap could truncate our dependent variable and
bias our estimation if many respondents maintain more
than 12 relationships. However, our sample contains only
68 responses with 12 active contacts (4 percent of the
sample). Our results were not materially different when we
excluded these responses (analysis available on request).
414 DecemberAcademy of Management Discoveries
it into one of the following categories: family, per-
sonal friend, client/customer, business contact, and
current coworker. Given that our posited arguments
pertain to workplace relationships, we excluded any
ties that were formed before or after one’s tenure at
DEC. In light of our interest in understanding the
survival of relationships in the absence of a shared
social context, we only counted former coworkers
who were identified as personal friends because
those associated with the other categories were po-
tentially anchored in another environment, such as
one’s home or current place of work. Because re-
spondents had to specify the nature of their current
relationships with individuals they listed, our measure
should not merely reflect names of individuals that
respondents could recall from their time as members.
Independent variables. We calculated age-at-
hire using the date respondents reported joining DEC
and their reported age at the time of the first survey.
To measure tenure, we calculated the number of
years an individual worked at DEC based on their
reported hire and exit dates, for up to three distinct
periods of employment. Based on scholarly argu-
ments that the midlife transition represents a multi-
year process that tends to commence as individuals
approach their forties (cf. Levinson, 1996), we cre-
ated a measure of employment during midlife tran-
sition, which indicates the number of years between
the ages of 35 and 45 each individual was employed
at DEC. We also counted the number of pre-midlife
transition years and post-midlife transition years
each individual worked at DEC (i.e., the number
of years employed at DEC when under the age of 35
and over the age of 45, respectively) and used these
as controls.
To assess DEC’s performance at the time of hire
and during employment, we gathered electronic or
paper copies of DEC’s annual reports, including
those that were published before it was a public firm.
These reports provided a host of metrics related to
DEC’s performance, including net profit, revenue,
and total employment. We used these data to calcu-
late the organization’sprofit margin at hire (i.e., the
year an individual was hired), and change in profit
margin over tenure (i.e., the difference in profit
margin between one’s final and first year of em-
ployment at DEC). We also captured employment
growth at hire, measured as the percentage change in
the number of employees in the year of hire over the
prior year and used the same data to calculate em-
ployment change over tenure (i.e., the difference in
employment between one’s final year and first year
of employment at DEC).
Control variables. In all of our analyses, we con-
trolled for gender and education. The latter measure
is a dichotomous variable that indicates whether
a respondent reported holding a bachelor’s degree or
above. In our exploration of the effect of maintained
ties on social integration, we controlled for age at exit
from DEC. Based on our interest in a respondent’s
experience in one’s subsequent workplace, this
measure should closely approximate the age of
joining it. To account for possible occupational ef-
fects, we used dummy variables capturing re-
spondents’reported area of specialization (e.g.,
marketing, finance, engineering). We also included
measures for two personality traits that could incline
an individual toward becoming socially integrated:
need for affiliation (Steers & Braunstein, 1976); and
self-monitoring (Snyder & Gangestad, 1986), which
encompasses three factors: acting ability, extraver-
sion, and other directedness (Lennox & Wolfe, 1984).
Finally, we controlled for the number of other types
of maintained relationships (other contacts) that re-
spondents had reported (e.g., business contacts,
family members) because such relationships likely
provide task and social support that are conducive
to becoming socially integrated (cf. Halbesleben,
2006). Preexisting connections with other former
DEC employees in a newly joined context may
provide a signal of one’sworthinessasarelational
partner, thereby, facilitating the process of social
integration.
In testing our ideas about the factors underlying
relationship maintenance, we controlled for some
individual characteristics that could systematically
influence the number of relationships one has with
former coworkers. Current age, as directly reported
by respondents, helps to isolate the effect of age-at-
hire from possible generational effects because it is
an individual’s current age that corresponds to one’s
generation. We accounted for the number of alumni
group affiliations each individual claimed because
such groups could be providing a shared context to
facilitate relationships (Bardon et al., 2015). In ad-
dition, we controlled for whether individuals were
seeking a job,currently unemployed, or working as
an entrepreneur because these factors could sys-
tematically prime individuals to maintain ties with
former coworkers. We also controlled for three other
work-related factors that could influence the number
of relationships one has with former coworkers: the
number of different functional areas in which one
worked, work location (USA 51), and length of
tenure with DEC’s acquirer.
We recognized the possibility that DEC’s dis-
tinctive culture (Schein, 2004) could produce idio-
syncratic effects that are less applicable to other
organizational contexts. Specifically, DEC’s reputa-
tion for adhering to the principles of an internal labor
market might systematically influence the number of
relationships its members developedand maintained.
2018 415Walsh, Halgin, and Huang
For instance, it is possible that newcomers to DEC
were particularly inclined to have a sizable number
of pre-entry ties with existing employees, which
would provide a natural advantage for forming addi-
tional valuable relationships. To partly account
for this possibility, we controlled for the number
of maintained relationships individuals reported
forming before working at DEC (pre-entry relation-
ships). Given that those working at higher echelons of
an organization may be more capable of taking ad-
vantage of internal labor market practices, we also
controlled for an individual’shighest hierarchical
position.
Statistical Analysis
To explore the impact of maintained ties on social
integration, we used an ordinary least squares re-
gression model with robust standard errors. To in-
vestigate factors that influence the maintenance of
ties, we used a Poisson regression model because the
survey data are cross-sectional, and the dependent
variable is a count measure. The insignificant results
of a chi-squared (x2) test suggest the dependent variable
does not exhibit serious overdispersion. Because our
analyses entail the testing of quadratic and interaction
effects, the potential for multicollinearity among our
predictor variables could mask the true effects of in-
dividual measures. However, Table 1 indicates the
included variables have relatively low correlations.
We also mean-centered our measure of age-at-hire
before forming its squared term. All variables in the
models reported in our article have variance in-
flation factor (VIF) scores ranging from 1.08 to
9.55 with the average VIF scores for all models
only hovering around two, suggesting that multi-
collinearity has not unduly influenced our results.
RESULTS
Table 2 presents the results of our analysis re-
garding the value of maintained ties in subsequent
workplaces. Model 1 presents only controls and
Model 2 includes our measures of maintained
friendships and the other types of relationships. It
indicates that the number of maintained friendships
with former coworkers had a significant positive ef-
fect on social integration into a subsequent work-
place, providing evidence of the value of such ties.
Table 3 presents analyses that test our ideas about
the factors underlying individuals’maintenance of
relationships with former coworkers. Model 3 only
includes our control variables. Model 4 introduces
a linear measure of age-at-hire, and Model 5 includes
a square term to test our ideas about its potential
curvilinear effect. Model 5 indicates that the esti-
mated coefficients of age-at-hire and its squared term
are significant with expected signs as illustrated
graphically in Figure 1, thus, strongly indicating
a curvilinear relationship between age-at-hire and
the number of maintained relationships. The peak
corresponds to a raw value of roughly 38. This value
closely corresponds with the transition from early to
middle adulthood (Levinson, 1986), lending cre-
dence to our posited ideas.
Our findings also provide support for our ideas
regarding the impact of organizational tenure. Re-
sults from Model 4 indicate a positive and significant
effect of tenure on the number of relationships one
maintains. Building on this analysis, Model 6 tests
our ideas about age at employment. It replaces tenure
with variables that measure the number of midlife
transition years an individual was employed at DEC
as well as the number of pre-midlife and post-midlife
transition years employed. As predicted, when
controlling for the latter two variables, the coefficient
for midlife transition years is positive and significant.
