third party employment branding: human capital …...folio of methods to self- promote and sustain...

r Academy of Management Journal2016, Vol. 59, No. 1, 90–112.http://dx.doi.org/10.5465/amj.2013.1091

THIRD PARTY EMPLOYMENT BRANDING:HUMAN CAPITAL INFLOWS AND OUTFLOWS FOLLOWING

“BEST PLACES TO WORK” CERTIFICATIONS

BRIAN R. DINEENPurdue University

DAVID G. ALLENRutgers University & Warwick Business School

“Best Places to Work” (BPTW) and similar competitions are a proliferating form of thirdparty employment branding. Little is known, however, about how single or repeatedthird party employment branding occurrences relate to key human capital outcomes.Extending signaling theory by considering signal credibility and comparability, we usearchival and survey data from 624 BPTW participants in 16 competitions across a three-year period to develop and test hypotheses linking BPTW certifications to collectiveturnover rates and key informant perceptions of applicant pool quality. We find thatcertifications are associated with lower turnover rates, and in addition, propose com-peting crystallization and celebrity hypotheses that model turnover trajectories withrepeated certifications, finding diminishing marginal turnover reductions across mul-tiple certifications. We also examine company size and industry job opening modera-tors, finding that as certifications increase, applicant pool quality is: (1) higher insmaller companies, and (2) higher when job openings are scarcer. Finally, beyond beingcertified or not, we find supplemental evidence for effects of the specific certificationlevel achieved (e.g., 2nd versus 15th). This investigation advances theory related tocollective turnover, applicant pool quality, and employment branding, and is relevant tocompany decisions about seeking or re-seeking third party certifications.

Individual and organizational labor market partici-pants routinely face two important behavior-shapingconditions: competition, whereby organizationscompete to attract and retain individuals and in-dividuals compete to obtain andmaintain jobs; anduncertain decision making, whereby organizationsand individuals select each other based on in-complete information. Signaling theory (Connelly,Certo, Ireland, & Reutzel, 2011; Spence, 1973) de-scribes how these two forces interact to create sig-naling systems evolving over time in response toorganizational and individual behavior; how and

why signals about unobservable characteristics aredeveloped, sent, received, and interpreted; andhow individual micro-oriented processes generatemore macro-oriented trends and outcomes. Webuild on signaling theory to examine the effects ofthird party generated signals regarding organiza-tions’ employment characteristics.

From a macro perspective, organizations are in-creasingly recognizing that they must strategicallymanage their human capital (e.g., Becker & Huselid,2006; Rousseau, 1995; Ulrich, Younger, Brockbank,& Ulrich, 2012). Strategic human resource manage-ment theory suggests that clusters or systems ofhigh-performancework practices increase employeeabilities, motivation, and opportunities to perform,resulting in outcomes such as higher productivity,better customer service, positive attitudes, improveddecision making, reduced turnover, and, ultimately,improved organizational performance (e.g., Combs,Liu, Hall, & Ketchen, 2006; Huselid, 1995; Jiang,Lepak, Hu, & Baer, 2012). Research in this areahas largely focused on employee perceptions ofand reactions to Human Resource (HR) systems

This research was funded by a grant from the SHRMFoundation. We thank Peter Burke, President of Best Com-panies Group, and his excellent staff, Abbie Shipp for herassistance with the data analysis process, and Chuck Pierce,Jason Shaw, and the faculties of the Krannert School ofManagementOrganizationalBehavior andHumanResourcesarea at Purdue University and the London Business SchoolOrganizational Behavior area for their helpful comments andsuggestions. Finally,we thankLyleHanna forhis support andJacquelyn Thompson for her editorial assistance.

90

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.2013.1091

(e.g., Bowen & Ostroff, 2004; Piening, Baluch, &Salge, 2013). However, scholars have argued that“strong” HR systems in which practices exhibit dis-tinctiveness, consistency, and consensus, while dif-ficult for firms to create, can be evenmore difficult tochange once in place (Bowen & Ostroff, 2004), andcan takemultiple years to affect distal organizationaloutcomes (Huselid & Becker, 1996). Thus, beyondthe influence of HR systems, or changes to thosesystems over a multi-year period, others have sug-gested that firm reputation is a critical means of un-derstanding sustained competitive advantage via theinfluence of high-performance HR practices on or-ganizational performance (Barney, 2001; Zorn,Roper, & Richardson, 2014). Germane to reputationare employers’ efforts to brand themselves as moreengaging, developmental, and employee-centered,and increased practical and scholarly attention hasturned to employment branding efforts as a keyconsideration for increasing the knowledge peoplehave about employers and for attracting and retain-ing talent (e.g., Cable & Turban, 2001; Collins &Kanar, 2014; Gardner, Erhardt, & Martin-Rios, 2011;Jones, Willness, & Madey, 2014).

The idea of employment branding draws frommarketing theory on brand image to suggest that or-ganizations can be associated with an employmentbrand consisting of individuals’ perceptions aboutwhat is distinctive, central, and enduring about theorganization as a place to work. Gardner et al. (2011:261) defined an employment brand as “names,terms, signs, symbols, or designs or a combination ofthem intended to identify the employment offeringof one employer and to differentiate it from the of-ferings of competing employers,” and organizationsput forth effort to create or alter these brand percep-tions to convey a favorable value proposition to po-tential or current employees (e.g., Edwards, 2010;Lievens, 2007; Van Hoye, Bas, Cromheecke, &Lievens, 2012). In doing so, an employment brandmay translate into brand equity by influencing de-cisions to apply for, accept, and remain with an or-ganization’s job opportunities (Cable & Turban,2001; Collins & Kanar, 2014).

Consistent with these ideas, organizations mustquestionhow theymaybest signal in the competitivemarketplace that they are particularly attractiveplaces in which to commence or continue employ-ment relationships. Organizations draw from a port-folio of methods to self- promote and sustain theiremployment brand. For example, they might em-phasize corporate culture initiatives or changes(e.g., Katzenbach, Steffen, & Kronley, 2012), such as

work–life flexibility; or leverage consumer brandingefforts by using social media to promote corporatesocial responsibility policies attractive to potentialconsumers and employees. However, rather thanfocusing on organizations’ self-branding efforts, wepropose that signaling theory provides a broadfoundation and theory-based account for how andwhy organizations use third party information toparlay strategic human capital investments into or-ganizational performance outcomes by signalingtheir employment brand.

We examine in this paper the signaling effects ofthird party employment branding on human capitalinflows to and outflows from organizations, definingit as organizations’ use of communications, claims,or status-based classifications generated by inde-pendent external parties to shape, enhance, anddifferentiate organizations’ images as favorable em-ployers. We specifically examine formal third partybranding in the form of certifications or other status-based awards or recognition, rather than informalword-of-mouth information (e.g., VanHoye & Lievens,2009).Wefocuson“BestPlaces toWork” (BPTW)asanevaluative certification bestowed on companies thatdemonstrate superlative employee relations practices.BPTW and similar programs have grown from fewerthan 10 nationwide in the late 1990s to more than 100in recent years (Burke, 2014). This reflects the grow-ing number of third party sources making cross-organizational comparisons for consumers (e.g., yelp.com) and job seekers (e.g., glassdoor.com) more at-tainable than ever. Like those we examine, most em-ployer certifications are conferred as a function ofemployee engagement and company HR practiceassessments.

