READY, AIM, ACQUIRE: IMPRESSION OFFSETTING AND ......READY, AIM, ACQUIRE: IMPRESSION OFFSETTING AND ACQUISITIONS SCOTT D. GRAFFIN University of Georgia JERAYR (JOHN) HALEBLIAN University

r Academy of Management Journal2016, Vol. 59, No. 1, 232–252.http://dx.doi.org/10.5465/amj.2013.0288

READY, AIM, ACQUIRE: IMPRESSION OFFSETTING ANDACQUISITIONS

SCOTT D. GRAFFINUniversity of Georgia

JERAYR (JOHN) HALEBLIANUniversity of California—Riverside

JASON T. KILEYOklahoma State University

Drawing on expectancy violation theory, we explore the effects of anticipatory im-pression management in the context of acquisitions. We introduce impression offsetting,an anticipatory impressionmanagement technique organizational leaders employ whenthey expect a focal event will negatively violate the expectations of external stake-holders. Accordingly, in these situations, organizational leaders will announce the focalevent contemporaneously with positive, but unrelated information. We predict im-pression offsetting will generally occur in the context of acquisitions, but also morefrequently for specific acquiring firms and acquisitions that are more likely to lead to anexpectancy violation. We also posit that offsetting will effectively inhibit observers’perceptions of events as negative expectancy violations by positively influencingshareholder reactions to acquisition announcements. Consistent with our hypotheses, ina sample of publicly traded acquisition targets, we find evidence for impression off-setting, in which characteristics of both acquirers and their announced acquisitionspredict its frequency of use. We also find evidence that impression offsetting is effica-cious; on average, it reduces the negative market reaction to acquisition announcementsby over 40%, which translates into approximately $246million in market capitalization.

When external stakeholders interact with a firm,they develop expectancies about organizationalleaders’ decisions. Meeting these expectations isa means by which organizational leaders maintaintheir position (Puffer & Weintrop, 1991) and buildtheir reputations (Fombrun, 1996). When organiza-tional leaders act in a manner inconsistent withhow stakeholders expect the organization to behave,stakeholdersmay perceive such inconsistent actionsas “expectancy violations.” Consequently, expectancyviolations trigger a range of organizational impression

management activities (seeBolino,Kacmar,Turnley, &Gilstrap, 2008 and Elsbach, 2006, 2012 for reviews).

In prior impressionmanagement literature, scholarsoverwhelmingly focused on reactive impressionmanagement (Bolino et al., 2008; Elsbach, 2006),which consists of activities initiated in response toan expectancy violation intended to positively in-fluence external perceptions of the organization.Events that trigger reactive impression managementare typically (a) negative and (b) unanticipated byorganizations. Perhaps counter-intuitively, given itspotential for effectiveness, this prior impressionmanagement research has only rarely consideredanticipatory impression management (AIM), inwhich organizations anticipate a negative reaction toan event that is not yet known by stakeholders. Thisform of impression management occurs in anticipa-tion of an event that may lead to an expectancy vio-lation (Elsbach, Sutton, & Principe, 1998). In thisstudy, we examine a previously unexplored form ofanticipatory impression management in which or-ganizational leaders seek to offset perceptions of

The authors gratefully acknowledge the insight andsupport provided by our editor, Tim Pollock, and threeanonymous reviewers.We also acknowledge the guidancereceived by seminar participants at the University of Vir-ginia, Arizona State University, Texas A&MUniversity, LaUniversidad de Especialidades Especial Santo, Universityof Oxford, and Rice University. Finally, we are grateful forthe financial support thatwas provided by a Terry-Sanfordresearch grant from the Terry College of Business at theUniversity of Georgia.

232

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 expresswritten permission. Users may print, download, or email articles for individual use only.

http://dx.doi.org/10.5465/amj.2013.0288

potentially negative expectancy violations—andassociated negative market reactions—by con-temporaneously releasing positive, but unrelated,announcements.

FirmsuseAIM techniques toproactively influencefirm evaluations. Prior research, however, has ex-amined only two forms of AIM. One form, “strategicnoise,” occurs when organizational leaders are un-clear about the market reaction an event will en-gender (Graffin, Carpenter, & Boivie, 2011). Strategicnoise involves obfuscating the connection betweenthe event and the market reaction by releasing posi-tive and negative information, so observers cannoteffectively evaluate the event in isolation, inhibitingperceptions of an expectancy violation. A secondform, “big bath” accounting, occurs when orga-nizational leaders announce significant bad news,usually in the form of disappointing earnings, con-temporaneously with other negative news, such aswrite-downs, to exaggerate a negative expectancyviolation (e.g., Elliott & Shaw, 1988).

In this study,we introduce “impressionoffsetting”as an AIM tactic. We define impression offsetting as:organizational actions initiated to positively influ-ence external perceptions of the organization byreleasing positive, but unrelated, information, inanticipation of an event becoming known that maynegatively violate external stakeholders’ expecta-tions. Offsetting thus takes the form of contem-poraneously releasing one or more positive, butunrelated, announcements. Since shareholdersoften react negatively to acquisition announce-ments (Haleblian, Devers,McNamara, Carpenter, &Davison, 2009), we theorize and test whether or-ganizations release positive information to “offset”the likelihood their announcement engenders anegative expectancy violation. To further test theunderlying theoretical mechanisms, we posit thatmore severe expectancy violations should increasethe frequency with which impression offsetting isemployed. Specifically, in this context, we suggestthat the riskier the acquisition, in terms of the antic-ipated stock market reaction to the announcement,the more likely organizational leaders will employthis tactic.

Our research has the potential to make severaltheoretical contributions to the impression manage-ment literature. First, we introduce the concept ofimpression offsetting, which is a previously un-explored AIM tactic. As part of this concept in-troduction, we organize previous AIM work, whichhas been conducted in isolation in various liter-atures, and we explain how impression offsetting

differs from other AIM techniques. Second, usingexpectancy violation theory, we predict a generalcondition under which negative expectancy viola-tions are likely. Accordingly, we test whether thereare prescriptive expectancies, based on generalnorms, that influence the likelihood that observerswill perceive an event—acquisitions—as a negativeexpectancy violation, and, thus lead to impressionoffsetting. We then focus our analysis to specificconditions under which observers are more likely toperceive an acquisition as a negative expectancy vi-olation. Hence, we also test if predictive expectan-cies, which are based on specific actor or situationcharacteristics, influence the increased likelihoodobservers will perceive a particular acquisition asa negative expectancy violation. Thus, our pattern offindings, as guided by expectancy violation theory,helps us better understand when and how offsettingis used. Third, we find that impression offsettinginfluences the stock market reactions to acquisitionannouncements. Despite the fact that researchershave documented that organizational leaders ac-tively engage in impression management activities,we know very little about the effectiveness of suchactivities on organizational outcomes (see Bolinoet al., 2008 for a recent review; see also Zavyalova,Pfarrer, Reger, & Shapiro, 2012). We find that im-pression offsetting attenuates the negative stockmarket reaction to an acquisition announcement,which, we argue, suggests inhibited perceptions ofa negative expectancy violation.

THEORY AND HYPOTHESES

Reactive and Anticipatory ImpressionManagement Surrounding Expectancy Violations

Expectancy violation theory suggests observershold expectations regarding how an actor shouldbehave in a given situation (e.g., Burgoon, 1993;Burgoon, Stern, & Dillman, 1995). According to thistheory, the term expectancy refers to an, “enduringpattern of behavior” (Burgoon, 1993: 31) to which anactor is expected to conform. These expectanciesare broadly classified as prescriptive or predictive(Burgoon & Hale, 1988; Floyd, Ramirez, & Burgoon,1999). Prescriptive expectancies are based on gen-eral norms for typical behavior (Burgoon & Hale,1988), whereas predictive expectancies refer to an-ticipated behaviors for specific actors or for specificsituations (Floyd et al., 1999; Kim, 2014). When be-havior is consistent with either type of expectancy,it provides observers little new information. In contrast,

2016 233Graffin, Haleblian, and Kiley

failure to conform to either typeof expectancy resultsin an expectancy violation that is “distracting andredirects attention toward the actor and the viola-tion” (Burgoon, 1993: 35). Expectancy violations arepositive or negative, depending on if an actor’s be-havior exceeds or violates the expectation (e.g., firmexceeds or misses earnings expectations). Researchalso suggests that when observers perceive inten-tionality on the part of the actor, the magnitude of anexpectancy violation, whether positive or negative,is amplified (Bachman & Guerrero, 2006).1

In an effort to manage perceptions of—and re-actions to—negative expectancy violations, firmsengage in impression management (Elsbach, 2006).Elsbach (2012) notes that two important factors inorganizing and understanding impression manage-ment activities relate to its timing (i.e., anticipatoryor reactive) and its valence (i.e., positive, neutral, ornegative). Using Elsbach’s (2012) factors, we reviewprior research on impression management. Reactiveimpression management occurs when organiza-tional leaders engage in impression managementactivities after a negative expectancy violation isknown to stakeholders with the intent of reducingstakeholders’ negative reactions to the event itself.Hence, the goal of reactive impression manage-ment is to, “provide clear, rational explanationsfor an organization’s actions. . . and focus on visiblechanges the organization has instituted to preventa similar crisis from occurring in the future”(Elsbach, 2012: 467). This tactic involves providingadditional, typically self-serving, information aboutan expectancy violation in an effort to lessen thenegativity of ongoing stakeholder reactions (e.g.,Porac, Wade, & Pollock, 1999). For instance, re-searchers found that organizational leaders tend toblame external forces for poor performance (Bettman&Weitz, 1983; Staw,McKechnie, & Puffer, 1983) andprovide self-serving accounts to justify compensa-tory decisions in response to the release of executivecompensation information (Porac et al., 1999; Zajac& Westphal, 1995). Reactive impression manage-ment thus involves the following sequence: (1) anevent becomes known; (2) stakeholders perceivea negative expectancy violation (i.e., evidenced by

a strong negative reaction to the violation); and (3)organizational leaders engage in impression man-agement tactics in an attempt to influence ongoingstakeholder reactions by providing additional in-formation about the violation. The goal is for leadersto reduce the perceived severity of the expectancyviolating event by providing a justification or ex-planation for their actions (Elsbach, 2006, 2012).

In contrast to reactive impressionmanagement,wefocus on AIM techniques. Consistent with prior re-search (e.g., Graffin et al., 2011) we define AIM asactivities that are undertaken in anticipation of, orcontemporaneously with, an event that organiza-tional leaders believemay be perceived as a negativeexpectancy violation. Unlike reactive impressionmanagement, which typically provides additionalinformation about a known expectancy violation(e.g., Zavyalova et al., 2012), AIM provides in-formation that isunrelated to the focal event. Indeed,if organizational leaders provide event-related in-formation before observers knowabout the violation,the observers may retrospectively view that in-formation as self-serving, which could negativelyaffect stakeholder impressions (Hovland, Janis, &Kelley, 1953). The contemporaneous aspect of AIMensures that stakeholders must consider multiplepieces of information at the same time, which pre-vents an evaluation of the focal event in isolation(Graffin et al., 2011).

