Apologies and Transformational Leadership

13 Pages·2005·0.15 MB·English

Journal of Business Ethics (2006) 63: 195–207 (cid:1) Springer 2006
DOI 10.1007/s10551-005-3571-0
Sean Tucker
Nick Turner
Julian Barling
Apologies and Transformational
Erin M. Reid
Leadership
Cecilia Elving
ABSTRACT. This empirical investigation showed that Introduction
contrary to the popular notion that apologies signify
weakness, the victims of mistakes made by leaders ‘‘If you have behaved badly, repent, make what
consistently perceived leaders who apologized as more amends you can and address yourself to the task
transformational than those who did not apologize. In a of behaving better next time.’’ – Aldous Huxley
field experiment (Study 1), male referees who were
perceived as having apologized for mistakes made offi- ‘‘No sensible person ever made an apology.’’ –
ciating hockey games were rated by male coaches Ralph Waldo Emerson.
(n = 93) as more transformational than when no apol-
With increasing media coverage of the self-serving
ogy was made. Studies 2 (n = 50) and 3 (n = 224)
replicated this effect in two vignette studies to enhance and often unrepentant behavior of certain corpo-
internal and ecological validity. Contrary to expecta- rate leaders, it would be easy to conclude that
tions in Study 3, there were no apology(cid:1)leader gender modern leadership reflects the views of Emerson
interactions. Theoretical and practical implications are rather than Huxley. While recent high profile
discussed. cases serve to reinforce waning public and em-
ployee confidence in corporate leadership, we
KEY WORDS: apologies, moral leadership, transfor-
suggest that ethical leaders who attempt to ‘‘do the
mational leadership
right thing’’ with their words and actions will be
perceived as better leaders by followers. Instead of
denying their mistakes, ethical leaders apologize,
make amends, and take steps to avoid repeating
transgressions in the future. Indeed, some popular
writers have recently argued that apologizing is a
Sean Tucker, MSc, is a PhD candidate in organizational be- prerequisite for high quality leadership (Blanchard
havior at the Queen’s School of Business, Queen’s Uni- and McBride, 2003; Lazare, 2004; Timson, 2003).
versity, Canada. Although providing a genuine apology can be a
Nick Turner, PhD, is an assistant professor or organizational humbling experience, particularly for individuals in
behavioratQueen’sSchoolofBusiness,Queen’sUniversity, leadership positions, research evidence is beginning
Canada. to suggest (e.g., Kim et al., 2004) that apologies
Julian Barling, PhD, is a professor of organizational behavior
are critical in rebuilding and sustaining long-term
at Queen’s School of Business, Queen’s University,
relationships.
Canada.
Whileresearchandpopularwritingseachpointto
Erin M. Reid, MSc, is an analyst in the Intergovernmental
the interpersonal benefits of an apology, reports
Relations group, Human Resources and Skills Develop-
suggest that sincere apologies actually occur rela-
ment Canada.
tively infrequently in organizations (Timson, 2003;
Cecilia Elving, MSc, is a researcher at the Swedish Guide
and Scout Council. Weeks, 2003). Two reasons might account for this.
196 Sean Tucker et al.
First, public admissions of fault are perceived by admission of responsibility but a denial of wrong-
manyleadersasembarrassing,asignofweaknessthat doing.Denialsrefutebothresponsibilityandwrong-
threatens their authority (Jackall, 1988), even wors- doing. Finally, in an apology, the transgressor both
ening sensitive situations (Folger and Skarlicki, admits the act was wrong, accepts responsibility for
2001). Second, organizational leaders are often the offence, expresses empathy, offers penance, and
counseled that apologizing to the aggrieved could promises not to repeat the untoward behavior in the
expose the individuals or their organizations to liti- future (Goffman, 1971; Schmitt et al., 2004).
gation(e.g.,Neckers,2002).Insharpcontrasttothis Most explanations for the nature and conse-
apparent conventional wisdom, we suggest that quences of an apology fall under impression man-
apologizing following wrong-doing will positively agement theory. Goffman (1971) classified apologies
influence leadership perceptions. In particular, we as a type of remedial work. Specifically, he used the
believe that leaders who apologize will be perceived image of ‘‘splitting of the self’’ (p. 113): one half of
as looking beyond self-interest for the good of the the individual representing the wrongdoing, and the
relationship, thus embodying many elements of other half sympathetic to the victim, hoping to be
transformational leadership (i.e., inspirational moti- forgiven. In a somewhat different view, Schlenker
vation, idealized influence, intellectual stimulation, (1980) argued that apologies are a more self-serving
and individualized consideration; Bass, 1998). impressionmanagementtacticthatindividualsuseto
While the interpersonal effects of apology have maintain their social standing and save face. From
garnered increased attention, surprisingly little re- both perspectives, apologies constitute a visible and
search has examined the implications of an apology unambiguousbehaviorthatenablesvictimstoviewa
on subsequent leader-follower relations. With few leader’s behavior as a socially responsible reaction to
exceptions (e.g., Giacalone and Payne, 1995; wrongdoing. A sincere apology signals vulnerability
Tomlinson et al., 2004), most studies on apologies and transmits moral meaning, allowing the repair of
involve participants of equal role status in non- interpersonal relationships to begin.