It is worth noting that the coefficient for pre-midlife
transition years is also positive and significant.
However, a post hoc Wald test indicates that the effect
of midlife transition years was marginally greater than
that of pre-midlife transition years (x253.81, p5
.06). It appears that working at an organization during
the midlife transition has a pronounced impact on tie
maintenance.
We also examined the effects of organizational
conditions at the time of hire and over the course of
one’s employment on tie maintenance. As shown in
Model 4, the estimated coefficient for employment
growth at hire is positive and marginally statistically
significant, suggesting that individuals who joined
the organization as its workforce was expanding
more rapidly went on to maintain a greater number of
relationships than others. We note that there is no
significant effect for change in employment growth
over tenure. In addition, we note that the coefficient
for profit margin at hire is not significant. However,
the coefficient for change in profit margin over one’s
tenure is positive and significant, suggesting that
individuals with tenures coinciding with periods of
financial growth maintained a greater number of ties
than others.
We also tested the moderating effect of organiza-
tional performance factors. Model 7 in Table 4 ex-
plores the effects of higher levels of employment
growth at the time of hire on maintained relation-
ships. Specifically, we found that employment
Author’s voice:
Was there anything that surprised you
about the findings? If so, what?
416 DecemberAcademy of Management Discoveries
TABLE 1
Descriptive Statistics and Pairwise Correlations
Mean SD 1 2 3 456789101112
1 Maintained relationships (friendships) 3.88 3.63
2 Age 55.67 7.57 0.05
3 Gender (male 51) 0.75 0.43 20.14 0.19
4 College (bachelor or above 51) 0.70 0.46 20.02 0.04 0.08
5 Alumni group affiliations 0.60 0.62 0.05 0.03 0.07 0.001
6 Entrepreneur 0.001 0.07 0.001 0.01 0.001 20.01 20.02
7 Currently unemployed 0.03 0.17 0.001 0.001 20.01 20.01 0.02 0.30
8 Seeking a job 0.04 0.19 0.02 0.001 0.01 20.01 0.02 20.01 0.74
9 Functional areas 1.13 0.45 0.001 0.07 0.001 20.03 0.01 20.02 20.03 20.01
10 Highest position 2.16 1.35 0.10 0.33 0.16 0.16 0.13 20.03 20.02 20.02 0.09
11 Work location (USA 51) 0.60 0.49 0.09 0.24 20.09 0.08 20.24 0.04 0.07 0.07 20.01 20.09
12 Tenure with DEC’s acquirer 1.86 3.23 0.06 20.11 20.02 20.05 20.08 20.01 0.001 0.01 20.01 20.13 0.05
13 Need for affiliation 1.58 0.6 0.13 20.03 0.01 0.04 0.06 20.02 0.02 0.01 0.05 0.11 20.09 0.02
14 Self-monitoring 28.49 3.37 0.06 0.09 20.03 0.04 20.05 0.01 20.04 20.05 20.01 20.04 20.03 0.001
15 Current distance to work location 2,086.06 2,626.09 20.05 20.17 0.08 20.08 0.13 20.03 20.03 20.05 0.04 0.07 20.68 20.07
16 Interconnectedness of relationships 3.41 0.82 0.15 20.03 0.04 20.01 0.07 20.04 20.08 20.04 0.04 0.04 20.06 0.02
17 Size of social network 3.78 0.71 0.27 20.08 20.04 20.01 0.07 20.01 20.03 0.001 20.02 0.13 0.05 0.09
18 Pre-entry ties 0.41 0.82 0.16 0.04 0.02 0.001 0.01 0.08 20.02 20.02 20.03 0.001 0.01 0.02
19 Age-at-hire 28.31 6.41 20.05 0.59 0.13 0.12 0.04 20.01 20.04 20.03 20.02 0.19 20.02 20.07
20 Employment growth at hire 0.19 0.18 0.11 0.43 0.09 20.04 0.01 0.001 0.01 0.02 0.10 0.14 0.25 20.02
21 Profit margin at hire 0.09 0.04 0.05 0.21 20.03 0.01 0.001 0.02 0.03 0.03 0.001 0.05 0.15 0.001
22 Organizational tenure 14.3 6.46 0.18 0.42 0.05 20.09 0.001 0.01 0.05 0.06 0.06 0.16 0.26 0.18
23 Change in employment over tenure 20.23 0.19 20.10 20.36 20.08 0.03 20.02 0.01 20.03 20.03 20.08 20.12 20.24 0.10
24 Change in profit margin over tenure 20.95 1.09 0.05 20.02 20.01 0.02 0.001 0.01 0.02 0.03 0.02 0.02 20.03 0.19
25 Midlife transition 6.30 4.53 0.13 0.33 0.06 0.03 0.02 0.01 0.02 0.04 0.03 0.12 0.12 0.36
26 Pre-midlife transition 6.76 5.44 0.09 0.04 20.02 20.07 0.001 0.01 0.08 0.08 0.02 0.04 0.12 20.17
27 Post-midlife transition 2.56 4.43 0.08 0.57 0.13 0.01 0.03 0.02 0.01 0.03 0.05 0.18 0.16 0.13
28 Social integration 3.88 0.66 0.10 20.03 0.07 0.07 0.10 0.001 0.05 0.02 20.02 0.03 0.11 0.05
29 Age at exit from DEC 42.08 10.29 0.11 0.61 0.14 0.04 0.04 0.02 0.02 0.05 0.03 0.21 0.16 0.29
30 Other contacts 1.47 2.75 20.27 20.02 20.06 0.02 0.07 0.02 0.01 0.04 20.02 0.07 20.06 20.01
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
14 Self-monitoring 20.04
15 Current distance to work location 0.02 0.05
16 Interconnectedness of relationships 0.07 20.03 0.03
17 Size of social network 0.19 20.10 20.01 0.38
18 Pre-entry ties 0.07 20.03 20.01 20.05 0.02
19 Age-at-hire 20.03 20.02 0.06 20.03 20.17 0.09
20 Employment growth at hire 0.001 0.06 20.2 0.02 0.07 20.03 20.19
21 Profit margin at hire 0.02 0.08 20.17 0.05 0.08 20.01 20.10 0.34
22 Organizational tenure 0.01 0.12 20.24 0.05 0.14 20.05 20.29 0.56 0.28
23 Employment change over tenure 20.02 20.05 0.17 20.04 20.09 0.04 0.15 20.88 20.30 20.50
24 Change in profit margin over tenure 20.04 0.05 0.01 20.02 20.01 0.02 20.03 0.03 0.001 0.04 0.18
25 Midlife transition 0.05 0.13 20.12 20.01 0.08 20.01 0.08 0.26 0.13 0.55 20.27 0.05
26 Pre-midlife transition 0.02 0.04 20.12 0.04 0.08 0.01 20.21 0.21 0.11 0.32 20.27 20.12 0.08
27 Post-midlife transition 0.001 0.07 20.13 0.06 0.02 0.04 0.36 0.23 0.11 0.42 20.21 0.06 0.39 0.28
28 Social integration 0.12 20.15 20.18 0.09 0.20 20.02 20.07 0.07 0.01 0.06 20.05 0.01 0.02 20.02 0.01
29 Age at exit from DEC 0.03 0.11 20.13 0.04 0.03 0.03 0.44 0.24 0.11 0.48 20.26 0.001 0.79 0.37 0.78 20.01
30 Other contacts 0.04 20.06 0.03 0.04 0.16 0.09 0.001 20.03 0.02 0.01 0.02 20.01 0.01 0.03 20.02 0.06 0.01
Correlation coefficients greater than 0.04 are statistically significant at p,.05.