BPTW certifications have been shown to posi-tively affect employee attitudes, applicant poolquality, and financial performance, but smallersamples have precluded adequate partialing oflagged outcomes (e.g., Collins &Han, 2004; Fulmer,Gerhart, & Scott, 2003), performance has beencompared across broadly defined peer groups(e.g., Standard & Poor’s; Lau & May, 1998), and littleresearchhas examined effects on company-providedhuman capital inflow and outflow data over a pro-longed period. We investigate BPTW certificationeffects on firm level outcomes, including turnoverrates and key informant perceptions of applicant poolquality, over a three-year period while controlling forlagged outcomes and a comprehensive set of HRpractices that might otherwise affect these outcomes.

We intend to contribute to signaling theory and theemployment branding literature in the following

2016 91Dineen and Allen

http://yelp.com

http://yelp.com

http://glassdoor.com

ways. First, signaling theory focuses on systemscomprising senders, receivers, and signals (Bangerter,Roulin, & Konig, 2012; Connelly et al., 2011).We build on this by recognizing that signalingsystems sometimes incorporate third parties thattransmit signals on the sender’s behalf. Second,signaling theory and employment branding researchhave tended to focus on attracting talent. For ex-ample, Chapman, Uggerslev, Carroll, Piasentin, andJones (2005) summarized the effects of 27 studieslinking organization image to applicant attraction.Much recruitment research has used the perspectivethat applicants face uncertain and incomplete in-formation about jobs and work environments toargue that organizations invest in signaling theirpositive qualities as they compete with other orga-nizations to attract applicants (Bangerter et al., 2012;Connelly et al., 2011; Jones et al., 2014; Rynes, 1991;Spence, 1973). We shift this focus to consider thatcurrent employees may also be signaling targetsby examining turnover rates, an approach thatpotentially advances the burgeoning literature oncollective turnover, or “aggregate levels of employeedepartures” (Hausknecht & Trevor, 2011: 353; seealso, Heavey, Holwerda, &Hausknecht, 2013; Shaw,2011). For example, while Hausknecht and Trevor’s(2011) model of collective turnover identifies vari-ousHRpractices, collective employee attitudes, andcollective business and environmental factors thatmay influence turnover rates, it does not addressemployer image or reputation. Heavey et al.’s (2013)meta-analysis of collective turnover antecedentscomprises six broad predictor categories, none ofwhich include employment branding concerns oreffect sizes.

Third, we consider the effects of signaling whilecontrolling for a host of HR practices that mightotherwise affect turnover and applicant pool qualityoutcomes. This extends work that has found in-cremental effects of symbolic characteristics, whichin our studypertain to thirdparty certifications of theemployment brand, beyond instrumental character-istics such as tangible benefits, training opportuni-ties, or work–life policies (e.g., Lievens, 2007;Lievens & Highhouse, 2003). Fourth, we considerthat signal interpretation may depend on certaincontextual characteristics. Thus, we consider con-ceptual extensions to signaling theory as well as twopotential moderators: company size in a sample ofmostly small to medium companies, and job open-ings across a range of industries. These extensionsbuild on recent work identifying how signals attractjob seekers (Jones et al., 2014). Fifth, signaling theory

explicitly describes signaling systems as dynamic(Connelly et al., 2011; Spence, 1973). We thereforeconsider the effects of repeated BPTW signals onturnover, investigating points at which effects mightcommence, and the sustainability of signals oncethey are sent (Rindova, Pollock, & Hayward, 2006).For example, we ask whether a single third partyemployment branding event relates to turnover, orwhether only repeated employment branding sig-nals relate to turnover.

Finally, we intend to make an important practi-cal contribution regarding the true value of thirdparty employment branding. Although companiesare attracted to BPTW recognition for its obviousappeal and marketing benefits, and although thirdparty organizers claim that companies will derivebenefits, few rigorous empirical investigations havedemonstrated the direction, strength, and validityof organizers’ claims (Roehling et al., 2005). Thus,companies entering competitions—or continuingto enter yearly—do so without clearly knowingwhether they will derive benefits, whether repeatedrecognitionsmatter, or evenwhether recognition haspotential downsides.

THEORETICAL FOUNDATION ANDHYPOTHESES DEVELOPMENT

Signaling Theory and Third Party EmploymentBranding

Strategic HR theory and research demonstrate thatHRpractices and systemscan relate toorganizationalperformance outcomes (Combs et al., 2006). How-ever, increasing attention has turned to “black box”explanations of how human capital investmentstranslate to outcomes. One important realization hasbeen that this translation likely depends in part onhow employees and other constituents perceive andinterpret such practices (Piening et al., 2013). Even iforganizations strongly invest in high performancework practices, it is possible that these practicescould be poorly or inconsistently implemented orconstituents may perceive or value them differentlythan intended. Thus, a key issue is how organiza-tions can effectively communicate to constituentsthe legitimacy of their human capital investmentsand practices, especially given that these attributesmay not be directly observable.

Signaling theory describes how labor market par-ticipants gain information when that informationis asymmetric and attributes are not directly ob-servable (Connelly et al., 2011). Originally, signaling

92 FebruaryAcademy of Management Journal

theory focused on investments in signals that in-crease job seekers’ attractiveness, such as educationand work experience (Spence, 1973). The systemwas seen as dynamic: if certain signals increase labormarket success, more people will acquire the signalswhich will diminish their effectiveness. In turn, or-ganizations will re-evaluate the utility of the signals.Seekers then acquire more of the signals, such asmore education, or acquire alternative signals thatfurther differentiate them in the marketplace. Re-peated cycles generate signaling systems that orga-nize the exchange of labor market information(Bangerter et al., 2012).

Subsequent recruitment research (e.g., Rynes,1991) has used this perspective to argue that, likeindividuals, organizations also invest in acquiringand communicating signals to job seekers about un-observable characteristics. For example, organiza-tionsmay signal their corporate social responsibility(Jones et al., 2014) as one of many appeals as theycompete with other organizations to attract talent(Bangerter et al., 2012). A foundational aspect of ourlogic is that individual reactions to organizationalsignals, such as job seekers reacting to employmentbranding signals, collectively form organization-level outcomes such as applicant pool quality. Thisidea of individual-level behaviors leading to higherlevel patterns is consistent with the evolution ofsignaling systems described in signaling theory.