organizational settings. In this paper, we present Researchshowsthatapologiesarecomplexspeech
three studies that examine follower perceptions of acts, which can have a range of positive effects,
leaders who apologize following misconduct. The including generating forgiveness (Exline et al.,
first study examines apology in a field context 2004), restoring trust (Kim et al., 2004), reducing
involving ice hockey coaches. Studies 2 and 3 aggression (Ohbuchi et al., 1989), enhancing future
attempt to replicate findings in Study 1, enhance relationship closeness, and promoting well-being
internal and ecological validity, and explore the (Hodgins and Liebeskind, 2003; Witvliet et al.,
effects of leader gender on perceptions of transfor- 2002). Individual and situational correlates of apol-
mational leadership. Collectively, these studies pro- ogies have included transgressor and victim gender
videcompellingevidencethatleadershipperceptions (e.g.,Gonzaleset al.,1990;HodginsandLiebeskind,
among followers are higher when an apology is 2003),contextofapology(FolkesandWhang,2003;
provided compared to when no apology is given. Sigal et al., 1988), apology timing (Skarlicki et al.,
2004), perpetrator autonomy (Folkes and Whang,
2003; Hodgins et al., 1996), offence severity (e.g.,
The nature of an apology Tomlinson et al., 2004), status of offender (Gonzales
et al.,1990),anddegreeofrelationship(Hodginsand
Apologies belong to the class of speech acts known Liebeskind, 2003).
as accounts, which are tools used to respond to While there has been considerable research
perceived offences or misunderstandings. Building attention placed on understanding the nature and
upon Scott and Lyman’s (1968) earlier work, consequences of apologies, research on the rela-
Schlenker (1980) classified accounts into four tionship between apologies and leadership is scant.
categories: excuses, justifications, denials, and apol- To our knowledge, only one laboratory-based study
ogies (or concessions). Excuses involve admitting of apologies has manipulated offender status
wrong-doing but refusing to take personal respon- (Gonzales et al., 1990), and one study examined
sibility. Justifications are the opposite, involving an follower reactions to excuses, denials, or justifica-
Apologies and Transformational Leadership 197
tions offered by organizational leaders for decisions responses of others, particularly those in positions of
(Schaubroeck et al., 1994). No studies directly authority. This perspective is consistent with a so-
examine how followers perceive leaders who ac- cial constructionist view of leadership, which con-
count for their mistakes with apologies. We believe ceptualizes leadership as constructed through
that apologies can play an important role in devel- language, social interaction, and the formation of
oping and repairing leadership perceptions in orga- shared meaning (Chen and Meindl, 1991). We
nizations, and ground this prosocial orientation in propose that followers’ lasting and most salient
transformational leadership theory. perceptions of their leaders are primarily based on
distinct interactions that, while occurring relatively
infrequently, serve to punctuate or reinforce the
Apologies and socially constructing leadership status quo. We will refer to these situations as critical
moments.
Among all theories of leadership, transformational In organizational contexts, significant points in
leadership theory was the most frequently studied in time in working relationships provide opportunities
the 1990’s (Judge and Bono, 2001) and has been for supervisors and employees to construct leader-
shown in various contexts to be related to higher ship. For example, how a manager copes with a
socio-moral reasoning (Turner et al., 2002), in- sudden drop in sales may function as an opportu-
creased motivation (e.g., Charbonneau et al., 2001), nity for the employee to evaluate the manager’s
higher business performance (e.g., Barling et al., leadership skills. The manager who unfairly blames
1996), andreducedworkplaceinjuries(e.g.,Barling employees for the decline is likely to decrease
et al., 2002. Transformational leadership is primarily subordinates’ perceptions of her leadership; the
distinguished from other theories of leadership byits manager who shares responsibility and apologizes
focus on follower development (Avolio, 1999). for acute problems may be perceived more posi-
Transformational leaders display four specific char- tively. In these hypothetical but highly salient sit-
acteristics: idealized influence, inspirational motiva- uations, the reaction to the critical moment
tion, intellectual stimulation, and individualized supersedes in importance previous taken-for-gran-
consideration. Idealized influence stems from lead- ted interactions with the leader, which followers
ers’ ethical behavior. As role models, leaders build use as a wellspring of cues for forming leadership
respect and trust among followers. Leaders create perceptions.