2018 417Walsh, Halgin, and Huang
growth rate negatively moderated the linear term
(which reflects the experience of those hired in early
adulthood) and positively moderated the quadratic
term (which reflects the experience of those hired in
middle adulthood and beyond). Although the results
only offered marginal statistical support (p5.05),
they suggest that the curvilinear effect of age-at-hire
on maintained ties is dampened by higher levels of
organizational growth at the time of hire. Figure 2 plots
the relationship at one standard deviation greater than
and below the mean organizational growth value. We
also explored whether profit margin at hire moderated
the effect of age-at-hire as presented in Model 8, but
the results were not statistically significant.
Models 9 and 10 in Table 4 explore whether the
effect of employment growth and higher levels
of profitability during one’s tenure moderated the
effect of organizational tenure. We find no signifi-
cant relationships. However, when splitting or-
ganizational tenure into pre-midlife transition,
midlife transition, and post-midlife transition, we
find more promising results. Although we find no
interactions with profitability over tenure (Model
11), we find that the positive impact of employ-
ment during the midlife transition was enhanced
by employment during growth periods (Model 12).
Individuals whose tenures coincided with the midlife
transition and were marked by workforce growth
maintained even more relationships than others.
Post Hoc Tests of Alternative Explanations
To explore the statistical sensitivity of our results,
specifically the extent to which the effects of our core
age-related variables are contingent on the inclusion
of controls, we also ran simplified models. Results of
models including only the posited independent
variables, were largely consistent with those that
include the control variables. To address three other
possible alternative explanations of maintained ties,
we conducted post-hoc tests using data gathered in
the second survey. First, we sought to investigate
whether our findings were merely an artifact of the
total number of relationships individuals originally
formed while they were members. Thus, using the
results of our second survey, we also conducted
an additional test about age-related effects that
attempted to control for the number and intercon-
nectedness of individuals’relationships during the
time of their employment. The lapse of time would
make it difficult for individuals to recall the exact
number of ties they had, so we instead developed
TABLE 2
Ordinary Least Squares Regression Predicting Social Integration into Subsequent Workplace
m1 m2
Maintained relationships (friends) 0.029** (0.01)
Other contacts 0.025* (0.01)
Need for affiliation 0.111* (0.05) 0.086 (0.05)
Self-monitoring 20.030** (0.01) 20.029** (0.01)
Male 0.025 (0.08) 0.043 (0.08)
College (bachelor or above 51) 0.137* (0.07) 0.117
1
(0.07)
Age at exit from DEC 0.001 (0.001) 20.001 (0.001)
Highest position (second level) 0.18 (0.11) 0.145 (0.11)
Highest position (Third level) 0.125
1
(0.07) 0.099 (0.07)
Highest position (Fourth level) 0.009 (0.09) 20.006 (0.09)
Highest position (Fifth level) 0.14 (0.12) 0.091 (0.12)
Administration 20.145 (0.15) 20.133 (0.15)
Finance 20.115 (0.15) 20.071 (0.15)
Human resources 20.011 (0.14) 20.035 (0.13)
Information technology 20.045 (0.1) 20.036 (0.1)
Customer service 0.103 (0.09) 0.125 (0.08)
Engineering 0.036 (0.10 0.069 (0.10)
Manufacturing 0.094 (0.12) 0.067 (0.11)
Sales 20.03 (0.09) 0.015 (0.09)
Marketing 20.105 (0.1) 20.082 (0.1)
Consulting 0.269 (0.25) 0.285 (0.24)
Other function 0.054 (0.15) 0.082 (0.15)
Constant 4.382** (0.28) 4.315** (0.28)
Rsquared 0.08 0.10
Note: Robust standard errors are in parentheses. Two-tailed tests for all variables. N5498. Note that the Nis reduced from 539 because
respondents who did work after leaving DEC were not included in this analysis.
1p,.10
*p,.05
**p,.01
418 DecemberAcademy of Management Discoveries
a six-item scale (a50.84), aboutthe size of their social
networks during their employment.5The items highly
load on one factor with loadings greater than 0.6. We
also adopted a visual scale item developed by Mehra
et al. (2014) to measure the interconnections among
their ties. The item contains visual representations of
five network structures with varying degrees of in-
terconnectedness of relationships among contacts.
Second, we wanted to explore whether mainte-
nance of ties was actually a function of present-day
geographical proximity with former coworkers, so
we also controlled for each individual’s current
distance from their primary work location while
employed at DEC. Including this measure allowed us
to account for the possibility that those who remain
physically near their former workplaces may be
particularly inclined to share other social contexts
(such as local civic associations or sporting leagues)
with their former coworkers that provide a basis for
their ongoing relationships. Third, we wondered
whether personality characteristics could be driving
relationships with former coworkers, so as in the
prior test of our ideas about social integration, we
also controlled for self-monitoring and need for af-
filiation. Because a smaller number of individuals
responded to the second survey, we focused on
investigating the main effects posited in our core
analysis.6In Table 5, Model 13 shows that the per-
sonality and social network controls are posi-
tively and significantly related to the number of
relationships one has with former coworkers. Even
with these controls and a much smaller number of
observations, Model 14 supports the effects of tenure
TABLE 3
Poisson Models Predicting the Number of Relationships with Former Coworkers
m3 m4 m5 m6
Age-at-hire 0.020** (0.001) 0.024** (0.001) 20.006 (0.001)
Employment growth at hire 0.346
1
(0.20) 0.357
1
(0.20) 0.173 (0.20)
Profit margin at hire 20.041 (0.36) 20.131 (0.37) 20.052 (0.36)
Organizational tenure 0.034** (0.001) 0.035** (0.001)
Employment change over tenure 20.001 (0.17) 0.001 (0.17) 20.121 (0.17)
Change in profit margin over tenure 0.031* (0.01) 0.032* (0.01) 0.040** (0.01)
Age-at-hire squared 20.001** (20.001)
Midlife transition 0.017** (0.001)
Pre-midlife transition 0.008** (0.001)
Post-midlife transition 0.002 (0.001)
Age 0.002 (0.001) 20.023** (0.001) 20.024** (0.001) 20.001 (0.001)
Male 20.329** (0.03) 20.332** (0.03) 20.338** (0.03) 20.333** (0.03)
College (bachelor or above 51) 20.068* (0.03) 20.03 (0.03) 20.048
1
(0.03) 20.044 (0.03)
Alumni group affiliations 0.090** (0.02) 0.075** (0.02) 0.072** (0.02) 0.077** (0.02)
Entrepreneur 20.014 (0.20) 20.028 (0.20) 20.075 (0.20) 20.054 (0.20)
Currently unemployed 20.203
1
(0.12) 20.195 (0.12) 20.179 (0.12) 20.206
1
(0.12)
Seeking a job 0.234* (0.09) 0.196* (0.09) 0.194* (0.09) 0.193* (0.09)
Functional areas 0.001 (0.03) 20.008 (0.03) 20.006 (0.03) 20.016 (0.03)
Highest position (Second level) 0.088
1
(0.05) 0.088
1
(0.05) 0.087
1
(0.05) 0.088
1
(0.05)
Highest position (Third level) 0.285** (0.03) 0.255** (0.03) 0.253** (0.03) 0.265** (0.03)
Highest position (Fourth level) 0.261** (0.04) 0.239** (0.04) 0.239** (0.04) 0.240** (0.04)
Highest position (Fifth level) 0.321** (0.05) 0.266** (0.05) 0.273** (0.05) 0.288** (0.05)
Work location (USA 51) 0.183** (0.03) 0.120** (0.03) 0.129** (0.03) 0.127** (0.03)
Tenure with DEC’s acquirer 0.020** (0.001) 0.002 (0.001) 0.002 (0.001) 0.010* (0.001)
Pre-entry ties 0.146** (0.01) 0.155** (0.01) 0.159** (0.01) 0.150** (0.01)
Constant 1.135** (0.10) 1.489** (0.12) 2.136** (0.21) 1.323** (0.14)
Log likelihood 25,097.41 25,019.89 25,012.47 25,043.70
Note: Standard errors are in parentheses. Two-tailed tests for all variables. N51,678.