AIM involves the following sequence: (1) organi-zational leaders become aware of a potential expec-tancy violation before it becomes known to externalstakeholders; (2) organizational leaders engage inimpression management tactics either prior to orcontemporaneously with the event; (3) stakeholdersreact to both the event and the other information theorganization released. Table 1 provides a summaryof reactive impression management and AIM. Wefirst discuss two forms ofAIM that have been studiedin prior work, and then introduce a third form: im-pression offsetting.

Strategic noise involves releasing positive andnegative information prior to or contemporaneouslywith a focal event to which the anticipated marketreaction is unclear (e.g., Graffin et al., 2011). Theonly empirical study to examine strategic noise didso within the context of CEO successions. Graffinand colleagues suggested the goal of this tactic is to“minimize direct scrutiny of the event” (2011: 749).Executives reported their motivation for employingstrategic noisewas to allowmore time to evaluate thenewly appointedCEOonhis or her ownmerits ratherthan being “unduly constrained by what the market

1 In the interest of presenting the full range of expec-tancies and expectancy violations, we mention positiveand negative expectancy violations in our review ofexpectancy violation theory. The focus of our study,however, is on negative expectancy violations and ourdiscussion for the remainder of the paper focuses on thisaspect of expectancy violation theory.

234 FebruaryAcademy of Management Journal

‘said’ about their potential” (Graffin et al., 2011: 751).Thus, the goal for strategic noise is to obfuscate thelink between a focal event and themarket reaction inorder to give the event time to unfold and be evalu-ated on its own. Organizational leaders could thusemploy this tactic in other anticipated occurrencesto which the expected market reaction is uncertainsuchas entering anewgeographicorproductmarket.

Big bath accounting occurs when organizationalleaders anticipate a potential negative expectancyviolation—typically disappointing performance—by engaging in a one-time effort to stimulate per-ceptions of a negative expectancy violation andoverstate the negativity of the event by releasingother negative announcements prior to or contem-poraneously with it (e.g., Elliott & Shaw, 1988).Specifically, firmswill “under-report earnings by themaximum amount possible, preferring to take a bigbath in the current period in order to report higherearnings in the future” (Kirschenheiter & Melumad,2002: 762). All bad news is released at once because“when circumstances are bad, making things justa little worse by cleaning out the rubbish does littleharm to either reputation or prospects” (Walsh,Craig, & Clarke, 1991: 174). This approach allowsmanagers to reserve funds for future periods todemonstrate a comparative increase in profit (Walshet al., 1991), signal to shareholders that they haveaggressively dealt with past problems (Strong &Meyer, 1987: 644), and to lower the future perfor-mance benchmark (Elliott & Shaw, 1988). Other ex-amples of negative expectancy violations that maytrigger the use of big bath accounting include bank-ruptcy or earnings misstatements.

Finally, we introduce a previously unexploredform of AIM—impression offsetting.We suggest thatorganizational leaders will engage in impressionoffsetting when they anticipate an upcoming eventmay lead to a negative expectancy violation. In aneffort to reduce the perceived size of this potentialnegative expectancy violation, organizational leadersmay release positive, but unrelated, information priorto or contemporaneously with the event becoming

known to external stakeholders (see Appendix A forthe list of offsetting announcements in our sample).Impression offsetting is thus an attempt to manageimpressions positively by offsetting an anticipatednegative reaction to an event, which reduces thelikelihood that observers will perceive it as a neg-ative expectancy violation. Expectancy violationtheory suggests violations lead to increased scrutinyfor the violator and cause observers to seek addi-tional information regarding the violation (Burgoon,1993). Thus, the primary motivation for offsetting isto inhibit perceptions of a negative expectancy vio-lation by reducing the perceived size of a negativereaction. Although releasing this unrelated positiveinformation may not directly change how investorsreact to the acquisition announcement, it does forcemarket participants to react to multiple pieces ofnews. Thus, by compelling investors to react both toa potential negative expectancy violation as well asto the offsetting announcement(s), the net effect is toreduce the overall negativity of the market reactionto the acquiring firm by requiring market partici-pants to react to multiple pieces of news simulta-neously. By doing so, organizational leaders mayalso diminish the additional scrutiny and undesir-able attention that typically follows negative expec-tancy violations.

Releasing positive information that is unrelatedhelps avoid the potential pitfalls associated withproviding further information to the focal event asprior impression management research suggests au-diences may view attempts to justify a focal eventwith skepticism (Tedeschi & Reiss, 1981). A key as-pect of a justification is an argument that the event’sperceived negative consequences are actually posi-tive.Adanger in such a justification is that audiencesperceive organizational leaders as “protesting toomuch” and thus view the organization more nega-tively than they would had the event not been justi-fied (see Ashforth & Gibbs, 1990). More generally, ifaudiences question the truthfulness of the justifica-tion, they may interpret it as deceptive and perceivethe organization as dishonest (e.g., Schneider, 1981).

TABLE 1Reactive versus Anticipatory Impression Management

Anticipatory Reactive

Timing of IM Before or contemporaneous with the focal event After the eventFocus of IM Away from the focal event On the focal eventIntent of IM Inhibit or stimulate expectancy violations related

to the focal eventReduce negative reactions to the focal event


As such, releasing unrelated positive informationmay allow organizational leaders to offset a poten-tially negative shareholder reaction without engen-dering the skepticism related to justifications. Table 2provides a summary comparison of these three AIMtactics.

As we noted earlier, observer expectancies can bebroadly classified as prescriptive (i.e., general ex-pectancies), or as predictive (i.e., specific to an actoror situation) (Burgoon&Hale, 1988). In the context offirms, investors hold the prescriptive expectancythat organizational leaders should act in a mannerthat increases profits (Friedman, 1970) and enhancesshareholder value (e.g., Fligstein, 1990). When or-ganizational leaders behave in amanner that violatesthis prescriptive expectancy, a negative expectancyviolation occurs. For instance, Zavyalova andcolleagues (2012) note that a product recall isa form of a negative organizational expectancyviolation because it reduces the positive mediacoverage firms receive, triggering firms to activelyengage in reactive impression management. Fur-ther, if the negative expectancy violation is viewed asintentional by observers, the magnitude of the per-ceived negative violationwill be amplified (Bachman&Guerrero, 2006). For example, if a firm fails tomeetsits earnings expectations, but the reason(s) for itsdisappointing performance is unclear or ambiguous,a minor negative expectancy violation may occur. Ifa firm intentionally engages in a behavior that ob-servers associate with impairing firm performance,however, the observers will perceive a more signifi-cant negative expectancy violation.

We expect organizational leaders to engage inimpression offsetting in an effort to reduce the like-lihood observers perceive an event as a negativeexpectancy violation. We test our theory in the con-text of acquisition announcements because researchsuggests the market typically reacts negatively toacquisition announcements (Haleblian et al., 2009)and this reaction is more consistently negative for

larger, publicly-held, targets (Moeller, Schlingemann,& Stulz, 2004). Thus, when firms undertake largepublic acquisitions, shareholders will likely reactnegatively to such announcements because they trig-ger a prescriptive negative expectancy violation. Duetowidely-reported figures suggesting the failure rateofacquisitions is between 70 and90% (e.g., Christensen,Alton, Rising, & Waldeck, 2011), as well as academicresearch suggesting acquisitions generally destroyshareholder value (e.g., Haleblian et al., 2009), man-agers likely anticipate anegative shareholder reaction.Further, due to the analyses of valuation and strategicfit, as well as the required due diligence that precedeslarge public acquisitions, shareholders will almostcertainly interpret acquisitions as intentional, whichlikely amplifies the prescriptive negative expectancyviolation (Bachman & Guerrero, 2006). In sum, thisresearch motivates organizational leaders to inhibitthe likelihood shareholders will perceive an acquisi-tion announcement as anegative expectancyviolationand their advanced knowledge regarding the an-nouncement of an acquisition gives them the oppor-tunity to engage in impression offsetting.

At the same time, however, reactive impressionmanagement techniques will not likely be effectivein this context. Specifically, organizational leaderswill likely not be able to positively sway themarket’sreaction to an acquisition announcement by pro-viding a post-hoc justification that attempts to castthe acquisition in a more favorable light. Indeed,as we noted earlier, when observers impute self-interest, after the fact justifications do not producepositive results (Schneider, 1981). In our context,organizational leaders may not be able to justify anacquisition by reactively providing additional rele-vant information, because most of the information isalready publicly available.

AIM techniques may be more effective, however.Decision makers could positively bias stakeholders’overall impression of the organization by announc-ing positive unrelated news such as an increase in

TABLE 2Anticipatory Impression Management Techniques

Strategic Noise Big Bath Impression Offsetting

Anticipated reaction No clear expectancy: positive or negative Negative expectancy violation Negative expectancy violationOrg. IM Response Negative to positive Negative PositiveGoal of IM Obfuscate: Make link between event and

market reaction unclear, inhibitingexpectancy violations ofeither valence

Exhaust all bad news: Stimulatenegative expectancy violationand increase negative marketreaction

Positively influence: Inhibit negativeexpectancy violation andreduce negative marketreaction


dividends or by announcing a new product contem-poraneously with the acquisition announcement.This positive bias may, in turn, inhibit perceptions ofa prescriptive negative expectancy violation and re-duce the amount of additional scrutiny and attentionthat typically follows expectancy violations by re-ducing the magnitude of a perceived violation. Wethus suggest:

Hypothesis 1. Organizational leaders will re-lease unrelated positive information aroundacquisition announcements more often than ispredicted by organizations’ baseline count ofpositive announcements.

Suppressing the Negative Expectancy Violation

Inherent in the logic associated with the argu-ments above is that the use of impression offsettingwill positively influence the market reaction to ac-quisition announcements.We contend that unrelatedpositive information released through impressionoffsetting generates additional information in themedia. This additional information forces investorsto assess and likely “average” multiple pieces of in-formation simultaneously. As such, the unrelatedpositive information reduces the magnitude of thenegative reaction to the expectancy violation, in theformof an acquisition announcement.More broadly,despite numerous studies examining the use of or-ganizational impression management, there is verylittle evidence regarding its effectiveness (Bolinoet al., 2008). Our context, however, allows us to ex-amine a quantifiable measure of the effectiveness ofimpression offsetting: the abnormal market returnfor the acquiring organization following the an-nouncement of an acquisition. Indeed, as we notedearlier, when actors attempt to manage stakeholderimpressions by releasing information that is directlyrelevant to a focal event, such announcementsare typically met with skepticism (e.g., Tedeschi &Reiss, 1981). This skepticism results from observersknowing the organization has a vested interest inpositively “spinning” news to make observers per-ceive a focal event more favorably. By releasing un-related positive information, however, such skepticismmay be reduced. Also, the newly released impressionoffsetting information means the market will havenew, positive information to process. By doing so,the net combined reaction to the impression off-setting and the acquisition announcement reducesthe likelihood of a strong market repudiation. Thereduced magnitude of the reaction, in turn, reduces

the likelihood that investors will perceive the ac-quisition as a negative expectancy violation.We thusexpect impression offsetting will positively influencethe market reaction to an acquisition announce-ment by forcing the market to react to the positiveoffsetting announcement(s) as well as the acquisitionannouncement and hypothesize:

Hypothesis 2. Releasing unrelated positive infor-mationwillbepositivelyassociatedwithabnormalreturns around acquisition announcements.