Despite the advances reviewed above, at least twoimportant questions remain: Why would organiza-tions turn to third parties to signal their attractive-ness as employers; and does signaling theory applyto both job seekers and current employees? Gardneret al. (2011) propose that employment brands at-tempt to identify employment offerings and differ-entiate them from competitors. With respect toorganizational motives for using third parties to sig-nal employment brands, we propose at least two keyfeatures grounded in signaling theory and particularto third party signals such as BPTW certifications:credibility and comparability.

Credibility pertains to Gardner et al.’s (2011)identification aspect and regards the relative au-thenticity or legitimacy of third party generated in-formation versus company generated claims. Key tosignaling theory is that parties have at least partiallymisaligned goals and limited motivation to providecompletely accurate information (Spence, 1973).Companies, for example, aremotivated to emphasizepositive rather than negative signals. Self-servingclaims may have questionable veracity but willseem credible if they can be identified as honest,

reliable, costly, or hard-to-fake (Bangerter et al., 2012;Connelly et al., 2011). Further, strategic HR researchsuggests that HR systems that create consensus andconsistency regarding constituent interpretation arestronger and more effective (Bowen & Ostroff, 2004).External certifications signal consistency by implyingthat HR practices are systematic and enduringrather than transient or idiosyncratic. Thus, organi-zations may seek third party employment brandingendorsements to communicate credible signals tojob seekers or employees that the organization is agreatplace toworkand is likely toremainso.Brandingendorsements incur costs, such as time and resourceinvestments, but the signals are seen as credible andstable because they are third party bestowed andlargely outside organizational control.

Comparability pertains to Gardner et al.’s (2011)differentiation aspect; that is, whether third partygenerated information provides both absolute andrelative metrics. If a company advertises that it isa great place to work or demonstrates this to currentemployees through branding initiatives, the com-pany will certainly seem more appealing. However,these messages provide little information about theorganization relative to other organizations. In-dividuals do not typically evaluate organizations inisolation, but rather in comparison with other po-tential employers; as Collins and Kanar (2014: 286)state, “brands need to be considered relative to theircompetition.” Indeed, strategic HR research on sys-tem strength also suggests that HR systems andpractices that are distinctive are stronger and moreeffective (Bowen & Ostroff, 2004). Thus, effectivesignaling goes beyond merely signaling attractive-ness; it requires positioning an organization relativeto other organizations (Bangerter et al., 2012). Thirdparty generated certifications indicate that compa-nies are both good places to work and that they havesuperior relative standing compared with othercompanies that lack certification. The power of rel-ative versus absolute comparisons is fundamentalto broader scientific thought (e.g., Festinger, 1954;Merton, 1968); we argue that these dynamics aregermane to third party generated signals.

With this foundation, and with respect to in-cumbents and job seekers, we consider the possiblerelationships of BPTW certifications with collectiveemployee turnover rates and applicant pool quality.Of course, signal receivers have to interpret signals(Bangerter et al., 2012; Connelly et al., 2011). Even ifa signal is verifiable and consistent, as in a BPTWcertification, it may be received and interpreted dif-ferently based on the context in which it is received


(Belogolovsky & Bamberger, 2014). Thus, we alsoconsider organizational size and available industryjob openings as contextual elements that could affecthow signals are received.

Turnover Following Third Party EmploymentBranding

Are current employees influenced by third partyemployment branding? Certainly, pre-hire percep-tions and ongoing work experiences influence em-ployee perceptions of the workplace (e.g., Allen,Mahto, & Otondo, 2007; Tews, Michel, & Allen,2014). At the same time, organizations routinely at-tempt to assure current employees that they areworking in the best of places. Employees form asso-ciations about their organization as an employerbrand, which then may influence their job-relatedbehaviors, beyond objective job attributes (Collins &Kanar, 2014). Employees may doubt the credibilityof internal branding efforts, but third party employ-ment branding endorsements may be more crediblein persuading them that the organization is a greatplace to work. Specifically, intermediary-bestowedcertifications can bring public recognition that im-proves a company’s reputation and increases publicperceptions about the company as an employer.Reputation includes “stakeholders’ perceptions aboutan organization’s ability to create value relativeto competitors” (Rindova, Williamson, Petkova, &Sever, 2005: 1033) and is “best understood as anintangible asset based on broad public recognitionof the high quality of its capabilities and outputs”(Pfarrer, Pollock, & Rindova, 2010: 1133). Media cer-tifications relate to organizational prominence(Rindova et al., 2005; see also, Graffin & Ward,2010). In turn, prominence likely increases the at-tractiveness of working for and staying associatedwith a company.

Second, consistent with creating value relativeto competitors (Rindova et al., 2005), third partybranding information provides a standard of com-parison incumbents can use to assess and comparetheir relative situation: employees may leave notonly because they dislike their current situation, butalso because they have more attractive alternatives(Griffeth, Steel, Allen, & Bryan, 2005). Employeeswho work for a certified BPTW company may bereluctant to consider moving to a non-certifiedcompany. Thus, relative assessments are more pow-erful than absolute assessments; for example, “mycompany treats employees better and offers moretraining than other companies” is more persuasive

than “my company treats me well and providestraining.”

When incumbents draw favorable and apparentlycredible comparisons, they are likely to feel moreidentified with their current employer. An identityperspective suggests that BPTW certifications maybe associated with reduced turnover rates as in-dividuals wish to identify and maintain their asso-ciation with a “winning” company visibly andcredibly burnished by reputational acclaim (Tajfel &Turner, 1979; see also, Cable & Turban, 2003). In-dividuals prefer to associate with organizations thathave positive images because the association en-hances social status and self-esteem (Collins &Kanar, 2014; Dutton, Dukerich, & Harquail, 1994;Gardner et al., 2011). BPTW recognition mayalso embed more employees in the organization(e.g., Mitchell, Holtom, Lee, Sablynski, & Erez, 2001)by reducing movement desirability (e.g., Heaveyet al., 2013; March & Simon, 1958) and highlightingincurred sacrifices if they leave. Our comparabilitylogic suggests that perceived sacrifices might be-come more salient as employees feel that they areunlikely to find a similar or better situation else-where. Leaving a BPTW organization might meanrelinquishing a position at a prestigious organizationand having to justify the decision to friends, family,or associateswhomight ask, “Howwill you ever findanother company that treats its workers so well?”Thus, we expect BPTW certification signals to con-strain individual turnover decisions, leading to re-duced collective turnover rates.

Hypothesis 1. Third party employment brand-ing, in the form of BPTW certifications, will benegatively associated with collective turnoverrates.