inspirational motivation by raising followers’ per- Surprisingly, researchers have yet to test the
sonal expectations and setting new, higher aspira- widespread belief in many managerial circles that
tions for individuals and groups. Intellectual leaders should avoid apologizing for mistakes lest
stimulation is enhanced when leaders challenge fol- they be seen as weak. In contrast, we hypothesize
lowers to think for themselves and re-consider old that a leader’s apology for a transgression will en-
problems in new ways. Lastly, transformational hance follower perceptions of their transformational
leaders grant their followers individual consideration leadership for two reasons. First, when individuals
by paying attention to their unique needs and abil- perceive their leader is fair based on the way he or
ities (Bass and Steidlmeier, 1999). she behaves (e.g., apologizing), they are likely to
While a great deal is known about the effects of believe their leader engages in those behaviors
transformational leadership, a key issue remains willingly. This is critical as one major component of
unexamined: when and how do followers begin to transformational leadership is idealized influence,
see leaders as transformational? In everyday life, which is manifested when leaders act on their values
people may pay little attention to routine social and are guided by their beliefs, choosing to do the
interactions (Berger and Luckman, 1966). In times right thing (Bass, 1998). Second, leaders who
of crisis, however, individuals become more cog- apologize will be seen as doing so because they care
nizant of the behavior of others as they seek to for the individual and the relationship, which re-
make sense of the abnormal situation. These situ- flects individualized consideration, another integral
ations may lead to the development of strong component of transformational leadership. We
impressions based on observed reactions and argue that mistakes, wrongdoing, and other unusual
198 Sean Tucker et al.
situations are critical moments that attune followers tionalleadershipthanwillmaleleaderswhodo
to the behavioral intentions of their leaders. Fur- notapologize.
ther, that apologies offered in response to these
We conducted three studies to test our three
types of incidents are associated with higher trans-
hypotheses (Hypothesis 1 and Hypotheses 2a and
formational leadership perceptions. Therefore:
2b). To our knowledge, no studies that investigate
Hypothesis 1: Leaders who apologize for mis- apologies have been conducted in field settings.
takes will be perceived as more transforma- Replication beyond the laboratory to situations that
tional than leaders who do not apologize for maximize ecological validity remains a next step.
mistakes. Thus, our first study tests the effect of apologies
offered by amateur ice hockey referees to competi-
Research on gender and leadership indicates that
tive level hockey team coaches for mistakes com-
women struggle to be perceived as leaders in orga-
mitted during officiating a hockey game.
nizations (Liu and Wilson, 2001). A number of
studies have shown that women are rated higher in
transformational leadership than males (e.g., Bass Study 1
et al., 1996; Eagly et al., 2003), and that the com-
ponents of transformational leadership are more Method
consistently linked with positive outcomes than the
components of transactional leadership, on which Overview and context
men on average score higher than women. There is This study was conducted in the context of referees
also some evidence suggesting that perceptions of apologizing (or not) to amateur competitive league
leaders who apologize may be influenced by leader hockeycoachesformistakesmadewhileofficiatinga
gender. Laboratory-based studies have found that game. Amateur hockey players who aspire to pro-
women tend to offer more frequent and richer fessionalcareerscompeteatthecompetitivelevelfor
apologies than men (Gonzaleset al., 1990, 1992). In positions on junior hockey teams, from which pro-
contrast, men are more likely to avoid apologizing, fessional teams recruit prospective players. The
and instead show a greater preference to use aggra- opportunity to play elite hockey, and potentially be
vating accounts such as denials (e.g., Hodgins and rewardedwithlucrativecontractsinthefuture,raises
Liebeskind, 2003). However, Tata (1998) illustrates the salience of errors committed by referees in call-
what may be a gender bias among followers against ing plays. Indeed, hockey referees are frequently
female managers. She found that female managers targets of strong criticism from spectators, players,
were evaluated more harshly by subordinates than and coaches for perceived errors (Irvin, 1997).