Impression Offsetting and Predictive Expectancies

Next, we test conditions under which organiza-tional leaders are more or less likely to engage inimpression offsetting. We do so to more directly testthe theoretical mechanisms that drive the use ofthis previously unexplored impressionmanagementtactic. Since we propose that the anticipation of anexpectancy violation drives the use of this tactic, itlogically follows that the greater the expected viola-tion, the more likely organizational leaders will en-gage in impression offsetting.

Such contentions are consistent with expectancyviolation theory which proposes that, in addition toprescriptive expectancies, which are based on gen-eral normative behavior across firms, observers alsohold predictive expectancies that an actor’s priorbehavior (Burgoon & Hale, 1988) or specific context(Floyd et al., 1999) can influence. In the followingfour hypotheses, we examine how such predictiveexpectancies, which influence the degree to whicha given occurrence will lead to an expectancy vio-lation, affect the use of impression offsetting. Withinour context, although markets often view acquisi-tions negatively, there is significant variation inperceived expectancy violations depending oncharacteristics of the firm and on context associatedwith the deal (Haleblian et al., 2009). Accordingly,acquisitions allow for the examination of conditionsunder which organizations are more or less likely toengage in impressionoffsetting as a given acquisitionmay trigger stronger or weaker negative expectancyviolations. We restrict our analyses to acquirer andtarget characteristics previously examined withinthe acquisitions literature. Specifically, we examineacquirer and target related variables that increase therisk associated with an acquisition, which makea perceived expectancy violationmore likely. Alongthese lines, we chose the variables of acquirerprior acquisition activity (e.g., Haleblian, Kim, &Rajagopalan, 2006; Hayward, 2002), acquirer reputation


(e.g., Saxton & Dollinger, 2004) target relatedness(e.g., Barney, 1988;Morck, Shleifer, &Vishny, 1990;Seth, 1990) and target pre-merger profitability(Capron&Shen, 2007).Aswedevelop below, as eachof these acquirer and target related variables increaseacquisition risk, they also increase the likelihood ofimpression offsetting.

Acquirer prior acquisition activity. As a firm’sprior actions inform its predictive expectancies(e.g., Burgoon, 1993), we suggest one particularlysalient source of such expectancies: firm acquisitionhistory. This contention is consistent with expec-tancy violation research that suggests as norm de-viating behavior accumulates through behavioralrepetition, a larger negative expectancy violationreaction is generated as compared to a single be-havior (e.g., Bachman & Guerrero, 2006). Repeatedviolations of investor expectations are thought toreduce stakeholder trust (Fehr & Gachter, 2000) inwhich stakeholder skepticism of subsequent viola-tions increases with behavioral repetition (Pfarrer,DeCelles, Smith, & Taylor, 2008; Simpson, 2002).Together, these ideas suggest that if a given actorrepeatedly violates a given expectation, observerswill perceive each subsequent violation more se-verely. Consistent with this tenet of expectancy vi-olation theory, prior researchers found that as thelevel of a firm’s acquisition activity increases,marketparticipants perceive the firm as carrying excessivegrowth related risk, and as such the market’s re-sponse to their subsequent acquisition activity be-comes more negative (Laamanen & Keil, 2008).

Sincemarkets see theacquisitionsof serial acquirersmore negatively than less frequent acquirers, theirannounced acquisitions should be more likely tostimulate negative expectancy violations as comparedto other acquirers. Moreover, from the perspective ofthe serial acquirer, as it increases its acquisition ac-tivity, and thushasahistoryof repeatedlyviolating theexpectation of engaging in activities that promote or-ganizational performance, it may learn that marketstypically perceive acquisitions negatively (Haleblian& Finkelstein, 1999), especially those of experiencedacquirers (Laamanen & Keil, 2008). Hence, basedon awareness of the negative perception of marketsthat derives from direct experiential learning, serialacquirers will be more prone to initiate impressionoffsetting than less experienced acquirers in order tocounteract negative perceptions associated withacquisition activity. Thus, relative to less experi-enced acquirers, serial acquirers generally havemore experience with negative acquisition out-comes, which results in an added motivation to

positively influence stakeholders to defend, andpotentially build, its identity as a competent serialacquirer. Combined, these ideas suggest:

Hypothesis 3.A firm’s recent acquisitionactivitywill be positively associated with releasing un-related positive information around acquisitionannouncements.

Acquirer reputation.We also argue that investorshold predictive expectancies for high reputationfirms.While some research suggests high-reputationactors are held to higher standards and are morestrongly repudiated formissteps than low-reputationfirms (e.g., Milbourn, 2003; Wade, Porac, Pollock, &Graffin, 2006), other studies suggest reputation pro-tects firms and reduces the negative outcomes experi-enced by such firms (e.g., Pfarrer, Pollock, & Rindova,2010).Weexpect, however, that organizational leadersat high reputation firmsmay bemore concerned aboutthe potential downside associated with a negative ex-pectancy violation and are thus more likely to engagein impression offsetting for two reasons.

First, because firm reputation is based upon theability to consistently deliver quality (Lange, Lee, &Dai, 2011), shareholders will likely hold the pre-dictive expectancy that high-reputation firms willcontinue to perform at a high level. Engaging in ac-quisitions, which are fraught with risk and oftendestroy shareholder value, is inconsistent with thispredictive expectancy. Indeed, as an actor’s prioractions influence predictive expectancies (e.g.,Floyd et al., 1999), shareholders expect higherreturns fromhigh reputation actors (Fombrun, 1996).Along these lines, prior research has shown thathigh-reputation firms engage in riskier investmentsthan other firms in order to meet the high expecta-tions they faced (Petkova, Wadhwa, Yao, & Jain,2014). Therefore, shareholders will more likely per-ceive risky acquisition actions from these high rep-utation actors as expectancy violations than theywould from firms held to lower expectations.

Second, based on the quality of their abilities andtheir outputs, high-reputation organizations benefitfrom substantial accumulated levels of public rec-ognition (see Barnett & Pollock, 2012 for a review ofthis literature). Given these benefits, high-reputationorganizations may have an enhanced motivationto manage impressions of their acquisitions toprotect their reputations by avoiding the percep-tion of a negative expectancy violation. Accordingly,Burgoon and Hale (1988) argued that positivelyevaluated actors have the potential to commit amoregrievous violation because the gap between actual


and expected behavior is likely greater for such ac-tors. Consistent with this idea, Rhee and Haunschild(2006) found that high-reputation automakers expe-rienced larger losses in market share from negativeimpressions than low-reputation organizations. Ac-cordingly, high-reputation organizations may bemore likely to attempt to mitigate an anticipatednegative market reaction—such as impressionoffsetting—given the potential for stakeholders to de-monstrate strong reactions. Thus, we hypothesize:

Hypothesis 4. High-reputation organizationswill be more likely than other organizations torelease unrelated positive information aroundacquisition announcements.

Target relatedness. Specific situation character-istics also inform predictive expectancies (Burgoon,1993). Based on this tenet of expectancy violationtheory,weargue thatmanagerswill perceive ahigherlikelihood that investors will view an acquisition asa negative expectancy violation as the risk associatedwith the target increases. Prior research suggests,within the context of acquisitions, there is significantvariation in the likelihood of a positive outcome(e.g., Pablo, 1994). As acquisitions increase in targetrisk, managers expect investors to recognize thisdifficulty and react negatively to the announcement.Although all acquisitions carry significant risk(Eisenmann, 2002), unrelated acquisitions are gen-erally considered riskier (Morck et al., 1990) and arethus more likely to lead to a negative expectancyviolation. Moreover, acquirers that have knowledgeof their own industry or industriesmay be less able toassimilate and use knowledge from a target organi-zation in a different, unrelated industry (Cohen &Levinthal, 1990; Vermeulen & Barkema, 2001). Con-sistent with this notion, prior research shows less-related acquisitions tend to produce more negativemarket reactions than more-related acquisitions(Morck et al., 1990; Seth, 1990).

As such undertaking an unrelated acquisitionrepresents a stronger potential predictive expec-tancy violation than the case of a related acquisition.Expectancy violation theory recognizes the valenceof the violation is an important determinant inwhichthe stronger the negative behavior, the more likelyobservers will perceive it as a negative expectancyviolation (Burgoon & Hale, 1988). Moreover, share-holderswill likely associate undertaking sucha riskyacquisition as intentional by organizational leadersin which intentional violations are typically associ-ated with increased evaluations of responsibilityand negative evaluations (e.g., Bachman & Guerrero,

2006: 946; Fincham & Bradbury, 1987). Thus, as thelikelihood for a predictive negative expectancy vio-lation increases for an acquisition, organizationalleaders will become more motivated to act pre-emptively in order to influence themarket’s reactionthrough the use of impression offsetting. Along theselines we hypothesize:

Hypothesis 5. Target unrelatedness will be posi-tively associated with releasing unrelated positiveinformation around acquisition announcements.Target pre-merger profitability. In a similar vein,

as with unrelated targets, unprofitable targets alsorepresent a high degree of acquisition risk, because itis difficult to turn around the performance of a firmexperiencing poor performance (Capron & Shen,2007). More specifically, if a target organization iscurrently profitable, an acquirer may be effectivelyacquiring a revenue stream and some opportunity toincrease efficiency by consolidating support func-tions such as marketing and accounting. A target’sprofitability may also signal that it has valuable re-sources that the acquirer can use (see Capron&Shen,2007). In this instance, it might be relatively easyfor organizational leaders to highlight the ongoingprofitability of the organization as part of the justifi-cation for entering into a relationship with a targetfirm. In contrast, acquiring an unprofitable targetcombines the challenge of integrating the two com-panies as well as transforming an acquired un-profitable organization into a profitable componentof the acquiring firm. Accordingly, organizationsmay anticipate acquiring an unprofitable target,which inherently involves greater risk than acquir-ing a profitable one. As such, it is more likely to leadto a negative expectancy violation than a profitabletarget. Indeed, a lack of profitability suggests that theburden for generating potential synergies with theacquiring organization increases, in that leaving anunprofitable target to operate independently is likelyto be value-destroying for the acquiring firm. In ad-dition, the challenge involved in creating a relation-shipwith an unprofitable acquired organizationmaylead markets to react more negatively to targets withhigher risk profiles. These factors also lead to anenhanced likelihood of a perceived negative expec-tancy violation; hence we expect acquirers will bemore motivated to use impression offsetting whenthey acquire an unprofitable organization.

Hypothesis 6. Unprofitable targets will be morelikely than profitable ones to be associated withreleasingunrelatedpositive informationaroundacquisition announcements.