We also propose that signal interpretation maydepend on contextual characteristics such as orga-nization size and the relative availability or scarcityof job opportunities. Signal receivers do not inter-pret signals in isolation, but in the context of otherrelevant information. First, small/large companydistinctions are important for studying companyreputation effects (Rindova et al., 2005), and orga-nization size may influence how incumbents in-terpret external third-party employment brandingsignals. One key signaling theory consideration issignal strength (Connelly et al., 2011). BPTW certi-fication signals may be relatively stronger in smallerthan in larger firms. Larger firms tend to alreadybenefit from reputational advantages (e.g., Cable& Graham, 2000); thus, the marginal gain from


additional signaling may be smaller. It may also bethe case that employees in smaller firms feel moreownership and identification with BPTW certifica-tions. In smaller companies, certification signals areless diffused among employees. The signal feelsmore strongly focused or directed at them person-ally. Therefore, because it is an employment brand-ing signal, it makes them feel they are better treatedon average than if they received the signal but wereamong many other people who also received it(i.e., in a larger company). For example, employeesmight think, “I’m partly responsible for this award”or “I’mpart of a select team that receives outstandingtreatment working for this company relative to othercompanies. Although I have other options, I’d befoolish to leave.” Conversely, in larger companies,company-directed signals are more diluted acrossemployees and therefore weaker and less personallyrelevant to any given employee. Also, because oflarger-company resource and reputational advan-tages, employees of larger companies likely alreadybenefit from social status and esteem associatedwithworking for these organizations (Turban & Cable,2003), reducing the marginal benefits of BPTW cer-tifications on identity.

Hypothesis 2. The negative relationship be-tween third party employment branding, in theform of BPTW certifications, and collectiveturnover rates will be moderated by firm sizesuch that the relationship is weaker in largerfirms and stronger in smaller firms.

In the context of evaluating the decision to main-tain or leave a job, another important consideration isthe existence and obtainability of employment al-ternatives (March&Simon, 1958; Swider, Boswell, &Zimmerman, 2011), and industry labor market con-ditions explain turnover rates (Griffeth et al., 2005).Based on our logic that BPTW certifications providecredible signals that the current employer is botha great place to work and better than non-certifiedalternatives, we suggest that the signaling effect ofcertification on turnover rates will be more pro-nounced when incumbents are motivated to makecomparisons with external alternatives, i.e., whenjob openings aremore plentiful. Withmore potentialopportunities, turnover in general should be higher,and it thus becomes more important for organiza-tions to signal the positive attributes offered to in-cumbents, particularly to the extent these attributescompare favorably with competitor offerings orwould be forfeited by leaving. As job openingsencourage individuals to make these alternative

comparisons, signal receiver attentionwill be higher(Connelly et al., 2011); thus, being able to signal thatthe “grass may not be greener” elsewhere shouldenable certified organizations to better compete toretain talent. On the other hand, when alternativesare relatively scarce, incumbents are less likely toengage in active comparisons of current to alterna-tive opportunities. Thus, although turnover in gen-eral is likely to be lower with scarcer job openings,signals about the relative favorability of employmentofferings may be relatively less important withregard to explaining turnover rates.

Hypothesis 3. The negative relationship be-tween third party employment branding, in theform of BPTW certifications, and collectiveturnover rates will be moderated by industry jobopenings such that the relationship is weakerwhen job openings are scarcer and strongerwhen job openings are more plentiful.

Repeated Signaling and Subsequent TurnoverRates

Positive publicity or endorsements have beenshown to have monotonic, simple linear relation-ships with employment brand image (e.g., Collins &Stevens, 2002). However, many organizations re-peatedly apply for BPTW certifications. This raisesthe unexamined question of whether subsequentcertifications continue to consistently relate toturnover, or whether BPTW signal utility waxes orwanes over multiple occurrences. The question istheoretically important because signaling systemstend to adapt over time (Bangerter et al., 2012). Forexample, if all organizations were designated aBPTW, the signal would lose its value. Thus, orga-nizations might logically inquire whether repeatedsignals continue to convey similar information.

One possibility, termed a crystallization effect, isthat recurrent signaling increases confidence that thesignal is credible. Thus, repeated signaling may berequired for third party branding efforts to crystallizeand translate into decreased collective turnoverrates. This relates to the definition of reputation,which implies that perceived quality and promi-nence develop only with time and consistent dem-onstration of firm capability (Pfarrer et al., 2010;Rindova et al., 2005). Specifically, just as consumers“wait to see how legitimate a new brand is goingto be” (e.g., GoogleGlass), employees might gradu-ally develop stability attributions about employee-friendly policies (Nishii, Lepak, & Schneider, 2008;


Weiner, 1986), leading them to weigh BPTW signalsmore heavily when making employment decisionsonly after their company earns multiple certifica-tions. For example, they might think, “this is oursecond certification in a row – our company musttreat us better than our competitors treat their em-ployees!” Without this consistency, however, em-ployee identity and thus turnover will stay the samebecause employees may perceive whimsy or in-stability in changes to the employment brand.

Hypothesis 4A. Turnover will be relativelyhigher with zero or one certifications, and rela-tively lower with repeated certifications.

Asecondpossibility, termedacelebrity effect, is thatrepeated transmissions dilute the relative utility of thesignal. Novel third party employment branding likelycommunicates useful, impactful, and affect-ladensignaling information about relative company com-parability, but the signaling utility may diminish overrepeated certifications. This is analogous to researchregarding celebrity firms that benefit from positiveportrayals: “those firms that attract a high level ofpublic attention and generate positive emotional re-sponses from stakeholder audiences” (Rindova et al.,2006: 51). Outside entities, such as the media, oftenconfer celebrity status associated with organizationalbrand images (Rindova et al., 2006). Initial BPTWcertification is likely tobringpublicityabout theactualannouncement and subsequent marketing highlight-ing the earned accolade. Initial certification-basedacclaim may be more likely to emotionally impactcurrent employees. That is, an initial BPTW certifi-cation reflects a substantive change in the perceptionof the employment relationship. According to ourcomparability logic, itprovidesnovelsignalsabout theemployee’s relationship with the employer and abouthow the relationship compares with other potentialemployment relationships. Rather than shift relation-ship perceptions, subsequent certifications only con-firm them. Thus, repeated signaling of the sameinformation loses novelty, emotional impact, and in-formational benefits.

Hypothesis 4B. Turnover will be relativelyhigher with zero certifications, and relativelylower with one or more certifications.

Applicant Pool Quality Following Third PartyEmployment Branding

In addition to retaining human capital, organiza-tions must compete to attract an essential but limited

resource: appropriately qualified employees (Carlson,Connerley, & Mecham, 2002). Along with economicincentives such as compensation and benefits, orga-nizations can attract applicants by signaling reputa-tion. Job seekers prefer firmswith positive reputationsbecause they associate such prestigious firms withsocial status and self-esteem (Cable & Graham, 2000;Cable & Turban, 2003; Jones et al., 2014). Previousresearch indicates that BPTW certifications arepositively associated with applicant pool quality(e.g., Collins &Han, 2004; Turban&Cable, 2003Study1); however, these studies included less than 100companies; neither controlled for ongoing peoplemanagement practices or applicant pool quality levelsprior to the certifications; and neither examinedmoderating influences of organization size or labormarket conditions.