1p,.10
*p,.05
**p,.01
5The six items were: I had a large number of friends who
also worked at the company; I was highly intertwined with
the social network of DEC employees; I was highly engaged
with other employees at the company; I only had a small
number of close contacts who worked at the company; I
regularly organized social activities with other employees;
I attended social events with other employees.
6Those who responded to the second survey are more
likely to have relationships with former colleagues,
thereby, limiting the variance in the dependent variable
and making it more difficult to find supportive evidence.
Thus, Models 13–16 constitute a conservative test of our
theory.
2018 419Walsh, Halgin, and Huang
and organizational growth at the time of hire and
Model 15 provides continued evidence of the cur-
vilinear effect of age-at-hire. Model 16 indicates that
the measure of midlife transition years is marginally
significant in this limited sample.
DISCUSSION AND CONCLUSION
This study recasts scholarly understanding of the
potential staying power of workplace relationships
in the absence of a shared social context, which has
long been considered a crucial and necessary anchor
for sustaining them (Dahlander & McFarland, 2013;
Feld, 1997). With few exceptions, studies of work-
place relationships have focused exclusively on the
experience of active organizational members, im-
plicitly assuming that relationships are likely to de-
grade as individuals exit and lose the structural
embeddedness and shared context afforded by an
organizational environment. Indeed, prior research
seems to confirm this perception by showing that
relationships decay significantly within the first year
following an individual’s departure from a shared
context (Burt, 2001). Against this backdrop, this
study provides the first systematic examination of
the potential “stickiness”of relationships with for-
mer coworkers. Thus, this article provides greater
clarity about the potential for relationships to sur-
vive without structural supports and to even outlast
the organizations in which they were created.
This expansion in scope is in no way trivial; re-
lationships with former coworkers offered material
advantages in newly joined contexts. Our findings
suggest that such ties are not just vehicles for nos-
talgia but instead enable individuals’functioning as
they become acclimated to new work roles (Burt,
2002). Although the socialization literature has long
recognized the importance of bonds with co-
workers in reducing uncertainties encountered by
newcomers (Morrison, 2002; Van Maanen & Schein,
1979), our findings suggest that relationships with
former coworkers may also play a formative role in
enabling individuals to navigate their way through
newly encountered social structures. Having a solid
foundation of close contacts from other contexts may
provide crucial support for individuals through the
access to diverse information and emotional support
they provide (Gershon, 2017). This finding is con-
sistent with the notion that oscillating between deep
engagement with a group of new colleagues and with
former colleagues may positively influence a new-
comer’s performance in a new work setting (cf. Burt
& Merluzzi, 2016). Rather than positioning ties with
former and current coworkers as a “zero sum game,”
it demonstrates how ties from one’s past serve as im-
portant catalysts in the creation of new relationships.
Our findings draw attention to the importance of
the timing of relationship formation and supports the
notion that when ties are formed and developed may
play a role in their capacity to last. The consistent
FIGURE 1
Graph of the Estimated Main Effect of Age-at-Hire
–14 –12 –10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16 18 20 22 24 26
Age-at-hire (mean centered)
54.543.532.5
Predicted number of relationships
420 DecemberAcademy of Management Discoveries
TABLE 4
Poisson Models Predicting the Number of Relationships with Former Coworkers
m7 m8 m9 m10 m11 m12
Age-at-hire
a
0.024** (0.001) 0.024** (0.001) 0.021** (0.001) 0.021** (0.001) 20.006
1
(0.001) 20.006
1
(0.001)
Employment growth at hire
a
0.202 (0.21) 0.354
1
(0.2) 0.396
1
(0.2) 0.385
1
(0.2) 0.172 (0.2) 0.274 (0.2)
Profit margin at hire
a
20.13 (0.37) 20.56 (0.46) 20.209 (0.39) 20.038 (0.36) 20.037 (0.39) 20.207 (0.36)
Organizational tenure
a
0.034** (0.001) 0.035** (0.001) 0.041** (0.001) 0.035** (0.001)
Employment change over tenure
a
20.021 (0.17) 20.007 (0.17) 20.027 (0.17) 0.039 (0.17) 20.121 (0.17) 20.209 (0.17)
Change in profit margin over tenure
a
0.034* (0.01) 0.033* (0.01) 0.031* (0.01) 0.024 (0.02) 0.043** (0.01) 0.039** (0.02)
Age-at-hire squared 20.001** (20.001) 20.001** (20.001)
Midlife transition 0.017** (0.001) 0.034** (0.001)
Pre-midlife transition 0.008** (0.001) 0.007** (0.001)
Post-midlife transition 0.002 (0.001) 0.004 (0.001)
Age-at-hire 3Employment growth at hire 20.022
1
(0.01)
Age-at-hire squared 3Employment growth at hire 0.003
1
(20.001)
Age-at-hire 3Profit margin at hire 0.035 (0.06)
Age-at-hire squared 3Profit margin at hire 0.004 (0.01)
Tenure 3Employment change over tenure 0.012 (0.01)
Tenure 3Change in profit margin over tenure 20.002 (0.001)
Midlife 3Change in profit margin over tenure 20.002 (0.001)
Midlife 3Employment change over tenure 0.039* (0.02)
Age 20.023** (0.001) 20.024** (0.001) 20.024** (0.001) 20.024** (0.001) 20.001 (0.001) 20.001 (0.001)
Male 20.341** (0.03) 20.341** (0.03) 20.330** (0.03) 20.333** (0.03) 20.333** (0.03) 20.333** (0.03)
College (bachelor or above 51) 20.048
1
(0.03) 20.047
1
(0.03) 20.031 (0.03) 20.029 (0.03) 20.044 (0.03) 20.047
1
(0.03)
Alumni group affiliations 0.070** (0.02) 0.071** (0.02) 0.075** (0.02) 0.075** (0.02) 0.077** (0.02) 0.075** (0.02)
Entrepreneur 20.055 (0.2) 20.069 (0.2) 20.025 (0.2) 20.026 (0.2) 20.052 (0.2) 20.057 (0.2)
Currently unemployed 20.187 (0.12) 20.178 (0.12) 20.198 (0.12) 20.194
1
(0.12) 20.207
1
(0.12) 20.212
1
(0.12)
m7 m8 m9 m10 m11 m12
Seeking a job 0.205* (0.09) 0.192* (0.09) 0.194* (0.09) 0.198* (0.09) 0.195* (0.09) 0.192* (0.09)
Functional areas 20.007 (0.03) 20.006 (0.03) 20.007 (0.03) 20.007 (0.03) 20.016 (0.03) 20.017 (0.03)
Highest position (second level) 0.085 (0.05) 0.087
1
(0.05) 0.090 (0.05) 0.088
1
(0.05) 0.086
1
(0.05) 0.091 (0.05)
Highest position (Third level) 0.256* (0.03) 0.255* (0.03) 0.254* (0.03) 0.256* (0.03) 0.265* (0.03) 0.266* (0.03)
Highest position (Fourth level) 0.239* (0.04) 0.240* (0.04) 0.239* (0.04) 0.238* (0.04) 0.239* (0.04) 0.241* (0.04)
Highest position (Fifth level) 0.281* (0.05) 0.276* (0.05) 0.270* (0.05) 0.267* (0.05) 0.289* (0.05) 0.292* (0.05)
Work location (USA 51) 0.127* (0.03) 0.127* (0.03) 0.119* (0.03) 0.120* (0.03) 0.128* (0.03) 0.127* (0.03)
Tenure with DEC’s acquirer 0.002 (0.001) 0.002 (0.001) 0.002 (0.001) 0.002 (0.001) 0.011* (0.001) 0.009 (0.001)
Pre-entry ties 0.158* (0.01) 0.159* (0.01) 0.155* (0.01) 0.156* (0.01) 0.151* (0.01) 0.151* (0.01)
Constant 2.162* (0.23) 2.116* (0.21) 2.003* (0.16) 1.967* (0.15) 1.363* (0.15) 1.361* (0.15)
Log likelihood 25,009.23 25,010.85 25,019.07 25,019.35 25,043.36 25,040.84
Note: Standard errors are in parentheses. Two-tailed tests for all variables. N51,678.