DATA AND METHODS

Sample and Data Collection

Our sample consists of acquisition announcementsfor deals in which both the acquirer and target werepublic companies. In addition, these deals met thefollowing criteria and were: (a) greater than $100million in value; (b) announced between 1995 and2009; and (c) subsequently completed. Our sampleconsists of 770 acquisition announcements fitting ourcriteria. Missing data reduces our sample to 758 formodels predicting impression offsetting and 701 formodels predicting cumulative abnormal returns.

We collected data from Thomson Reuters SDCPlatinum for each acquisition announcement in-cluding the date, acquirer, target, acquisition price,form of consideration offered, acquirer and targetindustries, and premium paid. Using Compustat, wealso collected data on the acquirers’ organizationsize, profitability, and level of diversification. FromRisk Metrics, we collected data on the acquirers’board characteristics. We employed ThomsonReuters I/B/E/S to capture whether earnings announce-ments indicate negative, zero, or positive earningsand whether those earnings miss, meet, or beatconsensus estimates.Wealso collectedandanalyzedpress releases to examine impression offsetting be-haviors and baseline counts of positive announce-ments by acquirers in our sample.We report variablemeans, standard deviations, and correlations inTable 3.

Press release collection. We collected press re-leases in twowaves. The firstwavecovered the three-day event window, covered the period from one daybefore to one day after the event for each acquisition,and the second wave, which we collected for ourbaseline positive announcements measure, coveredthe period from four months prior to the event win-dow through one month prior to the event window.

Event window press releases. For our eventwindow collection, we collected and analyzed 4,969press releases to isolate and code those from theacquirers in our sample. Our primary sources of datawere PR Newswire and Business Wire, the twodominant press release distributors. In the eventwindow collection, we accessed PR Newswire andBusiness Wire using both Factiva and LexisNexis.After substantial investigation, we found that Busi-ness Wire was substantially identical between Fac-tiva and LexisNexis, but PR Newswire in LexisNexiscovered more organizations. Accordingly, we gath-ered data for 666 acquisitions fromFactiva databasesand an additional 77 acquisitions from LexisNexis

databases (i.e., a total 743 acquisitions out of 770 or96.5%). As an additional robustness check, we latercompared our Factiva data (supplemented withLexisNexis data for 77 acquisitions) to LexisNexisdata for the full sample covering the same three-daywindow and found them to be substantially similaracross the entire sample (r 5 .98; p , .001). We alsoused organization websites when we could not findpress releases in PR Newswire or Business Wire. Wegathered data for 24 acquisitions directly from orga-nization press websites and an additional three ac-quisitions usingGoogle searches to find press releasepages on organization websites that were not listedon their press websites.

Based on the content of the collected press re-leases, we developed content categories. Thoughmost press releases cover one news announcement,we also include those that have multiple types ofnews in one press release (e.g., an earnings releasethat also includes an executive retirement anda revision of previously issued guidance for a futureperiod). We provide a list of all categories andthe number of occurrences of each category inAppendix A.

Baseline press releases. For our baseline collec-tion,wegatheredover 300,000press releases and,weused a subset of 53,200 press releases that fit a three-month date range for our Baseline positive an-nouncements control that we describe below. Likeour event window collection, our primary sources ofdata are PRNewswire andBusinessWire. Consistentwith our event window collection, we accessed PRNewswire and Business Wire using LexisNexis,having learned in our event window collectionthat the PR Newswire database in LexisNexiscovers more organizations. We gathered baselinedata for 765 organizations from LexisNexis data-bases (i.e., 765 acquisitions out of 770 or 99.4%).Like our event window collection, we also usedorganization press websites to gather data for theremaining five acquisitions.

We used computer-assisted technology to facili-tate processing themassive number of press releasesin the baseline collection. In particular, we devel-oped custom software, written in the Python 3.3programming language, to identify and isolate pressreleases released by focal organizations in our sam-ple. Our software uses custom per-organization pat-terns that we developed by manually examiningorganization press releases, coding custom rules,and iteratively checking them against our actualbaseline data. This process eliminated press re-leases that organizations did not directly issue in the


TABLE3

SummaryStatisticsan

dCorrelation

s

Variables

Mea

nSD

12

34

56

78

910

1112

1314

1516

1718

1920

2122

1.CEO

pay

8.93

21.00

41.00

2.CEO

tenure

7.05

07.12

50.03

1.00

3.Outsidedirec

torperce

ntage

0.83

20.09

90.13

20.28

1.00

4.Total

direc

tors

11.760

3.57

90.12

0.07

0.19

1.00

5.Lag

gedROA

0.05

00.07

60.11

0.01

20.10

20.10

1.00

6.Freeca

shflow

0.25

20.23

90.05

0.03

20.01

20.01

0.11

1.00

7.Stock

returns

0.25

31.04

10.06

0.01

20.18

20.12

20.16

20.04

1.00

8.Acq

uirer

size

9.17

61.26

00.48

20.05

0.27

0.25

0.17

0.09

20.07

1.00

9.Acq

uirer

industry

conce

ntration

0.11

40.12

120.04

20.05

0.01

20.10

20.01

20.05

20.04

0.05

1.00

10.D

eman

duncertainty

1213

8.74

930

69.623

0.23

20.05

0.30

20.07

0.01

0.02

20.06

0.26

20.01

1.00

11.B

aselinepositive

annou

nce

men

ts0.09

00.11

20.19

20.06

0.05

20.07

0.03

0.05

0.05

0.14

20.04

0.12

1.00

12.P

rior

acqu

isitionac

tivity

0.91

61.74

30.15

0.18

0.07

0.33

20.03

0.06

20.01

0.15

20.17

0.02

0.04

1.00

13.H

ighreputation

0.07

00.25

50.09

0.01

20.13

20.01

0.21

0.01

0.15

0.26

0.03

20.07

0.17

0.06

1.00

14.P

erce

nts

tock

47.301

43.703

20.10

0.07

20.09

0.25

20.16

20.06

0.10

20.16

20.13

20.31

20.01

0.10

0.06

1.00

15.A

ttitude

0.96

10.19

320.02

0.00

0.04

0.01

20.03

20.02

0.03

20.01

20.01

0.06

0.01

0.05

0.00

0.06

1.00

16.F

riday

annou

nce

men

t0.11

60.32

00.00

0.02

20.02

0.04

20.03

20.02

0.05

0.03

20.04

0.01

0.00

0.00

20.01

0.01

20.02

1.00

17.N

egativean

nou

ncemen

ts0.04

10.19

90.02

20.04

0.09

0.10

0.00

0.04

20.03

0.05

0.02

0.08

20.05

0.11

0.00

0.01

0.04

0.04

1.00

18.N

eutral

annou

nce

men

ts0.09

10.33

00.07

20.04

0.05

0.06

0.07

0.07

20.04

0.18

0.07

0.10

0.06

0.15

0.08

20.02

20.03

20.03

0.03

1.00

19.U

nrelatedness

0.63

90.48

10.11

20.09

0.07

0.08

0.05

20.02

0.01

0.18

0.09

0.00

20.04

0.05

0.04

20.09

20.01

20.01

0.07

0.01

1.00

20.U

nprofitabletarget

0.22

10.41

50.09

20.02

20.03

20.14

0.07

0.02

0.09

0.02

20.02

0.07

0.13

20.02

0.14

20.07

0.05

0.08

0.04

0.05

0.05

1.00

21.Impressionoffsetting

0.54

91.26

20.21

0.03

0.06

20.02

0.06

0.09

20.02

0.35

20.01

0.19

0.27

0.17

0.17

20.14

20.03

20.03

0.02

0.10

0.06

0.08

1.00

22.C

umulative

abnormal

returns

20.01

40.05

90.03

20.05

0.03

0.01

0.10

0.02

0.09

0.06

0.02

20.01

0.01

0.02

0.10

20.16

0.00

0.00

20.07

0.01

0.00

0.06

0.09

1.00

Note:

N5

701.

If|r|>

0.10

,then

p,

0.01

;if|r|

>0.07

,then

p,

0.05

.


sample and reduced our sample to 15,028 press re-leases for the baseline positive announcementsmeasure we describe below and two robustnessmeasures.2 Once processed, we exported the data inspreadsheet form for human coding of our measuresdescribed below.

Dependent Variables

Impression offsetting. We measure impressionoffsetting as the count of positive, material press re-leases issued by the acquiring firm within the eventwindow (i.e., day 21 to day 11). Employing thisthree-day window is consistent with prior researchonAIM (e.g., Graffin et al., 2011) and recognizes that,even if organizational leaders issue a press releaseone day after a focal event, they prepared that releasein advance of the event. Our results and conclusionsare substantively unchanged when we restrict thewindow to day 21 and day 0.

For clearly positive events (e.g., new product re-leases, social good efforts like donations) we codedpress releases as impression offsetting. For eventtypes that may or may not be positive, we looked foradditional information. Specifically, for earningsreleases, we compared announced earnings to con-sensus expectations to categorize them as positive,and thus as impression offsetting, or neutral or neg-ative, which are not coded as impression offsetting.For other events like divestitures, we coded pressreleases as impression offsetting if the organizationmade an unambiguously positive characterizationof the event in the text of the press release. In total,we categorized 415 events as positive, 49 as neutral,and 56 as negative. We provide a full list of pressrelease categories in Appendix A. Themost frequentcategories were new product announcements (197),

customer wins (79), earnings releases (51), socialgood events (41), and other acquisition announce-ments (34).

We used a primary coder to code all 4,969 pressreleases intooneormoreof the categories. Inaddition,we assessed the reliability of our primary coder byusing a secondary coder for 100 randomly chosenpress releases excluding those thatwe coded as eitherbeing from other parties or announcing the focal ac-quisition. To assess inter-rater reliability we em-ployedCohen’sk (Cohen, 1968).For the sampleof 100press releases thekwas0.99 andvalues above0.80 areconsidered excellent (Fleiss, 1981; Landis & Koch,1977). Given this high level of inter-rater agreement,wecoded the remainingpress releaseswithonecoder.

Cumulative abnormal returns.Abnormal returnsrepresent the part of the return on a security that isunanticipated by an economic model of expectedreturns for the same security. Cumulative abnormalreturns (CAR) are the sum of daily abnormal returnsfor a security over a period that captures the influ-ence of the event during that period. Specifically, weassessed returns of the security in our sample againstthe return of the value-weighted market portfoliousing the following formula:

CARtðT1,T2Þ5 +T2

t5T1

fRit 2 ðai 1biRmtÞg,

where Rit 5 the return on stock i for day t, Rmt 5 thereturn on the value-weighted market portfolio forday t; ai5 a constant, bi5 b of stock i, and T1 and T2

are the lower and upper limits of the event window,respectively.

We calculate the estimates of a and b during a 250-day window that falls between 295 and 45 days be-fore the focal acquisition. In line with McWilliamsand Siegel’s (1997) guidance on short event win-dows, we report a three-day event window centeredon the acquisition (i.e., day21 to day11). Our CARvariable is the average of the abnormal returns for theevent window. Our results are substantively un-changed to the alternative event windows using thedate of the acquisition (i.e., day 0 only) and a two-daywindow covering the day of the announcement andthe next day (i.e., day 0 to day 11).