Aswith certification effects on collective turnover,we propose that BPTW certifications credibly signalto potential applicants that a certified company iscomparatively attractive. However, whereas BPTWsignals should uniformly increase the quantity ofapplications received, we propose alternative per-spectives regarding the relationship between BPTWsignals and applicant pool quality. First, a heuristicperspective suggests that BPTW certifications mayact as cognitive cues or heuristics such that jobseekers will be less discerning in evaluating suit-ability and fit with companies. Because job seekingrequires the processing of abundant information, jobseekers, being “cognitive misers” (Fiske & Taylor,1984), will look for cognitive cues or shortcuts toreduce uncertainty and aid decision making (Rindovaet al., 2005; Roselius, 1971). This reliance onheuristicsmay lead job seekers to apply largely based on theirknowledge that companies are BPTWs. More sub-stantive company information, such as information forassessing fit, is likelyprocessedmoreperipherally, if atall. This may lead to substandard application de-cisions, lowering applicant pool quality among BPTWcompanies.

Conversely, a sorting perspective suggests thatBPTW certifications enhance applicant pool qualityfor two reasons. First, certifications signal that com-panies are highly desirable employers that canmakeparticularly selective hiring decisions. The literaturedetails how job seekers,while striving towork for thebest companies, also consider search utility. Giventime and resource constraints, they realisticallypursue opportunities with which they efficaciouslybelieve they have a chance (Kanfer, Wanberg, &Kantrowitz, 2001; Schwab, Rynes, & Aldag, 1987).For example, if a seeker has the resources to apply for


50 of the 200 job openings she identifies, she willapply for the 50 from which she feels she has thebest chance of receiving offers. With the perceptionthat BPTW companies selectively employ higher-quality workers, lower-quality job seekers havelower expectancies and are dissuaded from apply-ing, whereas better-qualified job seekers have theconfidence to apply (Rynes, 1991; Turban & Cable,2003).

Second, there is consistent evidence that jobseekers are more likely to carefully process highinvolvement recruitment information when theyhave pre-existing knowledge about the firm (Collins,2007) orwhen the firmhas a good reputation (Collins& Han, 2004). That is, communication effectivenessand message elaboration increases when signaledinformation is consistent with pre-existing knowl-edge structures; e.g., ads from familiar companiesattract more attention and trigger deeper informa-tional processing (Allen et al., 2007; Collins, 2007;Connelly et al., 2011; Rucker & Petty, 2006). Highinvolvement recruitment information refers to moredetailed information or arguments about companyattributes, such as extensive employer brochures oremployee endorsements (Collins & Han, 2004). Be-cause this information is more detailed, it requiresgreater job seeker cognitive involvement. Job seekerswho are more cognitively involved act on informa-tion more appropriately; i.e., by making job appli-cation decisions based more on useful aspects ofmessages, rather than peripheral aspects. As such,their application decisions should align with theirobjective fit (Kristof-Brown,Zimmerman,& Johnson,2005), rather than with more peripheral, and often-times inaccurate assessments of fit. Consistent withthis, Collins and Han (2004: 709) found “strongsupport” for the role of pre-established awarenesson the relationship between high involvementrecruitment practices and applicant pool quality.Specifically, when companies had establishedawareness via corporate advertising or firm reputa-tion, high involvement recruitment practices wereable to fulfill their purpose in attracting better-fittingcandidates to apply (i.e., increasing applicant poolquality). In our context, BPTW certifications mayconfirm pre-existing firm knowledge and similarlyencourage deeper information processing by jobseekers, leading to better application decisions andthus higher quality applicant pools (e.g., Dineen &Noe, 2009).

To disentangle the heuristic and sorting per-spectives, we examine company size and labormarket opportunities as potential moderators. In

terms of firm size, we propose that job seekers resortto heuristics when smaller BPTW companies ad-vertise jobs, whereas larger firms likely benefit fromsorting. As argued above, job seekers tend to haveless pre-existing knowledge about smaller compa-nies, and these companies tend to possess less le-gitimacy (Williamson,Cable, &Aldrich, 2002). Thus,with a greatermarginal need for job seekers to reduceuncertainty about applying to smaller firms, they aremore prone to rely on external signals such as BPTWcertifications. Moreover, because seekers tend tolack pre-existing coherent knowledge about smallercompanies, they are unable to match the BPTWreputational signal with prior knowledge. They arethus prone to less effortful processing of subsequentinformation (Collins, 2007; Collins & Kanar, 2014;Petty & Cacioppo, 1986), and will base applicationdecisions on the signal itself rather than on deeperprocessing of information regarding fit with sub-stantive vacancy characteristics.

Conversely, seekers will more deeply and care-fully process recruitment information from largercompanies that are BPTW-certified, consistent withthe sorting perspective. Larger companies tend to bemore familiar to job seekers. Certifications then co-here with pre-existing familiarity and reputationperceptions, and even bolster those perceptions be-cause they provide a relative comparison standardwhich complements the pre-existing absolute stan-dard. That is, as BPTW certifications confirm pre-existing perceptions (e.g., Collins &Han, 2004),morethoughtful and careful information processing oc-curs, yielding better application decisions.

Hypothesis 5. Company size will moderate therelationship between BPTW certifications andapplicant pool quality; the relationship will benegative for smaller companies and positive forlarger companies.

In terms of labor market opportunities, job seekersmaybemoreprone to cue-based (heuristic) or carefulinformation processing (sorting) depending on theextent of industry job openings. Plentiful job open-ings imply greater job vacancy information for jobseekers to process and therefore less time and re-sources to carefully consider each opening aswell asless pressure to carefully identify openings likely tobe a good fit. Faced with a plethora of informationand perceptions that there are likely many suitableopenings available, they may rely on heuristics,shortcuts, and cues when making application de-cisions (Fiske & Taylor, 1984; Petty & Cacioppo,1986; Rindova et al., 2005). Just as job seekers


use corporate reputation as a signal regarding jobattributes, and indicate job pursuit intentionsbased on that signal (Cable&Turban, 2003), plentifuljobs may induce job seekers to rely extensively oneasily observable and comparable signals such asBPTW certifications, consistent with the heuristicperspective.

Conversely, consistent with the sorting perspec-tive, when job openings are scarcer, job seekers maybe motivated to carefully process recruiting signalsin order to maximize search utility by targetingcompanies with which they believe they will likelyfit (Kanfer et al., 2001; Schwab et al., 1987). Wesuggested earlier that BPTW firms might be viewedas more selective, leading to more carefully con-sidered application decisions. With a more compet-itive labormarket, seekersmay be evenmore attunedto BPTW signals suggesting selectivity. Per searchutility arguments, higher-quality seekers may be lessconcerned about scarcity and selectivity and con-tinue to favor BPTW companies. However, lower-quality seekers may sharply and saliently perceivethat they face strong competition. Just as collegegraduates might focus their searches on companieswith which they believe they have a good chanceof attaining offers, less qualified job seekers will“hedge their bets” in a competitivemarket and focustheir efforts on non-BPTW companies relative toBPTW companies.