male managers when they used aggravating accounts Refereesdomorethanapplyrulesandprocedures
(denials). The inconsistency between gender-fo- toensurefairoutcomesbetweenteams;theyserveas
cused research on transformational leadership and leaders when they seek to influence coaches and
apologies leaves us with a necessarily more explor- players to engage in appropriate behaviors, and im-
atory set of research hypotheses regarding the pose penalties (discipline) when infractions occur.
interaction of apologies and leader gender on Indeed, research suggests that amateur players per-
transformational leadership. Therefore: ceive sports officials as influential figures during
games (Wann et al., 2000). Further, studies of both
Hypothesis 2a: Female leaders who apologize American college football officials (Ittenbach and
will be rated more highly on transformational Elter, 1988) and German referees of a range of
leadership than will male leaders who apolo- competitive sports (Brand and Ness, 2004) found
gize. that referees’ personality traits were similar to other
individuals with ‘‘well-developed qualities of lead-
Hypothesis 2b: Female leaders who do not ership’’ (Ittenbach and Elter, 1988, p. 121).
apologize will be rated lower on transforma- Reportsofapologiesofferedbyprofessionalsports
referees and umpires to coaches, players, and spec-
Apologies and Transformational Leadership 199
tators illustrate the complex roles and responsibilities their instructions were, on average, 2.59 years
oftheseofficials.Decisionsbyrefereescanhavelarge younger (p < 0.05) than those who did not, but did
consequences for the outcome of a game, and as not differ on amount of coaching experience. There
manyspectatorsandcommentatorsattest(e.g.,Irvin, were no significant differences on demographic
1997), mistakes are not uncommon although apol- variables for those in the apology condition versus
ogies are. Nonetheless, consistent with our first those the no apology condition. As transformational
hypothesis, apologizing may be an important way in leadership perceptions did not differ significantly
which referees can enhance their status as perceived across demographic variables, we collapsed analyses
by athletes and coaches. across age and coaching experience.
Participants Measure
Head coaches of competitive level hockey teams Transformational leadership was measured using a
located in two geographically separated hockey modified version of Carless et al. (2000) seven-item
associations in Canada were invited by telephone Global Transformational Leadership scale (see
to participate in the study. Coaches were ran- Appendix 1 for the items). We chose this shortened
domly assigned to receive one of two question- and validated scale instead of the more widely used
naires, which asked them to recall one of two Multifactor Leadership Questionnaire (Bass and
conditions: (1) a situation in which a referee with Avolio, 1995) because of its brevity and clear uni-
whom they are familiar apologized for making a dimensionality. We eliminated one of the seven-
mistake in a game; or (2) a situation in which a items (i.e., ‘‘My leader fosters trust, involvement,
referee with whom they are familiar did not and cooperation among team members’’) given its
apologize for making a mistake in a game. focus on team, instead of the individual relationship
Questionnaires provided no definitions of what with the leader represented in the remaining six
constituted an apology or a mistake, asked items. These items were adapted to fit the situation.
respondents to describe briefly how they felt about For example, one item was re-worded to state:
the situation in which the mistake had occurred, ‘‘Thisrefereeapproacheseachgamewithaclearand
and contained the transformational leadership positive outlook.’’ Participants responded on a
measure described below. In addition, participants 5-point Likert scale measuring behavioral frequency
indicated their own age, gender, number of years (0 = rarely or never to 4 = very frequently). The
coaching a competitive level team, and the gender coefficient alpha for this scale was 0.93.
of the focal referee. Coaches were asked to return
their survey completed if they could recall the
situation assigned to them, and return it blank if Results and Discussion
they could not. Four lotteries of fifty dollars each
were offered as an incentive for participation in Our first hypothesis was supported: referees who
this study. apologized were rated by coaches as more transfor-
Three hundred and thirty-eight head coaches of mational (M = 2.50, SD = 0.81) than those who
competitive level hockey teams (with players aged did not (M = 1.51, SD = 1.04), t(91) = 5.07, p <
8–17 years) agreed to participate and were sent 0.001.
questionnaires, and 156 surveys were returned (46% Results from Study 1 offered preliminary support
response rate).Ofthese, 94 wereusable asthe coach for the hypothesis that leaders who apologize for
had recalled the situation as described in their mistakes are viewed as more transformational.
instructions(43fortheapologycondition,51forthe Nonetheless, despite the fact that the study was
no apology condition). conducted in a field setting, some limitations limit
Themeanageofrespondentswithuseabledatawas therobustnessofthemainfinding.First,theabsence
42.35 years (SD = 6.83 years), all were male, had offormaldefinitionsofwhatconstitutedanapology,
coached an average of 7.89 years (SD = 6.66 years), andtoalesserextentwhatamistakewas,madeeach
and all game situations involved male referees. phenomenon open to individual interpretation,
Respondents who recalled the situation described in representing a possible threat to internal validity.