Independent Variables

Prior acquisition activity. Consistent with priorstudies (e.g., Laamanen & Keil, 2008; Vermeulen &Barkema, 2002), we measure prior acquisition ac-tivity as the count of acquisitions fitting our samplespecification the acquirer made in the three years

2 Although, we gathered press releases for an entire yearprior to the acquisition (–13 months through 21 month),our initial collection including this time window resultedin over 300,000 press releases. Since this was a massiveamount of data to process and manually code, we investi-gated whether it would be possible to scope the data anal-ysis to amore reasonable size. First, wemanually coded ourthree-monthmeasure. Then,we coded a 12-monthmeasurefor a 10% random subsample as a robustness check. Ouranalyses that showed the number of positive,material pressreleases issuedby firms in the12-monthwindowwashighlycorrelated with the number issued within a three-monthwindow (r5 0.87). Given the likely high level of similaritybetween these time windows as well as the highly labor-intensive nature of the coding, we opted to use our three-month measure.


immediately prior to the announcement date (i.e.,day –[365 * 3] to day 21).

High reputation. Following prior research, wecode acquirer high reputation using annual lists ofFortune’s Most Admired and theWall Street Journaland Harris Interactive’s corporate reputation rank-ings (Pfarrer et al., 2010). We code this variable as 1when the acquiring organization was within the top25 of either list in a particular year and as 0 other-wise. We lagged this variable in all analyses. Asa robustness check, we also created a variable thatcounts thenumberof times anorganization appearedon the list in the prior five years. Our results andconclusions are substantively unchanged when weemploy this alternate variable.

Unprofitable target. We code our unprofitabletarget variable as 1 if the target organization in anacquisition has a zero or negative net income in theirannual report in the year prior to acquisition an-nouncement and as 0 otherwise. To create this vari-able we flagged all acquisitions where the offerprice-to-earnings-per-share (EPS) ratio was missingin the Thomson Reuters SDC Platinum database weused to collect acquisition information. Once wemanually verified that this missing data indicatednegative earnings, we coded it as unprofitable.

Target unrelatedness. We code our unrelatedvariable as 1 for acquisitions inwhich the acquirer andtarget have different two-digit Standard IndustrialClassification (SIC) codes. Our results and conclusionsare substantively unchanged using three- and four-digit SIC codes.

Impression offsetting. We use our impressionoffsetting variable described above as an indepen-dent variable in our test of Hypothesis 2.

Control Variables

Baseline positive announcements. We measurebaseline positive announcements as the three-dayaverage count of positive material announcementsby the focal firm for a three-month period prior tothe acquisition announcement (i.e., day2121 to day230). We coded using a process identical to the onewe used for impression offsetting. In our sample, theoverall average three-day count is 0.0897 for a three-day eventwindow ([1.88333]/63,where 1.883 is theoverall average count for entire baseline periods inthe sample, 3 is the number of days in the eventwindow, and 63 is the number of business days ina three-month period).

As a robustness check, we also coded a 10% ran-dom subsample for similar six-month (i.e., day2213

today230) and12-month (i.e., day2395 today230)measurement periods. The three-month measure ishighly correlated with both the six-month measure(.91) and the 12-month measure (.87), suggestingstrong convergent validity between the three-monthcount and other longer interval counts, which sug-gests the use of the three-month measure as an accu-rate baselinemeasure. In total, we coded 15,028 pressreleases, which included the three-month measureand the two robustness measures discussed above.

Governance controls. We include four acquirercorporate governance controls in our analyses. Wecontrol for CEO pay using Compustat’s total CEOcompensation (TDC1) variable—which we logtransformed—and for CEO tenure as research sug-gests that each may generally influence a CEO’s stra-tegic decisions (Mueller, 1987), aswell as the specifictype of acquisitions a CEO pursues (e.g., Sanders,2001). We also capture outside director percentageand total directors to control for the relative strengthof monitoring by the board of directors as such gov-ernance characteristics have been associated withfirm acquisition activity (Hoskisson &Hitt, 1990).Welag each of these variables in all of our analyses.

Acquirer controls. We also include seven addi-tional acquirer controls inour analyses.Acquiring firmperformance may be associated with acquisition suc-cess (Morck et al., 1990); hence, we control for laggedROA to capture performance in the year prior to theacquisition announcement. Priorwork on acquisitionshas shown free cash flow influences acquisition like-lihood (e.g., Lang, Stulz, & Walkling, 1991), so wecontrol for the percentage of free cash flow, whichwe measure as (operating income – taxes – interestexpense – depreciation – common and preferred stockdividends)/common equity. In addition, it may be thatstock price influences acquisition likelihood (Harford,2005); hence, we control for stock returns, which wemeasure as an organization’s annual stock returns, as-suming reinvestment of dividends. There is also evi-dence that firm size influences acquisition behavior(e.g., Moeller et al., 2004). We control for acquirer sizeby using the log of the acquiring organization’s sales.We transformed this variable due to its skewed distri-bution, and we lagged it in all analyses. Because ac-quisitions in more concentrated industries may beperceived differently bymarkets (seeHou&Robinson,2006), we control for acquirer industry concentrationusing aHerfindahl index of the concentration (of sales)of the acquirer’s three-digit SIC industry code usingdata from Compustat, following Hou and Robinson(2006). Since environmental factors have also beenshown to influence acquisition activity (Harford,


2005), we control for environmental dynamism tocontrol for environmental influences on strategic de-cision making by examining the degree of sales in-stability in the markets in which the acquirer isaparticipant.Consistentwithprior studies (e.g.,Dess&Beard, 1984; Tosi, Aldag, & Story, 1973),wemeasure itas dispersion of sales over the prior five years andlagged it one year in all analyses. Also, in our test ofHypothesis 2, we use our high reputation and prioracquisition activity variables as controls.

Acquisition controls. We include four acquisitionlevel controls in our analyses. Percentage stock used inacquisitiondeals have been shown to influencemarketreactions (e.g., Travlos, 1987); as such we control forpercent stockwhich is thepercentageof the total cost ofthe acquisition that the acquirer pays using its ownstock. In addition, the attitude of an acquisition trans-action has been asserted to influence acquisitionperformance, so we control for the attitude of the ac-quisition classified as whether the acquirer’s bid isfriendly or hostile (Browne & Rosengren, 1987). Also,in our test of Hypothesis 2, we use our unrelatednessand unprofitable target variables as controls.

Announcement controls. We include three an-nouncement controls in our analyses. We measureFriday announcement as one when an acquisition isannouncedonaFridayandzerootherwise.Wecontrolfor Friday announcement to help rule out the alterna-tive explanation that managers will tend to announceanacquisitiononaFridaywhen theyexpect anegativereaction (i.e., “taking out the trash”), which may actas a substitute impression management tactic (e.g.,Dellavigna & Pollet, 2009). We also control for neutralannouncements and negative announcements, eachmeasured as a count variable and coded alongside—but not included in the count of—our impression off-setting variable described above. We control for theseannouncements to rule out the alternative explanationthat other AIM techniques are being used in the ac-quisition context and influence our results.

Year dummies.We include year dummyvariablesin all of our regression analyses to control for time-varying macroeconomic effects. For aesthetic rea-sons, we omit these year dummy variables from ourtables of regression results.

Analysis

We test Hypothesis 1 by comparing our observedcount of positive announcements around acquisitionswith baseline three-day average counts (a) from ourthree-month baseline measure using a paired t-testand (b) from a prior study using a two-sample z-test

for proportions. We test Hypothesis 2 using ordinaryleast squares (OLS) regression with robust standarderrors. We test Hypotheses 3 through 6 using nega-tive binomial regression with robust standard errorsbecause these models employ count dependentvariables with over-dispersion.

RESULTS

In Hypothesis 1, we predicted that impressionoffsetting around acquisitions will occur more oftenthan is predicted by organizations’ baseline count ofpositive announcements. We compare our observedcount of positive event announcements to a baselinethree-day average count of announcements from ourownbaselinemeasure and fromaprior study. For thethree-month period beginning four months priorand ending one month prior to the focal acquisitionevent, we constructed a baseline count of positiveevent announcements. We observed that in thisbaseline period, firms averaged 0.0897 positive an-nouncements every three days. In contrast to thebaseline period, during the impression offsettingwindow, which is also three days, we observed that205 out of 770 acquisitions, or 26.6% of firms, in oursample have at least one positive event announce-ment within the 21 day to 11 day event window.In addition, we found an overall average count of0.5221 positive announcements and our observedcount of positive press releases around acquisitionsis statistically significantly more positive than ourbaseline positive press release count (t 5 10.0619;p,0.001) and suggests that firms release 482%morepositive unrelated news surrounding acquisitionsthan suggested by our baseline. Finally, we alsocompared the number of negative press releases inthe eventwindow to a baseline negative press releasethree-day average count using a paired t-test. Wefound that negative announcements during the eventwindow (mean5 .0377) actually decrease relative tothe baseline three-day average count (mean5 .0627)of negative announcements (t 5 3.1198; p 5 0.000).This result suggests organizations release negativepress releases contemporaneously with acquisitionsin an amount 39.94% lower than suggested by ournegative announcements baseline measure. Thus,offsetting consists of both increased positive an-nouncements and decreased negative announce-ments relative to baseline levels. Combined, theseresults suggest that impression offsetting differs fromstrategic noise, which involves the release of bothpositive and negative announcements and providessupport for Hypothesis 1.


We predicted in Hypothesis 2 that impression off-setting will be positively associated with abnormalreturns around acquisition announcements. Themodels that testHypothesis2 arepresented inTable4.Model 1 in Table 4 presents the control model, whileModel 2 adds our measure of impression offsetting.The coefficient for impression offsetting in Model 2is positive and significant (p , 0.05), supporting Hy-pothesis 2. For instance, when an organization en-gages in impression offsetting at the mean level (twopositive announcements for thoseorganizationsusingany level of impression offsetting) with an acquisitionannouncement, the mean market reaction is reducedfrom21.4% to20.8%.At themeanof our sample thisrepresents apositive changeof $246million inmarketcapitalization and that impression offsetting reducesthe size of the negative market reaction by 44%. Asa robustness check, we also examined Hypothesis 2using a two-stage least squares (2SLS) model, andour results and conclusions were substantively un-changed when we used this alternative model. In our2SLS robustness model, we used shares outstandingand average top management team total compensa-tion as instruments for our impression offsettingvariable.3 Following Semadeni, Withers, and Certo(2014), we conducted two post-estimation tests re-garding the validity of our instruments, both of whichsupported their validity. First, we tested whether ourinstruments are weak using an F test (with Stata’s“estat firststage” post-estimation command). The testwas highly significant (F5 14.96;p, 0.001), rejectingthe null hypothesis of weak instruments. Second, wetested the over-identifying restrictions of the two-stage model using Wooldridge’s x2 test (with Stata’s“estat overid” post-estimation command). The testwas not significant (x2 5 1.21; p 5 0.271), failing toreject the null hypothesis that the instruments areuncorrelated with the structural error term and themodel is properly specified, and thereby supportingthe validity of our two instruments.