a
Variables are mean-centered before forming squared terms and interaction terms.
1p,.10
*p,.05
**p,.01
2018 421Walsh, Halgin, and Huang
patterns of support in our findings suggest that in-
dividuals who enter and work in an organization at
different ages vary in terms of the salience of re-
lationships with former coworkers in their post-exit
lives. Through our focus on the effect of age-at-hire,
we draw attention to how relationships are shaped
by the extent to which individuals are concerned
about their professional development when they join
an organization. As organizations socialize new-
comers of varying ages, it is those individuals whose
membership in an organization coincides with
the midlife transition who will be most prepared to
create valuable relationships that are worthy of being
maintained after they exit it. However, our findings
regarding the effect of tenure suggest that this value
accrues over time. We propose that this result owes
to the greater opportunities for meaningful interac-
tions that are afforded by working together for longer
periods of time within a shared context. Our findings
about the impact of working during the midlife
transition supports the notion that the years of the
midlife transition may be a particularly “fertile”pe-
riod of time for forging valuable workplace ties. This
finding affirms the importance of the midlife transi-
tion as a time when individuals are particularly
motivated and capable of investing themselves in
developing lasting relationships.
We also find evidence that the organizational
performance context influenced the development of
valuable relationships in our sample. Specifically,
the degree to which an organization’s workforce is
expanding at the time of hire and the growth of its
profitability during an individual’s tenure were
found to have a positive relationship with relation-
ship maintenance. Similarly, context enhanced the
positive effect of employment during the midlife
transition and moderated the effects of age such that
employment growth dampened the curvilinear
relationship of age-at-hire. This latter result suggests
periods of optimism and expansion counteract the
relative disinterest in creating workplace relation-
ships of newcomers who are at the earlier or later
points in their careers. This finding indicates the
merit of exploring additional contextual circum-
stances, such as organization-specific norms related
to age that might also impact relationship mainte-
nance behavior. For instance, Burt (2016) finds that
peak periods of network advantage vary across or-
ganizations based on how age is perceived within
each organization (i.e., “old valued,”“old deval-
ued,”“old and young devalued”). We speculate that
the extent to which relationships are valued in the
organizational context will influence how individ-
uals perceive value in the relationships formed in
that context, thus, enhancing or dampening the im-
portance of age on relationship development.
This study has a number of important theoretical
implications. First, it recasts scholarly thinking
FIGURE 2
Graph of the Estimated Effect of Age-at-Hire Moderated by Employment Growth at Hire
–14 –12 –10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16 18 20 22 24 26
Age-at-hire (mean centered)
Low employment growth High employment growth
65432
Predicted number of relationships
422 DecemberAcademy of Management Discoveries
TABLE 5
Post-hoc Poisson Models Predicting the Number of Relationships with Former Coworkers with Additional Controls
m13 m14 m15 m16
Age-at-hire
a
0.029** (0.01) 0.035** (0.01) 0.009 (0.01)
Employment growth at hire 0.962** (0.35) 0.938** (0.35) 0.829* (0.35)
Profit margin at hire 0.388 (0.7) 0.061 (0.71) 0.353 (0.7)
Organizational tenure 0.027** (0.01) 0.027** (0.01)
Employment change over tenure 0.323 (0.3) 0.265 (0.31) 0.185 (0.31)
Change in profit margin over tenure 0.03 (0.02) 0.038 (0.02) 0.035 (0.02)
Age-at-hire squared 20.001** (0.001)
Pre-midlife transition 0.002 (0.001)
Midlife transition 0.009 (0.01)
Post-midlife transition 0.008 (0.01)
Age 0.004 (0.01) 20.029** (0.01) 20.031** (0.01) 20.014* (0.01)
Male 20.203** (0.05) 20.209** (0.05) 20.218** (0.05) 20.208** (0.05)
College (bachelor or above 51) 20.084
1
(0.05) 20.06 (0.05) 20.075 (0.05) 20.076 (0.05)
Alumni group affiliations 0.035 (0.03) 0.014 (0.03) 0.01 (0.03) 0.016 (0.03)
Entrepreneur 20.413 (0.34) 20.406 (0.33) 20.443 (0.33) 20.404 (0.33)
Currently unemployed 0.055 (0.22) 0.145 (0.22) 0.169 (0.22) 0.092 (0.22)
Seeking a job 20.195 (0.18) 20.224 (0.18) 20.234 (0.18) 20.224 (0.18)
Functional areas 20.102* (0.05) 20.126* (0.05) 20.123* (0.05) 20.123* (0.05)
Highest position (second level) 0.089 (0.08) 0.105 (0.08) 0.101 (0.08) 0.098 (0.08)
Highest position (Third level) 0.211** (0.06) 0.182** (0.06) 0.180** (0.06) 0.190** (0.06)
Highest position (Fourth level) 0.141* (0.07) 0.139
1
(0.07) 0.152* (0.07) 0.132
1
(0.07)
Highest position (Fifth level) 0.279** (0.08) 0.235** (0.08) 0.253** (0.08) 0.259** (0.08)
Work location (USA 51) 0.183** (0.06) 0.150* (0.06) 0.163** (0.06) 0.155* (0.06)
Tenure with DEC’s acquirer 0.020** (0.01) 0.004 (0.01) 0.003 (0.01) 0.01 (0.01)
Need for affiliation 0.162** (0.04) 0.167** (0.04) 0.173** (0.04) 0.161** (0.04)
m13 m14 m15 m16
Self-monitoring 0.023** (0.01) 0.020** (0.01) 0.020** (0.01) 0.021** (0.01)
Current distance to work location 0.001 (0.001) 0.001 (0.001) 0.001 (0.001) 0.001 (0.001)
Interconnectedness of relationships 0.090** (0.03) 0.086** (0.03) 0.090** (0.03) 0.089** (0.03)
Size of social network 0.272** (0.04) 0.259** (0.04) 0.250** (0.04) 0.261** (0.04)
Pre-entry ties 0.121** (0.02) 0.122** (0.02) 0.127** (0.02) 0.124** (0.02)
Constant 21.092** (0.31) 20.38 (0.33) 0.649 (0.43) 20.378 (0.36)
Log likelihood 21,553.92 21,531.78 21,527.14 21,537.82
Note: Standard errors are in parentheses. Two-tailed tests for all variables. N5539.
aVariables are mean-centered before forming squared terms and interaction terms.