200 Sean Tucker et al.
Second, while many referees influence coach empathy and regret, and a promise to change his
understandings and behaviors through open com- behavior following the transgression. Twenty-four
municationandassessingfouls(e.g.,punishment),the participants faced the apology condition, with 26
relevance of this specific setting to the relationship participants in the non-apology condition. Both
between leaders and subordinates in organizations groups read a short scenario in which the trans-
remains to be demonstrated. To overcome these gression was described as follows:
limitations,weconductedasecondstudytoenhance
It is 2 months into your summer job at a large
the construct validity of an apology, provide
firm in Toronto. You have been working hard
respondentswithacommonmistakestimulus,andto
and getting along well with your manager, An-
better reflect a more traditional leader-subordinate
drew. Next week is your week off, and you and
interaction in an organizational setting.
your friends have planned your first ever trip to
Las Vegas. Your parents disapprove of Vegas, so
you’re paying for it all on your own. You ha-
Study 2
ven’t saved enough, but with this week’s pay,
you will be able to afford the trip. When you
Method
check your bank account, it turns out you have
not been paid. You approach Andrew and ask
Overview him where your pay check is. He says, ‘‘You
To overcome these three potential limitations, we skipped work last week. So I told the payroll
conducted a vignette study. First, construct validity department to dock your pay.’’ You did not skip
was enhanced because four components of an work that week; you were in fact working on a
apology wereincludedin thescenarioto ensure that special project in another department, to which
Andrew had assigned you several weeks ago, but
all participants in the apology condition faced the
he had apparently forgotten. You explain this to
same social account. Second, all respondents reacted
Andrew, then tell him you will have to cancel
to the same situation, involving an identical mistake
your trip if he can’t pay you by the next day.
scenario made by a supervisor. Third, ecological
He responds:
validity to leadership in organizational contexts was
enhanced because the scenario involved a supervisor Forthoseinthenon-apologycondition,thevignette
who either apologized or did not apologize to an closed with Andrew saying:
undergraduate student who did not receive their
wages due to an error maybe by the supervisor. ‘‘Oh – so then you didn’t skip work. Huh.
Unfortunately, I can’t do anything for you right
now – you won’t get the pay fixed for a month.’’
Participants
Fifty male students at a mid-sized Canadian business In contrast, Andrew responded as follows in the
school were recruited to participate in this vignette apology condition:
study. The average age of the participants was
23 years (SD = 4.38). All participants received a ‘‘Oh no! This is completely my fault – I shouldn’t
lotteryticketascompensationfortheirparticipation. have been so hasty. Unfortunately, I can’t do any-
There was no relationship between age and trans- thing for you – you won’t get the pay fixed for a
formational leadership thus we collapsed analyses month. I appreciate that this will make things dif-
ficult for you, I promise you, it won’t happen
across the demographic variable.
again. Please accept my sincerest apologies.’’
Instrument After reading the vignette, participants completed a
The study used a between-subjects factorial design. questionnairedesignedtoestablishthevalidityofthe
Participants were randomly assigned to either the manipulations and transformational leadership per-
apology or no apology condition. In the apology ceptions. The first three items on the questionnaire
condition, a supervisor was depicted as taking were manipulation checks. The first measured
responsibility for a management mistake, expressing whether participants believed that a mistake was
Apologies and Transformational Leadership 201
made: ‘‘Putting yourself in the shoes of the student, mational leadership after apologizing for a mistake
to what degree do you think that the manager made evenafterbeingcarefultoexcludepossiblethreatsto
a mistake?’’ The second item measured whether construct and ecological validity. Together, there-
participants believed that an apology had been of- fore, the results from Studies 1 and 2 suggested that
fered: ‘‘Putting yourself in the shoes of the student, male leaders who apologize for mistakes are per-
to what degree do you think the manager provide a ceived as more transformational than male leaders
sincere apology?’’ The third item assessed the per- who do not apologize for mistakes.
ceived plausibility of the vignette: ‘‘This situation is Nonetheless, all the participants in these studies
plausible.’’ Responses to all of these questions were were male, as were the referees and the manager
given on a 1 (not at all) to 7 (most certainly) scale. portrayed in the vignette, thus limiting the ability to
Participants then completed the Carless et al. generalize beyond male dyads because of findings
(2000) transformational leadership scale as described showing consistent gender differences in transfor-
in Study 1. We again scored each item on the 5- mational leadership (Eagly et al., 2003). As a result,
point Likert-type scale, and removed the item with we conducted a third study that included an explicit
theteamreferent,leavingsixitemswithacoefficient focusonwhetherleadergendermoderatestheeffects
alpha of 0.86. of an apology on perceptions of transformational
leadership.