Table 5 presents the models that test Hypotheses 3through 6. In Hypothesis 3, we predicted that a firm’srecent acquisition activity will be positively associ-ated with impression offsetting. Model 1 presents thecontrol model, while Model 2 includes all indepen-dent variables predicting strategic offsetting. The co-efficient for prior acquisition activity in Model 2 ispositive and statistically significant (p , 0.01), sup-porting Hypothesis 3. Accordingly, when acquisi-tions in theprior threeyears increase fromthemean inour sample (roughly one acquisition) to one standarddeviation above the mean (roughly three acquisi-tions), the number of offsetting announcements in-creases by approximately 0.4 announcements.

InHypothesis 4,wepredicted that high-reputationorganizations will be more likely to engage in impres-sionoffsetting aroundacquisitionannouncements.Thecoefficient forhigh reputation inModel2 ispositiveandstatistically significant (p , 0.01), supporting Hypoth-esis 4. This finding suggests the number of offsettingannouncements increases by roughly one announce-ment for high reputation organizations relative to non-highreputationfirms.WepredictedinHypothesis5 thatacquisition target unrelatedness will positively influ-ence the likelihood of the acquiring organization en-gaging in impression offsetting around acquisitionannouncements. The coefficient for unrelatedness inModel 2 is positive but not statistically significant, fail-ing to support Hypothesis 5. In Hypothesis 6, we pre-dicted that, when acquisition targets are not profitable,organizations will be more likely to engage in impres-sion offsetting around acquisition announcements. Thecoefficient for unprofitable target inModel 2 is positiveand statistically significant (p , 0.05), supporting Hy-pothesis 6.This result suggests thatwhenanacquisitiontarget is unprofitable, the number of a firm’s positiveunrelated news announcements increases by approx-imately 0.3 releases. We also conducted a series ofrobustness checks that specified impression off-setting as binary (tested with a legit model) or right-censored at 2, 3, 4, and 5. Across these models, wecontinued to see a broad pattern of support for Hy-potheses 3 through6.4We also conducted robustnesschecks in whichwe controlled for the premium paid

3 We selected our two instruments for a number of rea-sons. First, Average TMT total compensation and Numberof shares outstanding may be plausibly related to theamount of scrutiny that a firm’s actions receive from au-diences, and, consequently,maybeplausibly related to thedecision touse impressionmanagement. The signs on eachcoefficient align with this intuitive rationale (see, Murray,2006). They are also plausibly unlikely to have a directinfluence on market reactions to an acquisition an-nouncement. Further, they are available for most of ourobservations, and they satisfy the mathematical relation-ships needed to support their validity and strength as in-struments as we describe above.

4 Weconsidered a robustness checkusing a zero-inflatednegative binomial model, but the descriptive statistics ofour press release raw data suggest that there is no excesszero problem. A total of 98.6% (i.e., 759 of 770) of obser-vations have at least one press release (without regard totone ormateriality) in the baseline period. A total of 77.3%(i.e., 595 of 770) of observations have a positive, materialannouncement in the three-month baseline period.


by the acquirer (as a proportion of the target’s marketvalue), and our robustness results are substantiallysimilar to our main models.

Our results suggest consistent support for our ar-gument that organizational leaders release positiveinformation content to offset potential negative ex-pectancy violations like acquisitions. In addition,our results suggest the use of impression offsettingincreases as characteristics of the acquirer or theacquisition increase the likelihood of an anticipatednegative expectancy violation. Finally, impressionoffsetting appears to be an effective AIM tactic be-cause it positively influences the market reaction toacquisition announcements.

DISCUSSION

We introduced a new form of AIM—impression off-setting: an AIM tactic organizational leaders em-ploy when they expect a focal event will negatively

violate the expectations of external stakeholders.Our results suggest when firm executives antici-pate that observers may view an upcoming an-nouncement as a deviation from expected behavior(i.e., a negative expectancy violation), they are proneto employing impression offsetting in which theycontemporaneously issuepositive, butunrelated, pressreleases with an acquisition announcement.

Ours is the first study to quantify the benefits of anyimpressionmanagement technique.Wefoundevidencethat impression offsetting limits negative abnormalmarket returns to acquisition announcements. Specifi-cally, we found that impression offsetting inhibitedperceptionsof anegativeexpectancyviolationsuch thatwhen firms issued one positive release, the negativemarket reaction associated with acquisition announce-ments decreased by roughly 44%.

In addition, we theorized and found that organiza-tional leaders are more or less likely to engage in

TABLE 4Effects of Impression Offsetting on Cumulative Abnormal Returns

Model 1 Model 2

Constant 0.0113 (0.0490) 0.0101 (0.0494)Governance ControlsCEO pay 0.0008 (0.0031) 0.0008 (0.0030)CEO tenure 20.0004 (0.0003) 20.0004 (0.0003)Outside director percentage 0.0181 (0.0347) 0.0226 (0.0348)Total directors 0.00141 (0.0008) 0.0016* (0.0008)Acquirer ControlsLagged ROA 0.06031 (0.0337) 0.06361 (0.0339)Free cash flow 0.0033 (0.0075) 0.0042 (0.0074)Stock returns 0.0078** (0.0029) 0.0078** (0.0029)Acquirer size 20.0010 (0.0022) 20.0019 (0.0023)Acquirer industry concentration 0.0082 (0.0241) 0.0105 (0.0242)Demand uncertainty 20.00001 (0.0000) 20.00001 (0.0000)Baseline positive announcements 20.0067 (0.0174) 20.0168 (0.0171)Prior acquisition activity 0.0025* (0.0011) 0.0025* (0.0012)High reputation 0.01371 (0.0072) 0.01311 (0.0073)Acquisition ControlsPercent stock 20.0002** (0.0001) 20.0002** (0.0001)Attitude 0.0019 (0.0097) 0.0035 (0.0098)Unrelatedness 20.0052 (0.0048) 20.0048 (0.0047)Unprofitable target 0.0057 (0.0050) 0.0062 (0.0049)Announcement ControlsFriday announcement 0.0006 (0.0067) 0.0015 (0.0064)Negative announcements 20.0289* (0.0119)Neutral announcements 20.0009 (0.0060)Independent VariableImpression offsetting 0.0029* (0.0013)N 701 701r2 0.1076 0.1194

Note: Year dummy variables are present in the analyses but omitted from the tables.1 p , 0.1* p , 0.05

** p , 0.01 (standard errors in parentheses)


impression offsetting depending upon the anticipatedseverity of an expectancy violation. These findingsspeak directly to the theoretical mechanisms under-lying the use of this previously unexplored impressionmanagement tactic. Since it is the anticipation of a po-tential expectancy violation that drives the use of thistactic, it logically follows that the more severe the vi-olation, the more likely organizational leaders willengage in impression offsetting. In our context, wefound that the higher the riskiness of the acquisi-tion, the more likely organizational leaders were toengage in impression offsetting. Specificallywe foundthat serial acquirers make greater use of impressionoffsetting because markets often respond negativelyto acquisition programs (e.g., Laamanen &Keil, 2008).Hence, employing impression offsetting to a currentacquisition may positively influence shareholders’perceptions of the firm’s ability to execute subsequent

acquisitions, which may improve the reaction to fu-ture acquisitions. Moreover, we found that high-reputation organizations that derive value from theirreputation as an intangible asset and likely perceivegreater risk to their social evaluations, act to protect itusing impression offsetting. We also found a lack ofprofitability led to a greater use of impression off-setting, which may suggest a relative lack of valuableresources or amore difficult and/or riskier integrationprocess contributed to managers’ perceptions ofa higher risk of negative expectancy violation.

Implications for ImpressionManagement Research

Our study contributes to the literature on AIM inanumberofways.First,weorganizedandextended thelimited and disparate AIM literature. Only two AIMtactics have been explored previously in the

TABLE 5Effects of Social Evaluation Risk Factors on Impression Offsetting Count

Model 1 Model 2

Constant 26.4440** (1.1176) 25.7956** (1.1343)Governance ControlsCEO pay 0.0824 (0.0840) 0.0868 (0.0838)CEO tenure 0.0154 (0.0106) 0.0115 (0.0109)Outside director percentage 20.1077 (0.9198) 0.2215 (0.9792)Total directors 20.0690* (0.0291) 20.0764** (0.0292)Acquirer ControlsLagged ROA 20.7940 (1.2429) 21.2690 (1.0627)Free cash flow 0.3412 (0.2773) 0.2787 (0.2818)Stock returns 0.0275 (0.0863) 20.0149 (0.0764)Acquirer size 0.5647** (0.0827) 0.4674** (0.0820)Acquirer industry concentration 20.1024 (0.5155) 0.2303 (0.4730)Demand uncertainty 0.0000 (0.0000) 0.0000 (0.0000)Baseline positive announcements 3.1194** (0.4949) 2.7464** (0.4624)Acquisition ControlsPercent stock 20.0024 (0.0019) 20.0023 (0.0019)Attitude 20.4114 (0.3205) 20.57311 (0.3166)Announcement ControlsFriday announcement 20.1582 (0.2149) 20.1264 (0.2120)Negative announcements 0.2909 (0.2288) 0.1912 (0.2496)Neutral announcements 0.1702 (0.1784) 0.0623 (0.1682)Independent VariablesPrior acquisition activity 0.1614** (0.0388)High reputation 0.6466** (0.2395)Unrelatedness 0.0860 (0.1543)Unprofitable target 0.2836* (0.1404)lnalphaConstant 0.1348 (0.1817) 20.0871 (0.2171)N 758 758?2 211.2766** 249.3262**

Note: Year dummy variables are present in the analyses but omitted from the tables.1 p , 0.1* p , 0.05

** p , 0.01 (standard errors in parentheses)


accounting and strategic management literatures: bigbath accounting and strategic noise. Strategic noiseinvolves organizational leaders releasing of bothpositive and negative unrelated information whenthe anticipated response is unclear. Big bath ac-counting occurs when organizational leaders releaseadditional negative information when the expectedreaction is negative. Impression offsetting representsa previously unexplored AIM tactic employed byorganizational leaders. Second, the results of thecurrent study, when combined with other recent re-search onAIM (e.g., Elsbach et al., 1998; Graffin et al.,2011), suggest that firms engage in a range of im-pression management activities in anticipation ofa potential expectancy violation.

In the case of impressionoffsetting, future researchcould explore prescriptive and predictive charac-teristics that trigger this AIM technique. Since man-agers are often aware of investors’ negative reactionsto acquisitions and likely wish to avoid market re-pudiations of their actions, future researchers maywish to examine how managerial characteristics,such as a CEO’s reputation or tenure, influence thelikelihood they engage in impression offsetting. Inaddition, although we only focused on negative ex-pectancy violations, positive expectancy violationsmay also occur when firm behavior exceeds expec-tations. We thus encourage future researchers toexplore conditions under which firms exceed ex-pectations, and the resulting reactions by marketsand stakeholders to this new information.