1p,.10
*p,.05
**p,.01
2018 423Walsh, Halgin, and Huang
about the relational implications of employment and
aging. Scholars have found evidence that progres-
sion through the life cycle substantively reorients
individuals’work-related motives (Kooij et al.,
2011). Indeed, age shapes the ways in which in-
dividuals relate to their work roles (Nakai et al., 2011;
Ng & Feldman, 2008) and their colleagues (Avery
et al., 2007). As individuals age, they are inclined to
more effectively contribute to “noncore domains of
performance”(Ng & Feldman, 2008: 413), such as by
increasing their efforts at citizenship behavior, ad-
herence to safety standards, and aversion to counter-
productive behaviors. At the same time, although
many stereotypes about older workers have been
debunked, there is evidence that they become less
inclined toward training programs and professionally
developing themselves (Ng & Feldman, 2012). Our
work confirms and advances this line of work by
drawing attention to the ways in which individuals’
concerns about relationships with their coworkers
evolve over the course of their careers, with older
workers seemingly becoming less oriented toward
forming lasting relationships with new colleagues.
The curvilinear effect of age-at-hire suggests that
the flow and ebb of individuals’interest in forging
new ties is demarcated by their passage through the
midlife transition. Our work in part corroborates the
idea that, on realizing their goals with regards to
professionally establishing themselves, individuals
progressing through middle adulthood may prefer to
invest their time in curating the connections they
have already established (cf. Nakai et al., 2011). It
also echoes evidence from prior studies about the
curvilinear influence of age on individuals’behavior
in organizations (Ng & Feldman, 2008). The nature of
this effect suggests that the qualitatively different
experiences of moving through early and middle
adulthood motivates evolutions in how individuals
interact in organizational environments and in-
terpret their connections in them.
By attending to the impact of relationships at
a point in time that is years or even decades after they
were formed, our findings draw attention to a latency
effect of workplace ties. Our findings suggest that the
value embedded in workplace relationships provides
a crucial touchstone for sustaining them amidst con-
ditions where they would typically be expected to
atrophy (Dahlander & McFarland, 2013; Levin et al.,
2011). Our findings provide insights about the value
of relationships as a rationale to explain their survival
against the odds (Burt, 2001; Hobfoll, 2001) and as
a mechanism that supports individuals in their efforts
to form ties with others. Furthermore, our work sug-
gests that the availability of this value owes in part to
the personal and organizational circumstances that
defined an individual’s entry into an organization.
Specifically, our work suggests that individuals who
are closer to the midlife transition when they enter
organizations that are growing rapidly may more
readily initiate the development of ties that have the
potential to be maintained. Furthermore, exposure to
longer periods of high organizational performance
may strengthen such ties by providing more oppor-
tunities for high quality interactions, thereby, en-
abling them to withstand the risks of dissolution that
emerge when one or both relational partners exit.
Our research also advances scholarly understand-
ing of the role of shared social contexts in relational
processes. Through its demonstration of the potential
for ties to thrive in the wake of exit and the closure of
the organization where they formed, it suggests that
shared social contexts may not be as crucial in the
maintenance of relationships as previously posited.
Furthermore, our work advances scholarly un-
derstandings of the role played by shared contexts by
drawing attention to their potential to change over
time. Although DEC existed as an organizational en-
tity for four decades, it provided distinctly different
conditions to its employees for building relationships
over time. Our findings suggest that organizational
success can heighten individuals’attentiveness to
developing relationships with their coworkers.
Practical Implications
Our findings can inform how organizations socialize
new members in two respects. First, considering age-at-
hire can help employers design more suitable sociali-
zation processes that reflect a newcomer’s openness to
developing value-laden relationships. In particular,
those individuals who are the closest to the midlife
transition may respond in a particularly favorable
manner to collective and investiture tactics that provide
opportunities for supportive interactions with other
members (Van Maanen & Schein, 1979). On the other
hand, younger or older entrants, who may be more ret-
icent to invest in socialization (Boumans et al., 2012; Ng
& Feldman, 2012), may benefit from an initial dose of
individualized and divestiture tactics to enhance their
openness to it. Second, our theory suggests that expe-
rienced newcomers have likely developed valuable ties
in other workplaces. Hiring a middle era adult’sformer
coworkers may create a more amenable relational con-
text for them, perhaps making them more inclined to
socially connect with others in a new workplace. Such
practices may be particularly feasible given that em-
ployees with strong social bonds may be amenable to
changing jobs together (Bartunek et al., 2008).
Limitations and Areas for Future Research
The limitations of this study suggest opportunities
for future research. First, particular characteristics of
424 DecemberAcademy of Management Discoveries
our focal firm have shaped our findings in ways that
may not hold in other organizational contexts. Most
importantly, prior research on DEC (Schein, 2004)
has noted its adherence to cultural principles that
emphasized internal labor market practices (Pfeffer
& Cohen, 1984). The use of such approaches, which
may systematically prime individuals to develop
valuable ties in a workplace, has attenuated in recent
years, as evidenced by the rise of labor market
intermediaries (Bonet, Cappelli, & Hamori, 2013),
virtual work, and portfolio careers (Hall, 2002), as
well as rapidly declining expectations regarding
lifetime employment. These changes potentially re-
duce individuals’opportunities for the routine in-
teractions that enable their workplace ties to become
valuable (Dahlander & McFarland, 2013). It could
thus be questioned whether our findings would
generalize to contexts that are less oriented toward
internal labor markets. Although further research on
relationship formation and maintenance in such
contexts would provide a substantive complement to
our findings, there is evidence that valuable work-
place ties are a pressing concern and a feasible goal in
geographically dispersed work contexts (Halgin,
Gopalakrishnan, & Borgatti, 2015; Joshi, Lazarova, &
Liao, 2009). These findings suggest that the motives
to form and maintain such ties that we found in an
internal labor market-driven context are likely to
shape individuals’behavior in workplaces that pro-
vide fewer tangible occasions to develop them.
Although contemporary workplaces may make the
development of valuable ties a more challenging
endeavor, we also recognize that the importance of
support from former coworkers is stronger than ever
(Kleinbaum, 2012). For instance, in a study of con-
temporary careers, Gershon (2017) noted that more
than 60 percent of the narratives she analyzed in-
volved a former coworker helping an individual find
a job. She proposes that this finding represents
a dramatic change from the 1970s, when “network-
ing was a strategy for learning that a job existed in the
first place, not, as it primarily is today, a strategy for
calling a hiring manager’s attention to a resume”(p.
107). She attributes the changes to the contemporary
media ecology and points out the advantages during
the job search process for applicants who are rec-
ommended by former coworkers. Indeed, although
the rise of social media may make it less difficult for
individuals to maintain ties with former coworkers,
we believe that our findings regarding age-at-hire,
tenure, and age during employment can still shed
light on the propensity of individuals to take up such
opportunities. We encourage future research that
explores the effects of changes in the employment
contract that are shaping individuals’expectations and
intentions with regards to workplace relationships.