Results
Manipulation checks Study 3
The manipulation checks confirmed the validity of
mistake, apology, and plausibility manipulations. Overview
First, there was no difference between scenarios on
the perceptions that the supervisor had made a Study 3 had two aims. The first was to replicate the
mistake, t(47) = 1.06, ns. Across both the apology findings of Study 2 in a vignette context in which
conditions, participants believed that the manager we provided participants with a common mistake
had made a mistake (M = 6.31, SD = 1.18). Sec- and a complete apology definition. The second
ond, participants were more likely to believe that a investigated whether the gender of the leader in the
sincere apology had been offered in the apology vignette would moderate any effects of an apology
condition (M = 5.29, SD = 1.52) than the no on perceptions of transformational leadership by
apology (M = 2.44, SD = 1.39) condition, t(47) = using both male (Andrew) and female (Anne)
6.87, p < 0.001. Third, the perceived plausibility of supervisors in the vignette.
the two vignettes was high (M = 5.69, SD = 1.23),
with no differences between the two apology con-
ditions, t(47) = 0.81, ns.
Method
Transformational leadership perceptions
We then tested the hypothesis that perceptions of Participants
transformational leadership would differ across con- Two hundred 24 students at a mid-sized Canadian
ditions.Intheapologycondition,themeanscoreon business school were recruited for this electronic
transformational leadership (M =3.41, SD = 1.04) survey. The average age of the participants was
was higher than in the no apology condition 20.62 years (SD = 3.38). Of the participants, 48.2%
(M = 2.79, SD = 0.97), t(47) = 2.17, p < 0.05. were male and 51.8% were female. All participants
received either bonus course credit or a lottery
ticket as compensation for participation. Transfor-
Discussion mational leadership perceptions did not differ sig-
nificantly by age or respondent gender, as a result of
The results of Study 2 replicate and extend those of which we did not include either variable as cova-
Study 1: leaders were seen to be higher in transfor- riates in the analyses.
202 Sean Tucker et al.
Design ratedmoretransformationalthanleaderswhodidnot
This study used a 2 (apology condition: apology apologize for the mistake (M = 2.57, SD = 1.06).
versus no apology)(cid:1)2 (leader gender: male versus Male leaders (M = 2.99. SD = 1.10) received
female) between-subjects factorial design in which higher transformational leadership scores than
participants were randomly assigned to one of four female leaders (M = 2.71, SD = 1.09), regardless
conditions. There were between 49 and 65 respon- of whether the leader apologized for the mistake
dents in each condition. or not. There was no effect for the interaction
between apology and leader gender,
Instrument F(3,220) = 0.68, ns, providing no support for
Participants were randomly assigned to either of the hypotheses2aand2b.
two scenarios (apology and no apology condition)
described in Study 2 and completed the same ques-
tionnaire. As already noted, the scenario described a
General discussion
mistake that was made over a summer student’s pay,
who as a result could not afford a much anticipated
These studies represent a first attempt to examine
vacation.Participantsintheapologyconditionreada
follower assessments of leaders after they apologize
scenariothatdescribedthesupervisorapologizingfor
for a mistake. In Study 1, we found that male ice
the mistake (i.e., admission of mistake, accepting
hockey referees who apologized to male coaches for
responsibility, expressions of remorse, and an action
mistakes they made during games were rated higher
plan for the future). As in Study 2, the no apology
in transformational leadership than referees who did
condition excluded these elements.
not apologize for an error. Study 2 replicated and
extended the findings of Study 1. As the vignette
Results was situated in an organizational context dealing
with a supervisor-employee issue, and because a
Manipulation checks specific definition of an apology was provided, both
Consistent with Study 2, the manipulation checks internal and ecological validity were enhanced.
confirmed the validity of the manipulations. There Again, the positive effect of apologizing on trans-
was no difference among scenarios on the percep- formational leadership perceptions was found.
tions that the supervisor had made a mistake, Finally,inStudy3,wereplicatedandextendedthese
F(3,220) = 0.94, ns. Participants believed that the findings by manipulating leader gender using the
manager had made a mistake irrespective of the vignette used in Study 2. Collectively, these results
condition to which they had been assigned offer strong support that apologizing after wrong-
(M = 6.42, SD = 1.15). With respect to apology doing is related to higher perceptions of transfor-
making,participantsweremorelikelytobelievethat mational leadership.