Our results also have implications for the broaderimpression management literature in two respects.First, although we discussed anticipatory and re-active impression management as distinct and sep-arate forms of influencingmarkets, itmay be the casethat impression management has both reactive andanticipatory elements. For instance, organizationalleaders may delay announcements for those eventsfor which they control the timing of the announce-ment if anunanticipatedexpectancyviolationoccurs.Moreover, firmsmightdelayannouncingacquisitionsif an unanticipated expectancy violation occurs, suchas a scandal or environmental accident. Hence, theanticipated timing of the announcement of an eventthat is under the control of organizations could actu-ally be a form of impression management, even if theevent itself were unanticipated. Along these lines,future researchers may wish to explore the joint re-lationships between anticipatory and reactive im-pression management in more detail.

Second, our supplementary analyses suggest thatfirms are not only more likely to release positive

unrelated information around acquisition announce-ments, but also that the release of negative informa-tion decreases in a statistically significant manner.This finding suggests there is agency behind thetiming of the release of a firm’s significant negativeoccurrences. How and when firms elect to releasesignificant bad news has received little direct atten-tion. Therefore, a better understanding of when firmsare more or less likely to release negative informationwould seem to represent fertile ground for future im-pression management research.

Implications for Acquisitions Research

Our findings also have implications for researchexamining acquisitions. To the extent that priorstudies of acquisition announcements did not con-trol for confounding events, the results of thesestudies may have been positively biased in light oforganizations employing impression offsetting. Thissuggests the stock market reaction for a significantpercentage of acquisition announcements—roughly27% in our sample—in prior studies may have beenpositively biased. In other words, despite the fact theprior research findsagenerallynegativemarket reactionto acquisitions, the size of the observed negative re-actionmay beunderstated. Our results also suggest thisbias occurs in systematicways aswe foundmore activeand higher reputation acquirers were more apt to en-gage in impression offsetting. This finding, combinedwith our results on the effectiveness of impression off-setting, suggests prior studiesmayhave reported biasedabnormal returns particularly for these types of acqui-sitions. Future research may wish to re-examine find-ings relating to the market reaction for active acquirersand high-reputation firms. Indeed, it may be the casethat,due to theirhigher levelsof acquisitionexperience,serial acquirers may have developed an impressionoffsetting capability,whichupwardly biases themarketreaction of their acquisitions.

Practical Implications

Our results also have practical implications formanagers. We found that impression offsetting suc-cessfully reduced the negative market reaction toa potential negative expectancy violation in oursample. Hence, our results may encourage organi-zational leaders to employ this tactic. For investorsand other market participants, however, impressionoffsetting—and its effectiveness—suggests that theyshould be aware of how an organization is managingthe event, and whether this needs to be taken intoconsiderationwhenvaluing the organization’s actions.


This exploration may be particularly important whenorganizations take on projects with high likelihoods ofbeing perceived as negative expectancy violations.Thus, investors should interpret press releases aroundpotentially negative events with caution.

Our research also has implications formanagers interms of the timing of announcements, which areunder their control. Indeed, our results combinedwith other studies of AIM (e.g., Elliott & Shaw, 1988;Graffin et al., 2011) suggest that organizational leaderswill actively attempt to influence stakeholder reac-tion to significant organizational events when theseleaders have advanced knowledge of the event be-coming known to stakeholders. While the currentstudy focuses on how organizational leaders proac-tivelymanage impressions associatedwith a potentialnegative expectancy violation, managers may wish toengage in this sort of impressionmanagement in othercontexts. Indeed, organizational leaders cananticipatenumerous occurrences, such as earnings announce-ments, alliance announcements and new productannouncements, and may involve positive and/ornegative potential expectancy violations.

Limitations and Future Research

One limitation of our study is that we were not abletodirectlyobserveorganizational leaders intentionallyreleasing information to offset the market’s negativereaction to acquisition announcements. We ratherinferred this motivation based upon an observed em-pirical pattern relative to a baseline behavior. Specifi-cally, we calculated a baseline count of confoundingpress releases around any announcement and ob-served a significant higher count of offsetting an-nouncements around acquisitions. Based on thedeviation from the baseline, we inferred agency on thepart of acquiring firms. Our approach, however, isconsistent with recent work in finance that inferredmanagerial malfeasance when CEOs repeatedly expe-rienced “lucky stock grants,” defined as receiving anoption at the lower stock price that month (Bebchuk,Grinstein, & Peter, 2010). Future researchersmaywishto leverage this tactic to better understand potentialmotives for other patterns of behavior within firmssuch as executive or board members’ actions.

Given this AIM tactic is both inexpensive and ef-fective, it is not clear why more firms do not use it. Inaddition, though, we explored the frequency to whichorganizational leaders engage in impression manage-ment in other contexts. Porac et al. (1999) found thatorganizations selected31%of itspeergroup fromfirmsoutside of its industry, which they took to represent

a form of impression management. Similarly, Graffinet al. (2011) found that 20% of firms engaged in im-pression management surrounding CEO announce-ments. While our results seem to be consistent withother studies that have measured the percentage offirms that do or do not engage in impression manage-ment, it is unclear why so many firms do not engagemore actively in impression management. Might thehigh and increasing rate of turnover in the executivesuite be limiting the development of such impressionmanagement capabilities? Could the strengthening ofgovernance and regulations regarding the release ofinformation be limiting executives’ discretion to en-gage in such tactics? Future research may wish to ex-plore thedifferencesbetween thesegroupsof firmsandthe rate of diffusion for such tactics.

Conclusion

We introduced a new AIM tactic—impression off-setting. Our results suggest when firm executives an-ticipate that observers may view an upcomingannouncement as a negative expectancy violation,they contemporaneously issue positive, but unrelated,press releases with an acquisition announcement. Wealso found that impression offsetting is effective, as itsignificantly decreases the negative abnormal returnaround an acquisition announcement. Finally, wefound that riskier acquisitions were more associatedwith this AIM technique. Stated simply, our findingsindicate that firms can successfully influence marketreactions to organizational announcements. To buildon our findings, we encourage future work on AIM inorder to better understand how organizational leadersattempt to influence market participants.

REFERENCES

Ashforth, B. E., & Gibbs, B. W. 1990. The double-edge oforganizational legitimation.Organization Science, 1:177–194.

Bachman, G., & Guerrero, L. 2006. An expectancy viola-tions analysis of relational quality and communica-tive responses following hurtful events in datingrelationships. Journal of Social and Personal Re-lationships, 23: 943–963.

Barney, J. 1988. Returns to bidding firms in mergers andacquisitions: Reconsidering the relatedness hypothe-sis. Strategic Management Journal, 9: 71–78

Barnett, M. L., & Pollock, T. G. (Eds.) 2012. The Oxfordhandbook of corporate reputation. Oxford, U.K.:Oxford University Press.

Bebchuk, L. A., Grinstein, Y., & Peter, U. 2010. LuckyCEOsandluckdirectors.TheJournalofFinance, 6: 2363–2401.


Bettman, J. R., & Weitz, B. A. 1983. Attributions in theboard room: Causal reasoning in corporate annual re-ports.Administrative Science Quarterly, 28: 165–183.

Bolino, K., Kacmar, M., Turnley, W. H., & Gilstrap, J. B.2008.Amulti-level reviewof impressionmanagementmotives and behaviors. Journal of Management, 34:1080–1109.

Browne, L. E., & Rosengren, E. S. 1987. Are hostile take-overs different? In L. E. Browne & E. S. Rosengren(Eds.), The merger boom: 199–229. Boston: FederalReserve Bank.

Burgoon, J. K. 1993. Interpersonal expectations, expectancyviolations, and emotional communication. Journal ofLanguage and Social Psychology, 12: 30–48.

Burgoon, J. K., & Hale, J. L. 1988. Nonverbal expectancy vio-lations:modelelaborationandapplication to immediacybehaviors. Communication Monographs, 55: 58–79.

Burgoon, J. K., Stern, L. A., & Dillman, L. 1995. Interper-sonal adaptation: Dyadic interaction patterns. NewYork: Cambridge University.

Capron,L.,&Shen, J.-C.2007.Acquisitionsofprivatevs.publicfirms: Private information, target selection, and acquirerreturns. Strategic Management Journal, 28: 891–911.

Christensen,C.M.,Alton,R.,Rising,C.,&Waldeck,A.2011.Thebig idea: The new M&A playbook. Available from http://hbr.org/2011/03/the-big-idea-the-new-ma-playbook/ar/1“ \t ”_blank.Harvard Business Review, 89: 3.

Cohen, J. 1968. Weighed kappa: Nominal scale agreementwith provision for scaled disagreement or partialcredit. Psychological Bulletin, 70: 213–220.

Cohen, W. M., & Levinthal, D. A. 1990. Absorptive capac-ity: A new perspective on learning and innovation.Administrative Science Quarterly, 35: 128–152.

Dellavigna, S., & Pollet, J.M. 2009. Investor inattention andFriday earnings announcements. The Journal of Fi-nance, 64: 709–749.

Dess, G. G., & Beard, D. W. 1984. Dimensions of organiza-tional task environments. Administrative ScienceQuarterly, 29: 52–73.

Eisenmann, T. R. 2002. The effects of CEO equity owner-ship and firm diversification on risk taking. StrategicManagement Journal, 23: 513–534.

Elliott, J. A., & Shaw, W. H. 1988. Write-offs as accountingprocedures to manage perceptions. Journal of Ac-counting Research, 26: 91–119.

Elsbach, K. D. 2006. Organizational perception man-agement. Mahwah, NJ: Lawrence Erlbaum.

Elsbach, K. D. 2012. A framework for reputation manage-ment over the course of evolving controversies. InM. L. Barnett & T. G. Pollock (Eds.), The Oxfordhandbook of corporate reputation: 466-486. Oxford,U.K.: Oxford University Press.

Elsbach,K.D., Sutton, R. I., &Principe, K. E. 1998.Avertingexpected challenges through anticipatory impressionmanagement: A study of hospital billing. Organiza-tion Science, 9: 68–86.

Fehr, E., & Gachter, S. 2000. Fairness and retaliation: Theeconomics of reciprocity. The Journal of EconomicPerspectives, 14: 159–181.

Fincham, F., & Bradbury, T. 1987. The assessment ofmarital quality: A reevaluation. Journal of Marriageand the Family, 49: 797–809.

Fleiss, J. L. 1981. Statistical methods for rates and pro-portions (2nd ed.). New York: John Wiley.

Fligstein, N. 1990. The transformation of corporate con-trol. Cambridge, MA: Harvard Press.

Floyd, K., Ramirez, A., Jr., & Burgoon, J. K. 1999. Expectancyviolations theory. InL.A.Guerrero,M.Hecht& J.DeVito(Eds.), The nonverbal communication reader (2nded.): 437–444. Prospect Heights, IL: Waveland Press.

Fombrun, C. J. 1996. Reputation: Realizing value from thecorporate image.Boston:HarvardBusinessSchoolPress.