In addition, demographic trends, such as extend-
ing life expectancies and retirement ages, could give
rise to doubts about whether the curvilinear re-
lationship we find would continue to hold true in the
future. However, given the longstanding evidence of
distinct life stages (Hall, 2002; Levinson, 1986, 1996;
Nakai et al., 2011), we found the effect of age-at-hire
should remain even if the ages that define the life
stages evolve. In particular, the inverted U-shaped
effect of age-at-hire might elongate as the age bound-
aries of early and middle adulthood potentially
change. Future research could corroborate this belief
through studies of populations in countries with dif-
ferent life expectancies and retirement practices.
It is also important to note possible explanations of
our results that we cannot definitively rule out with
our analyses. First, it is possible that some of the re-
lationships that individuals are maintaining have
other contemporary bases that are sustaining them.
Scholars have long noted the prevalence of multi-
plex ties, which are anchored by interactions that
transpire in more than one context. Thus, the sur-
vival of a given relationship may owe to an un-
reported affiliation, such as mutual membership in
a social club or ties among similarly aged children.
Although the control measures and post hoc tests we
have employed were designed to mitigate this pos-
sibility, future research could corroborate our find-
ings by directly inquiring about the particular ways
in which individuals interact with their maintained
ties. Such work holds great promise to advance un-
derstanding of the effects of sequentially working
together with individuals in multiple work contexts.
Second, the effect of tenure that we found could owe
to individuals with longer tenures having fewer op-
portunities to form workplace ties after they exit and
thus are more likely to maintain their relationships
with former colleagues from DEC. It is worth noting
that in an unpresented model we found a significant
and positive interaction between tenure and age-at-
hire that is consistent with this speculation. In gen-
eral, the positive impact of tenure on maintained
relationships is enhanced when individuals join
later in life and (by definition) exit at an older age.
Although the interdependence of our measures of
age-at-hire, tenure, and age at exit preclude us from
simultaneously studying the effects of all three fac-
tors, future research could provide greater insight
about how individuals construe their opportunities for
relationship formation after exit by examining the ef-
fect of the age one leaves. Finally, our research sheds
light on specific causes and effects of maintaining ties
with former coworkers, but it is important to recognize
that investigating other contributing factors would
broaden scholarly understanding of this phenome-
non. We encourage future research on other potential
2018 425Walsh, Halgin, and Huang
individual-level explanations, such as gender, orga-
nizational identification, and occupation, as well as
contextual characteristics that may shape tie main-
tenance, such as industry and national culture.
Lastly, the design of our study introduces the
prospect for retrospective bias. Given that respondents
exited DEC many years before we conducted our survey,
it is possible that they are not accurately recalling their
experiences at the organization. Although we sought to
mitigate this problem by focusing on measures that were
relatively objective (i.e., age, tenure) and on an in-
dividual’s maintenance of relationships in the
present day, our measure of social integration is
grounded in one’s perception of an experience that
happened at some point in the past. Future research
could more fully redress the potential for retrospec-
tive bias by studying the experiences of individuals
who are currently acclimating to a new organiza-
tional context and gathering data about the prior re-
lationships that remain salient in their lives.
Conclusion
Workplace relationships are both crucial sources of
support in the rapidly changing world of work and diffi-
cult to maintain. Maintained workplace relationships
enable individuals to socially integrate themselves in new
contexts, and such ties serve purposes that transcend the
shared contexts that gave rise to them. The particular
points in time at which individuals work in an organiza-
tion have lasting implications for their preparedness to
invest themselves in maintaining their ties after exit. This
research points to the formative and lasting influence of
the value of ties formed in a workplace. When individuals
join a new workplace, they are faced with opportunities to
create relationships that develop lives of their own, as
evidenced by their ability to survive despite the loss of
a shared context. This research illuminates how the tim-
ing of one’semploymentinanorganizationleadssome
members to carry forward valuable workplace relation-
ships throughout their careers.
REFERENCES
Armstrong-Stassen, M., & Schlosser, F. 2011. Perceived
organizational membership and the retention of older
workers. Journal of Organizational Behavior, 32:
319–344.
Arnett, J. J. 2015. Socialization in emerging adulthood:
From the family to the wider world, from socialization
to self-socialization. In J. E. Grusec & P. D. Hastings
(Eds.), Handbook of socialization: Theory and re-
search:85–108. New York: Guilford.
Avery, D. R., McKay, P. F., & Wilson, D. C. 2007. Engaging
the aging workforce: The relationship between
perceived age similarity, satisfaction with coworkers,
and employee engagement. Journal of Applied Psy-
chology, 92: 1542–1556.
Bardon, T., Josserand, E., & Villes`
eche, F. 2015. Beyond
nostalgia: Identity work in corporate alumni net-
works. Human Relations, 68: 583–606.
Bartunek, J. M., Huang, Z., & Walsh, I. J. 2008. The devel-
opment of a process model of collective turnover.
Human Relations, 61: 5–38.
Blau, P. M. 1964. Exchange and power in social life. New
Brunswick, NJ: Transaction Publishers.
Bonet, R., Cappelli, P., & Hamori, M. 2013. Labor market
intermediaries and the new paradigm for human re-
sources. The Academy of Management Annals,7:
341–392.
Boumans, N. P. G., de Jong, H. J., & Janssen, S. M. 2012. Age
differences in work motivation and job satisfaction.
The influence of age on the relationships between
work characteristics and workers’outcomes. The In-
ternational Journal of Aging and Human Develop-
ment, 73: 331–350.
Burt, R. S. 2001. Attachment, decay, and social network.
Journal of Organizational Behavior, 22: 619–643.
Burt, R. S. 2002. Bridge decay. Social Networks,24:333–363.
Burt, R. S. 2016. Life course and network advantage: Peak
periods, turning points, and transition ages. In D. F.
Alwin, D. H. Felmlee, & D. Kreager (Eds.), Social
Networks and the Life Course: Integrating the Devel-
opment of Human Lives and Social Relational Net-
works. Cham, Switzerland: Springer.
Burt, R. S., & Merluzzi, J. 2016. Network oscillation.
Academy of Management Discoveries, 2: 368–391.
Dahlander, L., & McFarland, D. A. 2013. Ties that last: Tie
formation and persistence in research collaborations over
time. Administrative Science Quarterly, 58: 69–110.
Dutton, J. E., Roberts, L. M., & Bednar, J. 2010. Pathways for
positive identity construction at work: Four types of
positive identity and the building of social resources.
Academy of Management Review, 35: 265–293.
Feld, S. 1997. Structural embeddedness and stability of in-
terpersonal relationships. Social Networks, 19: 91–95.
Gershon, I. 2017. Down and out in the new economy: How
people find (or don’t find) work today. Chicago:
University of Chicago Press.
Granovetter, M. S. 1973. The strength of weak ties. Amer-
ican Journal of Sociology, 78: 1360–1380.
Granovetter, M. 1983. The strength of weak ties: A network
theory revisited. Sociological Theory: 1: 201–233.
Haggerty, J. J., & Wright, P. M. 2009. Strong situations and firm
performance: A proposed re-conceptualization of the role
426 DecemberAcademy of Management Discoveries
of the HR function. In A. Wilkinson, N. Bacon, T. Redman,
& S. Snell (Eds.), The SAGE handbook of human resource
management:100–114. Thousand Oaks, CA: Sage.