Hypothesis 6. Industry job openings willmoderate the relationship between BPTWcertifications and applicant pool quality; therelationship will be negative when job open-ings are more plentiful and positive when jobopenings are scarcer.

METHOD

We obtained data for 2011–2013 from an in-dependent company that organizes and conductsannual city-, state-, and industry-based BPTWcompetitions. Depending on their size category,organizations pay a $600 to $1,200 entry fee. Orga-nizational representatives complete an employersurvey with items pertaining to employment prac-tices (e.g., work–life initiatives, salaries) and out-comes (e.g., turnover). Concurrently, randomlyselected employees complete an employee engage-ment survey. For companies with fewer than 24employees, at least 80% of the employees must par-ticipate. For larger companies, at least 40% mustparticipate. In the months after data are submitted,

the competition organizer weights the company data25% and employee engagement data 75% to de-termine whether entrants are certified that year.Certifications for companies “making the list” arepublicized, and the organizer sends all companiesa detailed feedback report, regardless of outcome.

In 2011 there were 1,815 participants that pro-vided data on at least one dependent variable. Fromthis initial sample, 834 participants across 15 BPTWprograms in theUnitedStates andone inCanada alsoparticipated in 2012 and 2013, providing data on atleast one dependent variable in 2013. We sought touse industry job opening rates from the Bureau ofLabor Statistics (BLS) as a moderator variable andcontrol for existing firm HR practices standardizedby industry; however, the 23 BPTW industries rep-resented in our data did not directly parallel BLSindustries. Thus, to match BLS job openings datawith relevant industries, we independently codedtheBPTWindustries into sevenmainBLS industries,retaining and re-coding 19 agreed-upon BPTW in-dustries (BLS, 2014a; Cohen’s k 5 .74; see Table 1).For companies in industrieswewereunable tomatch,we also coded “other” responses where respon-dents provided open-ended industry descriptions,retaining agreed-upon responses (Cohen’s k 5 .71).This process yieldedBLS industry classifications for697 of the original 834 cases (83.6%). Missing dataon other study variables yielded 624 cases used in atleast one analysis and thus comprising the finalsample.1

Measures

Voluntary collective turnover rates. The volun-tary turnover measure is similar to McElroy,Morrow, and Rude (2001). Each year, company rep-resentatives completing the employer survey wereasked, “What was your organization’s percentage ofvoluntary turnover in the last fiscal year?” Partici-pants completed surveys online and additional

1 Additional information aboutmissing data for specificanalyses appears in Tables 3–5. These sample sizes ac-count for the fact that companies can enter multiplecompetitions, such as the New York and Florida compe-titions. We judged entries to be duplicating firm-wideturnover figures rather than reporting entry-specific fig-ures if they reported the same turnover for each entry in2011 and 2013. We eliminated 63 entries in the originalsample from 29 companies using this criterion, andretained 74 entries from 31 companies that reported dif-ferent turnover per entry.


“mouse-over” information was available to enableconsistent responding. The mouse-over specificallydefined voluntary turnover as “instances wheremanagement agrees that the employee had the op-tion to continue employment with the organizationat the time of separation,” specified inclusion of full-and part-time employees who were on the payroll,and provided calculation instructions. The compe-tition organizer does not include reported turnoverin its BPTW certification algorithm.

Applicant pool quality. A four-item measureadapted fromCollins andHan (2004) asked companyrepresentatives to “Consider the job applicationsyour company received during the last fiscal year.”Specific items, rated on a 1 (strongly disagree) to 5(strongly agree) scale were: (1) Overall, the skills ofthe applicants for the positions applied for meet thecompany’s needs; (2) The applicant pools we gen-erate through our recruitment efforts are high in

quality; (3) A large percentage of the applicants gener-ated for posted positions are suitable for interviewing;and (4) The candidates who apply for positions withour organization have levels of skills and abilities thatare in line with the requirements for those positions.Coefficient alphas in 2011 and 2013 were .89 and .86respectively, and the competition organizer again doesnot use this measure in the certification algorithm.

Company size. This measure combined 2013 to-tals of full- and part-time employees in the UnitedStates (or Canada for that competition), with full-time (part-time) items as follows: “At registration,your organization submitted the number below asthe total number of full-time (part-time) permanentstaff employed in theUnitedStates. Please verify thatthis number is current and accurate, and makechanges as necessary.” Because the distribution washighly right skewed (skewness5 5.69; SE5 .10), weperformed a natural log transformation of these data.

TABLE 1BPTW Sample Information (N 5 624)

Percentage of Sample by BLS Industry

Construction 3.4%Manufacturing 3.5Trade, transportation, and utilities 3.4Professional and business services 67.0Education and health services 20.8Leisure and hospitality 1.4State and local government 0.5

Percentage of Sample byCompany Size BPTW Program Informationa

Name Year Started Percentage of Sample

Less than 25 3.8% Accounting 2008 18.4%25–49 11.6 Atlantic Canada 2007 2.250–99 20.8 Florida 2009 9.8100–249 21.0 Hawaii 2005 4.5250–499 15.1 Healthcare 2008 12.3500–999 8.0 Insurance 2009 4.31000–1999 7.4 Kentucky 2005 5.82000–4999 5.9 Los Angeles 2007 6.35000 or more 6.4 Maine 2006 3.2

New York City 2008 2.6Orange County 2009 3.2Pennsylvania 2000 9.5San Diego 2009 4.2South Carolina 2006 2.2Texas 2006 9.0Virginia 2011 2.6

a Programswere classified into the following geographic regions (BLS, 2014a; industry-based programswere the baseline for dummycodingpurposes):

South: Florida, Kentucky, South Carolina, Texas, Virginia.West: Hawaii, Los Angeles, Orange County, San Diego.Northeast: Maine, Atlantic Canada, New York City, Pennsylvania.


Industry job openings.We retrieved BLS industryjob opening rates for each month in 2013 (BLS,2014b), and averaged the 12 rates to create an overallrate for each BLS industry. This measure is based ona stratified random sample of over 16,000 nonfarmbusiness and government establishments. A jobopening requires that a specific position exists andthere is work available for that position, work couldstart within 30 days, and the employer is activelyrecruiting to fill theposition (seeBLS, 2014a formorespecific information about this variable).

BPTW certifications. To assist our research, thecompany running the BPTW competitions suppliedcomplete certification data for each year of thecompetition. We used these data to construct theprimary independent variable, operationalized asthe number of certifications (0, 1, or 2) organizationsachieved during the 2011–2012 competitions.