Friedman,M. 1970. The social responsibility of business isto increase its profits. The New York Times Maga-zine, September 13.

Graffin, S.D., Carpenter,M.A.,&Boivie, S. 2011.What’s allthat (strategic) noise? Anticipatory impression man-agement in CEO succession. Strategic ManagementJournal, 32: 748–770.

Haleblian, J., Devers, C. E., McNamara, G., Carpenter,M. A.,& Davison, R. B. 2009. Taking stock of what we knowaboutmergers and acquisitions: A review and researchagenda. Journal of Management, 35: 469–502.

Haleblian, J., & Finkelstein, S. 1999. The influence of or-ganizational acquisition experience on acquisitionperformance: A behavioral learning perspective. Ad-ministrative Science Quarterly, 44: 29–56.

Haleblian, J., Kim, J., & Rajagopalan, N. 2006. The influence ofacquisition experience and performance on acquisitionbehavior: Evidence from the U.S. commercial banking in-dustry.Academy ofManagement Journal, 49: 357–370.

Harford, J. 2005. What drives merger waves? Journal ofFinancial Economics, 77: 529–560.

Haspeslagh, P. C., & Jemison, D. B. 1991. Managing ac-quisitions. New York: Free Press.

Hayward, M. L. A. 2002. When do firms learn from theiracquisition experience? Evidence from 1990–1995.Strategic Management Journal, 23: 21–39.

Hoskisson, R. E., & Hitt, M. A. 1990. Antecedents andperformance outcomes of diversification: A reviewand critique of theoretical perspectives. Journal ofManagement, 16: 461–509.

Hou, K., & Robinson, D. 2006. Industry concentration and av-erage stock returns. Journal of Finance. 61: 1927–1956.


http://hbr.org/2011/03/the-big-idea-the-new-ma-playbook/ar/1&tnqh_x201C; \t &tnqh_x201D;_blank



Hovland, C., Janis, I., & Kelley, H. 1953. Communicationand persuasion. New Haven, CT: Yale.

Kennedy, M. 2008. Getting counted: Markets, media, andreality.American Sociological Review, 73: 270–295.

Kim, S. 2014. The role of prior expectancies and relationalsatisfaction in crisis. Journalism & Mass Communi-cation Quarterly, 91: 139–158.

Kirschenheiter, M., & Melumad, N. D. 2002. Can “bigbath” and earnings smoothing co-exist as equilib-rium financial reporting strategies? Journal of Ac-counting Research, 40: 761–796.

Laamanen, T., & Keil, T. 2008. Performance of serialacquirers: Toward an acquisition program perspec-tive. Strategic Management Journal, 29: 663–672.

Landis, J. R., &Koch,G.G. 1977. Themeasurement of observeragreement for categorical data. Biometrics, 33: 159–174.

Lang, L., Stulz, R.M., &Walkling, R. 1991. A test of the freecash flow hypothesis. Journal of Financial Econom-ics, 29: 315–335.

Lange,D.,Lee,P.M.,&Dai,Y.2011.Organizational reputation:A review. Journal of Management, 37: 153–184.

McWilliams, A., & Siegel, D. 1997. Event studies in man-agement research: Theoretical and empirical issues.Academy of Management Journal, 40: 626–657.

Milbourn, T. T. 2003. CEO reputation and stock-based com-pensation. Journal of Financial Economics, 68: 233–262.

Moeller, S. B., Schlingemann, F. P., & Stulz, R. M. 2004.Firm size and the gains from acquisitions. Journal ofFinancial Economics, 73: 201–228.

Morck, R., Shleifer, A., & Vishny, R. W. 1990. Do manage-rial objectives drive bad acquisitions? Journal of Fi-nance, 45: 31–48.

Mueller, D. C. 1987. The corporation: Growth, diversifica-tion and mergers. New York: Harwood Academic.

Murray, M. P. 2006. Avoiding invalid instruments andcoping with weak instruments. The Journal of Eco-nomic Perspectives, 20: 111–132.

Pablo, A. 1994. Determinants of acquisition integrationlevel.Academy of Management Journal, 37: 803–836.

Petkova, A., Wadhwa, A., Yao, X., & Jain, S. 2014. Repu-tation and decision making under ambiguity: A studyof U.S. venture capital firms’ investments in theemerging clean energy sector. Academy of Manage-ment Journal, 57: 422–448.

Pfarrer, M. D., DeCelles, K. A., Smith, K. G., & Taylor, M. S.2008. After the fall: Reintegrating the corrupt organiza-tion.Academy of Management Review, 33: 730–749.

Pfarrer, M. D., Pollock, T. G., & Rindova, V. P. 2010. A taleof two assets: The effects of firm reputation and ce-lebrity on earnings surprises and investors’ reactions.Academy of Management Journal, 53: 1131–1152.

Porac, J. F., Wade, J. B., & Pollock, T. G. 1999. Industrycategories and the politics of the comparable firm inCEO compensation. Administrative Science Quar-terly, 44: 112–144.

Puffer, S. M., & Weintrop, J. B. 1991. Corporate perfor-mance and CEO turnover: The role of performanceexpectations. Administrative Science Quarterly, 36:1–19.

Rhee, M., & Haunschild, P. R. 2006. The liability of goodreputation: A study of product recalls in the U.S. auto-mobile industry.Organization Science, 17: 101–117.

Sanders,W.G. 2001.Behavioral responses ofCEOs to stockownership and stock option pay. Academy of Man-agement Journal, 44: 477–492.

Saxton, T., & Dollinger, M. 2004. Target reputation andappropriability: Picking and deploying resouces inacquisitions. Journal of Management, 30: 123–147.

Schipper, K., & Thompson, R. 1983. The impact of merger-related regulations on the shareholders of acquiringfirms. Journal of Accounting Research, 21: 184–221.

Schneider, D. J. 1981. Tactical self-presentation: Towarda broader conception. In J. T. Tedeschi (Ed.), Impres-sion management theory and social psychologicalresearch: 23–40. New York: Academic Press.

Semadeni, M., Withers, M. C., & Certo, S. T. 2014. Theperils of endogeneity and instrumental variables instrategy research: Understanding through simulations.Strategic Management Journal, 35: 1070–1079.

Seth, A. 1990. Value creation in acquisitions: A reexami-nation of performance issues.StrategicManagementJournal, 11: 99–115.

Simpson, S. S. 2002.Organization crime, law, and socialcontrol. Cambridge: Cambridge University Press.

Staw, B. M., McKechnie, P. I., & Puffer, S. M. 1983. Thejustification of organizational performance. Admin-istrative Science Quarterly, 28: 582–600.

Strong, J. S., & Meyer, J. R. 1987. Asset writedowns: Man-agerial incentives and security returns.The Journal ofFinance, 42: 643–661.

Tedeschi, J. T., & Reiss,M. 1981. Identities, the phenomenalself, and laboratory research. In J. T. Tedeschi (Ed.),Impression management theory and social psycho-logical research: 3–22. New York: Academic Press.

Tosi, H., Aldag, R., & Story, R. G. 1973. On the measure-ment of the environment: An assessment of the Law-rence and Lorsch environmental uncertainty subscale.Administrative Science Quarterly, 18: 27–36.

Travlos, N. 1987. Corporate takeover bids, methods ofpayment, and bidding firms’ stock returns. The Jour-nal of Finance, 42: 943–962.

Vermeulen, F., & Barkema, H. 2001. Learning through acqui-sitions.Academy ofManagement Journal, 44: 457–476.


Vermeulen, F., & Barkema, H. 2002. Pace, rhythm, and scope:Processdependence in building aprofitablemultinationalcorporation.StrategicManagement Journal, 23: 637–653.

Wade, J. B., Porac, J. F., Pollock, T. G., & Graffin, S. D. 2006.The burden of celebrity: The impact of CEO certifica-tion contests on CEO pay and performance.Academyof Management Journal, 49: 643–660.

Walsh, P., Craig, R., & Clarke, F. 1991. “Big bath account-ing” using extraordinary items adjustments: Austra-lian empirical evidence. Journal of Business Finance& Accounting, 18: 173–189.

Westphal, J. D., & Deephouse, D. L. 2011. Avoiding badpress: Interpersonal influence in relations betweenCEOs and journalists and the consequences for pressreporting about firms and their leadership. Organi-zation Science, 22: 1061–1086.

Zajac, E. J., & Westphal, J. D. 1995. Accounting for the expla-nations ofCEOcompensation: Substance and symbolism.Administrative Science Quarterly, 40: 283–308.

Zavyalova,A., Pfarrer,M.D., Reger, R. K., &Shapiro,D. L. 2012.Managing the message: The effects of firm actions andindustry spillovers on media coverage following wrong-doing.AcademyofManagement Journal, 55: 1079–1101.

Scott D. Graffin ([email protected]) is an Associate Pro-fessor at the University of Georgia’s Terry College ofBusiness. He received his PhD in Strategic Managementfrom the University of Wisconsin–Madison. His re-search interests include corporate governance, as wellas the impact of reputation, status, and organizationalimpression management activities on organizationoutcomes.

Jerayr Haleblian ([email protected]) is an AssociateProfessor at the University of California-Riverside. He re-ceived his PhD from theUniversity of Southern California.His research focuses on strategic decisions and their per-formance outcomes in the contexts of mergers and acqui-sitions and strategic leadership.

JasonT. Kiley ([email protected]) is anAssistant Professorat the Spears School of Business at Oklahoma State Uni-versity. He received his PhD in Business Administrationwith a focus on Strategic Management from the Univer-sity of Georgia. His research interests include organiza-tional impression management, reputation, and socialperceptions, particularly in the context of mergers andacquisitions.

APPENDIX A APPENDIX B

TABLE A1Initial Categorization of Press Releases

Category Positive Neutral Negative

Earnings releases 34 8 9Earnings guidance 16 16 2Change in dividend rate 4New product 197Customer win 79Social good (e.g., donation,

sponsorship)41

Received award from thirdparty

27

Buyback or stock split 10Results of a sponsored study 7New executive or director 11Divestiture or plant closing 9Settlement of litigation or

other legal dispute2

Executive retirement 1Change of stock exchange

listing1

Debt issuance 1Other acquisition 34Completion of another

acquisition10

Recall or safety issue 1Totals 415 49 56

TABLE B1Organization Frequency by Total Acquisition Count

Acquisitions Frequency (Orgs.)

1 1762 743 294 205 126 77 58 49 2

10 111 212 114 115 119 1


mailto:[email protected]



Copyright of Academy of Management Journal is the property of Academy of Managementand its content may not be copied or emailed to multiple sites or posted to a listserv withoutthe copyright holder's express written permission. However, users may print, download, oremail articles for individual use.

Documents

READY, AIM, ACQUIRE: IMPRESSION OFFSETTING AND ......READY, AIM, ACQUIRE: IMPRESSION OFFSETTING AND ACQUISITIONS SCOTT D. GRAFFIN University of Georgia JERAYR (JOHN) HALEBLIAN University