Halbesleben, J. R. B. 2006. Sources of social support and
burnout: A meta-analytic test of the conservation of
resources model. Journal of Applied Psychology, 91:
1134–1145.
Halgin, D. S., Gopalakrishnan, G., & Borgatti, S. P. 2015.
Structure and agency in networked, distributed work:
The role of work engagement. American Behavioral
Scientist, 59: 457–464.
Hall, D. T. 2002. Careers in and out of organizations.
Thousand Oaks, CA: Sage.
Hobfoll, S. E. 2001. The influence of culture, community
and the nested-self in the stress process:
Advancing conservation of resources theory. Applied
Psychology: An International Review, 50: 337–421.
Ibarra, H. 1999. Provisional selves: Experimenting with
image and identity in professional adaptation. Ad-
ministrative Science Quarterly, 44: 764–791.
Iyer, V. M., Bamber, E. M., & Barefield, R. M. 1997. Iden-
tification of accounting firm alumni with their former
firm: Antecedents and outcomes. Accounting, Orga-
nizations, and Society, 22: 315–336.
Joshi, A., Lazarova, M. B., & Liao, H. 2009. Getting everyone
on board: The role of inspirational leadership in geo-
graphically dispersed teams. Organization Science,
20: 240–252.
Kanfer, R., & Ackerman, P. L. 2004. Aging, adult develop-
ment, and work motivation. Academy of Manage-
ment Review, 29: 440–458.
Kleinbaum, A. M. 2012. Organizational misfits and the
origins of brokerage in intrafirm networks. Adminis-
trative Science Quarterly, 57: 407–452.
Kooij, D. T. A. M., de Lange, A. H., Jansen, P. G. W., Kanfer,
R., & Dikkers, J. S. E. 2011. Age and work-related mo-
tives: Results of a meta-analysis. Journal of Organi-
zational Behavior, 32: 197–225.
Kram, K. E. 1988. Mentoring at work: Developmental
relationships in organizational life. New York: Uni-
versity Press of America.
Lennox, R. D., & Wolfe, R. N. 1984. Revision of the self-
monitoring scale. Journal of Personality and Social
Psychology, 46: 1349–1364.
Levin, D. Z., Walter, J., & Murnighan, J. K. 2011. Dormant
ties: The value of reconnecting. Organization Sci-
ence, 22: 923–939.
Levinson, D. J. 1986. A conception of adult development.
American Psychologist, 41: 3–13.
Levinson, D. J. 1996. The seasons of a woman’s life. New
York: Knopf.
Marquis, C., & Tilcsik, A. 2013. Imprinting: Toward a multilevel
theory. Academy of Management Annals, 7(1): 195–245.
Marsden, P. V. 1990. Network data and measurement.
Annual Review of Sociology, 16: 435–463.
Mehra, A, Borgatti, S. P., Soltis, S., Floyd, T., Halgin, D. S.,
Ofem, B., & Lopez-Kidwell, V. 2014. Imaginary world:
Using visual network scales to capture perceptions of
social networks. Research in the Sociology of Orga-
nizations, 40: 315–336.
Mischel, W. 1977. The interaction of person and situation.
In D. Magnusson & N. S. Endler (Eds.), Personality at
the cross-roads: Current issues in interactional psy-
chology: 333–352. Hillsdale, NJ: Lawrence Erlbaum.
Morrison, E. W. 2002. Newcomers’relationships: The role
of social network ties during socialization. Academy
of Management Journal, 45: 1149–1160.
Mortimer, J. T., & Simmons, R. G. 1978. Adult socializa-
tion. Annual Review of Sociology, 4: 421–454.
Nakai, Y., Chang, B., Snell, A. F., & Fluckinger, C. D. 2011.
Profiles of mature job seekers: Connecting needs and
desires to work characteristics. Journal of Organiza-
tional Behavior, 32: 155–172.
Ng, T. W. H., & Feldman, D. C. 2008. The relationship of age
to ten dimensions of job performance. Journal of Ap-
plied Psychology, 93: 392–423.
Ng, T. W. H., & Feldman, D. C. 2010. The relationships of
age with job attitudes: A meta-analysis. Personnel
Psychology, 63: 677–718.
Ng, T. W. H., & Feldman, D. C. 2012. Evaluating six com-
mon stereotypes about older workers with meta-
analytical data. Personnel Psychology, 65: 821–858.
Pfeffer, J., & Cohen, Y. 1984. Determinants of internal labor
markets in organizations. Administrative Science
Quarterly, 29: 550–572.
Salganik, M. J., & Heckathorn, D. D. 2004. Sampling and es-
timation in hidden populations using respondent-driven
sampling. Sociological Methodology, 34: 193–239.
Schein, E. H. 1985. Organizational culture and leader-
ship. San Francisco, CA: Jossey-Bass.
Schein, E. 2004. DEC is dead, long live DEC: The lasting
legacy of digital equipment corporation. San Fran-
cisco, CA: Berrett-Koehler.
Snyder, M., & Gangestad, S. W. 1986. On the nature of self-
monitoring: Matters of assessment, matters of validity.
Journal of Personality and Social Psychology, 51:
125–139.
Song, Y., Liu, Y., Shi, J., & Wang, M. 2017. Use of proactive
socialization tactics and socialization outcomes: A
latent growth modeling approach to understanding
newcomer socialization process. Academy of Man-
agement Discoveries,3:42–63.
2018 427Walsh, Halgin, and Huang
Steers, R. M., & Braunstein, D. N. 1976. A behaviorally-
based measure of manifest needs in work settings.
Journal of Vocational Behavior, 9: 251–266.
Van Maanen, J., & Schein, E. H. 1979. Toward a theory of
organizational socialization. Research in Organiza-
tional Behavior, 1: 209–264.
Walsh, I. J., & Bartunek, J. M. 2011. Cheating the fates: Or-
ganizational foundings in the wake of demise. Acad-
emy of Management Journal, 54: 1017–1044.
Walsh, I. J., & Glynn, M. A. 2008. The way we were: Legacy
organizational identity and the role of leadership.
Corporate Reputation Review, 11: 262–276.
Ian J. Walsh (iwalsh@isenberg.umass.edu) is an assistant
professor of Management at University of Massachusetts
Amherst. He received his PhD in Organization Studies
from Boston College. His research interests include social
cognition, workplace relationships, and consequences of
organizational demise.
Daniel S. Halgin (danhalgin@uky.edu) is an assistant pro-
fessor of management at the University of Kentucky and
a member of the LINKS Center for Social Networks Anal-
ysis. He received his PhD in Management from Boston
College. His research interests include social networks and
identity dynamics.
Zhi Huang (zhi.huang@uky.edu) is an associate professor
in the Department of Management at University of Ken-
tucky and holds a PhD in organization studies from Boston
College. He examines how status and power affect orga-
nization, group, and individual behaviors and outcomes in
various contexts (e.g., U.S. college sports, corporate boards,
work team).
428 DecemberAcademy of Management Discoveries
CopyrightofAcademyofManagementDiscoveriesisthepropertyofAcademyof
Managementanditscontentmaynotbecopiedoremailedtomultiplesitesorpostedtoa
listservwithoutthecopyrightholder'sexpresswrittenpermission.However,usersmayprint,
download,oremailarticlesforindividualuse.