Control variables.We included dummyvariablesfor geographic regions (see Table 1) and used archi-val data supplied by the organizing company tocreate a dummyvariable representing organizationalstatus (BPTW-certified or not) as of 2011 whenthe study began. Next, we included lagged (2011)values of our dependent variables (Wright, Gardner,Moynihan, & Allen, 2005). These were critical toassessing potential relationships between certifi-cations and 2013 outcomes while controlling forbaseline levels of those outcomes.

We also controlled for 2011 and 2013 indicesrepresenting salaries, diversity, and other HR prac-tices utilized by participating companies, in theevent those practices changed over the course of thestudy and thereby affected subsequent outcomes.For the salary index, specific items on the 2011 and2013 employer surveys were, “Average annual sal-ary for exempt employees (including partners if sal-aried)”; and “Average annual salary for non-exemptemployees.” Indices in both years averaged exemptand non-exempt salaries, each standardized by in-dustry participants. The diversity index comprisedfour items; one assessing the percentage of em-ployees that were female (which we standardized byindustry), and three assessing whether or not com-panies actively engaged in practices to recruit/retainemployees of varying ethnic backgrounds, age, ordisability status (also standardized by industry be-fore averaging it with the standardized gender item).The HR practices index comprised 65 items in fivesub-categories, including insurance benefits (14items; e.g., “vision coverage (employee)”; 1 5 notoffered; 65 employer pays 100% of premium); paidtime off (five items: numbers of paid holidays, paid

time off days after one year, vacation, sick, and per-sonal days); training (“What is the average number ofannual training and development hours received peremployee?”); performance evaluation (two itemsassessing employee performance appraisals and360-degree appraisals); and other general HR prac-tices (43 items; e.g., “onsite fitness programs,”“flexible hours”; 0 5 not offered; 1 5 offered). Westandardized groupsof similarly scaled items in eachsub-category by industry, and then averaged acrossthe five standardized sub-category scores.2

RESULTS

Means, standard deviations, and correlations ap-pear in Table 2. Of the eligible sample of 624, 17.5%were never certified, 18.9%were certified once, and63.6% were certified twice in the 2011–2012 certi-fication cycles. We conducted t-tests to assess anydifferences between the initial 2011participants thatwere and were not part of the final sample, on vari-ables as of 2011. A greater percentage of participantsin the analysis samplewere certified as of 2011whenthe study began, and had higher diversity and HRpractice indices compared with those not in thissample. Salaries, size, turnover, and applicant poolquality did not significantly differ.

Hypothesis 1 proposed that certifications would benegatively associated with turnover rates and Hy-potheses 2 and3proposed that this relationshipwouldbe moderated by firm size and industry job openings.After entering control variables in a first step, weregressed 2013 turnover on the number of certifica-tions achieved in 2011–2012 in a second step, the twomoderators in a third step, and the interaction terms ina fourth step. Results revealed a negative relationshipbetween certifications and turnover (DR2 5 .01; b 52.10;p, .05; seeTable 3,Model 2), providing supportfor Hypothesis 1. However, Table 3 indicates that nei-ther Hypotheses 2 nor 3 were supported. Hypotheses4A and 4B proposed competing crystallization andcelebrity effects: turnover would be relatively lowerfollowing twocertifications, compared to either zeroorone certification; or relatively higher with zero certifi-cations, compared to either one or two certifications.To test these hypotheses we constructed and enteredtwo effects coded contrasts after the control variables

2 Detailed information about the index items and de-velopment process are available from the first author.Company representatives also provided survey comple-tion times, and results were unchanged when controllingfor this.


TABLE2

Mea

ns,StandardDev

iation

s,an

dCorrelation

sam

ongStudyVariables

Mea

nSD

12

34

56

78

910

1112

1314

1516

1.Sou

thregion

dummy

.29

.46

2.Westregion

dummy

.18

.39

–.30*

*

3.Northeast

region

dummy

.17

.38

2.30*

*–.22*

*

4.Certified

asof

2011

.55

.50

–.03

.04

.04

5.Salaryindex

2011

–.03

.63

–.13*

*.071

–.09*

.071

6.Salaryindex

2013

–.05

.52

–.071

–.00

–.03

.03

.67*

*

7.Diversity

index

2011

–.00

.69

.00

–.10*

.05

.09*

–.09*

–.11*

*

8.Diversity

index

2013

.01

.70

.02

–.10*

.04

.10*

*–.11*

*–.08*

.71*

*

9.HRpractices

index

2011

–.01

.43

–.06

–.00

.071

.15*

*.08*

.13*

*.15*

*.17*

*

10.H

Rpractices

index

2013

–.00

.41

–.03

.03

–.01

.11*

*.09*

.11*

*.071

.15*

*.73*

*

11.A

pplicant

pool

quality

2011

3.55

.80

–.04

–.03

–.11*

–.02

–.02

–.03

.07

.06

.13*

*.13*

*

12.T

urno

ver

2011

10.29a

11.34

.081

–.03

.08*

–.03

–.16*

*–.11*

*.03

.03

–.04

–.05

–.07

13.C

ertification

s20

11–20

121.46

.77

.15*

*.18*

*–.06

.32*

*.02

–.03

.02

.05

.18*

*.23*

*.01

–.05

14.O

rgan

ization

size

5.49

b1.69

.12*

*–.04

.01

.04

.13*

*.12*

*.12*

*.15*

*.27*

*.18*

*.06

.13*

*–.071

15.Industry

job

open

ings

2013

3.36

a.44

–.06

–.01

–.11*

*.06

–.03

.00

–.01

–.00

.01

–.01

–.081

.01

.09*

–.23*

*

16.A

pplicant

pool

quality

2013

3.51

.76

.07

–.10*

–.08*

–.05

.00

–.06

.09*

.05

.081

.14*

*.41*

*–.04

–.00

.12*

*–.02

17.T

urno

ver

2013

12.46a

14.88

.03

.04

–.05

.01

–.04

–.11*

*.03

.06

–.02

–.03

.00

.26*

*–.071

.05

.03

.00

Note:

n5

betw

een46

6an

d62

4.E.g.,ap

plicantp

oolq

ualitywas

not

assessed

in20

11forfourof

the16

BPTW

program

s.aListedas

aperce

ntage

(i.e.,“10

.29”

means10

.29%

).bNaturallog

valuesh

own.

1p,

.10

*p,

.05

**p,

.01


(see Table 4). These represented differences in 2013turnover between: (1) participants not certified, andcertified once or twice; and between (2) participantscertified twice, and once or not at all (Cohen, Cohen,West, & Aiken, 2003).3 Results indicated an overallsignificant effect for the contrasts (DR25 .01; p, .05),with support for the celebrity perspective (Hypothesis4B): the coefficient for the effects coded contrastbetween 0 and 1 or 2 certifications was significant(b 5 4.36; SE 5 1.63; b 5 .13; p , .01) and thecontrast between 2 and 1 or 0 certifications was not(b5 .10; SE5 .57; b5 .01; n.s.). As Figure 1 shows,turnover is lower with one certification comparedto none, but is essentially the same with one or twocertifications.