an apology had been offered in the two conditions Transformational leadership has received consid-
containing an apology (M = 4.69, SD = 1.86) than erableresearchattentioninthelastdecademorethan
the two conditions containing no apology (M = all leadership approaches combined (Judge and
2.02, SD=1.29) condition, t(220) = 12.37, p < Bono, 2001). However, studies have focused pre-
0.001. Participants saw all scenarios as equally rea- dominantly on demonstrating the effects of trans-
listic (M = 5.23, SD = 1.39), F(3, 220) = 0.53, ns. formational leadership. Our research considered
behavioral incidents that inform follower percep-
Transformational leadership perceptions tions of transformational leadership. These findings
A 2(cid:1)2analysisofvariancewasconductedonpartic- contribute to our understanding of the development
ipants’ perceptions of transformational leadership, of leadership perceptions in organizational settings
yieldingasignificanteffectofapologizing,F(1,220) = via interpersonal accounts in response to critical
17.29,p < 0.001,andleadergenderF(1,220) =3.66, moments. Within this perspective, followers evalu-
p < 0.05. As in Studies 1 and 2, leaders who apol- atetheirleadershipqualitiesbasedonleaderbehavior
ogized for the mistake (M = 3.16, SD = 1.07) were tosuchuniquemoments.Thesesituationsaresalient
Apologies and Transformational Leadership 203
to subordinates, and result in close evaluation of the of apology and victim perceptions is an important
leader’s words and actions. area that warrants future research.
Any effects of an apology on leadership may be A second possible explanation parallels Tata’s
especially important for transformational leadership (1998) findings regarding the harsher evaluations of
theory, as apology is particularly consistent with females’socialaccountsinaggravatedsituations.The
two of the four components of transformational vignetteusedinStudies2and3concernedamistake
leadership. First, apologies demonstrate idealized made by the supervisor, with all respondents
influence, by emphasizing the importance of regardless of condition in these studies rating the
behaving in an ethical manner and taking respon- presenceofamistakebythesupervisorassignificant.
sibility for one’s actions. As Mills (2001) suggests, Under these conditions, the expectation of stereo-
leaders are ‘‘humanized by apologizing in a way typical female managers’ behaviors might be differ-
that a wrongdoer who remains silent and appears ent than stereotypical male managers’ behaviors.
indifferent to public opinion is not’’ (p. 115). More specifically, female managers might not be
Second, apologies also show caring for the em- expected to make such an error compared to male
ployee and the leader-employee relationship, managers, and are thus rated more harshly on lead-
exhibiting individualized consideration. ership behaviors. The more general evidence that
The results of Study 3 did not support hypotheses female managers are rated as more transformational
that leader gender interacted with apologizing to than male managers (Eagly et al., 2003) reflects a
explain additional variance in perceptions of trans- range of circumstances, rather than a specific con-
formational leadership. We found that respondents text like in Studies 2 and 3 in which a mistake has
rated male supervisors in a vignette more transfor- been made.
mational in both the apology and no apology con- Third and finally, the design of Studies 2 and 3
ditions. These findings are somewhat surprising implicitly assumed that no gender differences would
given findings from other studies which suggest that exist in terms of the manager’s behavior when the
femalesaregenerallyratedhigherintransformational mistake became known. All vignettes conditions
leadership(Eaglyet al.,2003)andtendtooffermore were held constant, and we changed only the man-
and richer concessionary accounts for perceived ager’s names (i.e., Andrew or Anne). In an organi-
interpersonal breaches than males (Gonzales et al., zational context, female leaders may apologize in a
1990,1992).Weofferthreepossibleexplanationsfor qualitativelydifferentwaythan maleleaders;indeed,
these unexpected findings. First, we believe that the apology research suggests this to be the case (e.g.,
relative infrequency of apology as an artifact of Hodgins and Liebeskind, 2003). This finding is also
everydayorganizationallifemaypartiallyexplainthis consistent with research that suggests that women’s
finding. On one hand, there seems to be reluctance perceptionsaremorestronglyaffectedbyexpressions
amongmanyorganizationalleaderstoofferapologies of responsibility and regret than are men’s (Tata,
to subordinates because leaders are averse to volun- 2000). Thus, leader gender may be more salient in a
tarily admitting personal mistakes. On the other less controlled, more ecologically rich research
hand, while followers may desire that a leader make design.Inthisvignette,weattemptedtochangeonly
amends for offences, they may also not expect an gender, and not leader behavior, preferring to mea-
apology based on their past experiences with leaders sure effects prompted only by gender categories,
and cynicism about leaders in general who appear to ratherthanbystereotypicalbehavior.Webelievethis
put self-interest ahead of moral action. Thus, when austerity is another possible reason why an apology
leaders voluntarily apologize for their misdeeds, it by leader gender effect was not isolated.
can make a powerful impression on followers who
were not anticipating the apology. In Study 2, there
may have been an additive effect for male leader Limitations
apologies because participants expected them to
apologize lessfrequently for transgressions compared Previous apology research has identified a range of
to females. The relationship between the frequency correlates of apology effectiveness. Across the
204 Sean Tucker et al.
current studies, we focused only on leadership and inreallife.InthecaseofStudies2and3,participants
gender. In Study 1, for example, the seriousness of may have felt it socially desirable to respond favor-
thereferee’smistakeandthenumberofpastmistakes ably to the supervisor who apologized. It is possible
werenotcontrolled.Itispossiblethattheseverityof that study participants who were confronted by a
the mistake during a game (e.g., failure to assess a breach that negatively affected them may respond
penalty against an opponent early in a game versus less favorably to an apology. Moving research on
failure to call a penalty in the last minute of a close apologies in organizational contexts into the field, as
championship game) influences transformational opposed to simulations such as vignettes, will
leadership perceptions. Similarly, in Studies 2 and 3, provideopportunitiestoexternallyvalidatetheeffect
we did not manipulate the severity of the supervi- of apologies on leadership perceptions.
sor’s breach. Gonzales et al. (1990) examined dif- Finally, while the apology condition in Studies 2
ferences between low- and high-status apologizers and 3 included many key elements of apology, the
and found that high-status actors may feel less pres- respondent was not offered penance, which in this
sure to offer concessionary accounts for less severe case could have been a monetary amount in lieu of
offences. For more severe offences, they found no the full amount of their pay check. Our intention
differences between low- and high-status actors. In was to make the apology as salient as possible. To
line with these findings, we would propose (all else do this, we intentionally left out an offer of resti-
being equal) that leaders who apologize for less tution to the aggrieved because we believed it
serious offences would be rated as more transfor- would dominate the four other elements of apology
mational compared to leaders who do apologize in in essence undoing the mistake from the perspective
these situations. This is an important issue which of the respondent. Two recent studies found that
future research might address. this specific element of apology has a particularly
In Studies 2 and 3, it is possible that the favorable strong effect (Bottom et al., 2002; Schmitt et al.,
results are due in part to an unfair comparison 2004).
between the apology and no apology conditions
(Cooper and Richardson, 1986). Study participants
responding to the no apology condition were pro- Directions for future research
vided with a hurried and insensitive sounding
response by the supervisor who withheld part of There is a need for more research on leadership and
their pay. Some may see this response as unrealistic apology, and in general, on the role of apologies in
and provoking negative feelings in the participant, organizations both as a conflict management and
thus setting up a situation in which the apology relationship repair device. First, while there is ample
condition produced markedly better leadership evidence that leaders are averse to apologizing, we
perceptions.Notwithstandingtheseconcerns,wedo know very little about what leaders think of apolo-
not think that the vignettes make an unfair com- gizing to subordinates. Qualitative research on
parison for two reasons. First, the manipulation responses to critical moments would provide insight
check of the apology and no apology conditions intoareassuchasconsequencesonleaderwell-being
showednosignificantdifferencesinrealism.Second, and self-perceptions before and after social accounts
we argue that it is not unusual for some supervisors are offered to followers.
to demonstrate a high level of insensitivity and Second, more research is needed on the contex-
indifference in situations where they have made a tual boundary conditions that may influence the
mistake (Jackall, 1988). This may occur when the effectivenessofleaders’apologies.Forexample,Sigal
aggrieved subordinate is perceived as non-threaten- et al. (1988) found that political candidates were
ing (e.g., a summer student) or in times of organi- rated higher by voters when they denied rather than
zationalperformanceproblems(FolgerandSkarlicki, apologizedforallegedmisconduct.Thissuggeststhat
2001). followers may becynical of leaderapologies in some
As with all vignette studies, there are questions settings.Otherevidencesuggeststhatifanapologyis
about whether participants would respond in the perceived to have a manipulative intent it may
samewaytothehypotheticalsituationastheywould aggravate a situation by decreasing follower percep-