Table 5 andFigures 2Aand2B illustrate the resultsof Hypotheses 5 and 6, which proposed moderatingeffects of company size and industry job openings onthe BPTW certification-applicant pool quality re-lationship. Specifically, the overall D R square was

significant (DR2 5 .01, p, .05, Model 4) and the testof Hypothesis 5 was significant but counter to ourprediction (b 5 2.22, p , .05). As shown inFigure 2A, BPTW certifications were positivelyrather than negatively associated with applicantpool quality among smaller companies (simpleslope: t 5 2.10; p , .05). Certifications and appli-cant pool quality showed no relationship amonglarger companies (t 5 21.17; n.s.). Table 5 andFigure 2B show that Hypothesis 6 was partiallysupported. Specifically, a marginally significantoverall interactioneffectoccurred (b52.15;p, .08),with a simple slope indicating a positive relation-ship between certifications and applicant poolquality when job openings were scarcer (t 5 1.73,p , .09). The relationship was non-significant,however, when job openings were more plentiful(t 5 2.63; n.s.).

Supplemental Analyses

We conducted supplemental analyses to furtherexamine ways certifications might manifest in turn-over and applicant pool quality. Data were available

TABLE 3Regression Results: Turnovera

Variables Model 1 Model 2 Model 3 Model 4

Control VariablesSouth region dummy –.00 .02 .03 .02West region dummy .03 .06 .06 .06Northeastern region dummy –.07 –.06 –.06 –.06BPTW-certified as of 2011 .02 .04 .04 .04Salary index 2011 .08 .09 .09 .09Salary index 2013 –.13* –.14** –.14** –.14**Diversity index 2011 –.04 –.04 –.04 –.04Diversity index 2013 .09 .09 .09 .09HR practices index 2011 .02 .03 .02 .02HR practices index 2013 –.04 –.02 –.02 –.02Turnover 2011 .27** .26** .26** .26**Independent VariableBPTW certifications 2011–2012 –.10* –.10* –.10*Moderating VariablesOrganization size .01 –.05Industry job openings .03 –.01Interaction TermsBPTW certifications x size .06BPTW certifications x job

openings.05

R2 .09 .10 .10 .10DR2 .09** .01* .00 .00

Note: n 5 617. This sample is reduced from its original size because seven companies did not report turnover in either 2011 or 2013.a Standardized regression coefficients shown. Dependent variable is 2013 turnover.* p , .05

** p , .01

3 We used weighted contrasts because of the unequalcell sizes among participants certified 0, 1, and 2 times;results were similar using non-weighted contrasts.


among BPTW-certified companies indicating spe-cific rankings (e.g., 1st, 5th, 20th). We sought,therefore, to analyze whether rankings were associ-atedwith the predicted trends, in twoways. First, weretested hypotheses among certified companies,replacing the 2011–2012 certifications variable withan average 2011 and 2012 ranking variable (e.g., ifa company was ranked 5th in 2011 and 11th in 2012

its average rank was 8th). We found a main effect onturnover (DR25 .01;b5 .13;p, .05), such that betterrankings (e.g., 8th is better than 18th), related tolower 2013 turnover. Second, we created a vari-able representing the percentage difference betweenrankings these two years, multiplying “worse”rankings differences (e.g., from 8th to 18th) by 21.We found a main effect on applicant pool quality(DR2 5 .01; b 5 .11; p , .05), such that percentagedifferences in rankings across years (in a favorabledirection) were associated with higher 2013 appli-cant pool quality.

DISCUSSION

We combine and advance the turnover, recruit-ment, and employment branding literatures by de-veloping and testing predictions regarding thirdparty employment branding as it relates to turnoverrates and applicant pool quality across a three-yearperiodwhile controlling for ongoing investments inHR practices. Specifically, we draw on signalingand strategic HR theories to explain why organiza-tions might engage third parties to provide crediblesignals to alert potential or current employees that

TABLE 4Regression Results: Turnover Effects Coded Contrastsa

Variables Model 1 Model 2

Control VariablesIndustry dummy variables b b

South region dummy –.13 (1.53) .87 (1.57)West region dummy 1.12 (1.73) 2.15 (1.77)Northeastern region dummy 22.71 (1.80) 22.39 (1.79)BPTW-certified as of 2011 .40 (1.20) 1.26 (1.25)Salary index 2011 1.26 (.80) 1.26 (.79)Salary index 2013 21.97 (.78) * 22.00 (.78) *Diversity index 2011 –.58 (.83) –.63 (.83)Diversity index 2013 1.34 (.83) 1.33 (.83)HR practices index 2011 .40 (.87) .47 (.86)HR practices index 2013 –.64 (.86) –.44 (.87)Turnover 2011 4.05 (.60) ** 4.00 (.60) **Effects Coded Terms0 versus 1 or 2 certifications 4.36 (1.63) **2 versus 0 or 1 certifications .10 (.57)R2 .10 .11DR2 .10** .01*

Note: n 5 617. This sample is reduced from its original size because seven companies did not report turnover in either 2011 or 2013.a Because standardized coefficients are not invariant to changes in group size when weighted contrasts are used, unstandardized re-

gression coefficients (standard errors) are shown. Dependent variable is 2013 turnover.b Because this analysis did not include industry job openings, we controlled for potential industry effects using industry dummy controls.

Individual industry coefficients not shown in the interest of space.* p , .05

** p , .01

FIGURE 1Celebrity Effect of Certifications on Turnover:

2011–2013

BPTW Certifications: 2011–2012

2013

Tu

rnov

er

20%

18%

16%

14%

12%

10%

8%0 1 2


the company’s employment brand compares fa-vorably across companies. These signals declarethat the company is a great place to work andsuggest that current and future employees arebest served by joining or remaining with thecompany. Moreover, the signals likely createperceptions of consensus, consistency, and dis-tinctiveness, three key aspects of constituent in-terpretation which can increase the efficacy ofHR systems.

Our results have a threefold impact. First, we rec-ognize that companies use employment branding toattract talent, but we also consider that companieshave at least an equal interest in using branding toretain talent. Second, in addition to consideringorganization-generated branding in the form of HRpractice constellations provided to attract and retainemployees, we specifically consider incrementaleffects of third party-initiated branding, which iscritically different in that it provides enhancedcredibility and comparability. Here we find thatBPTW certification signals explain key organiza-tional outcomes beyond the effects of existing

systems ofHRpractices,whichmight be “stronger”and more immutable in BPTW-participating com-panies (Bowen & Ostroff, 2004). Third, we providea unique study of third party employment brand-ing that extends prior work (e.g., Collins & Han,2004; Fulmer et al., 2003) by using actual datagathered during multiple certification processes,and addresses calls from strategic HR scholars forlongitudinal studies that account for lagged per-formance metrics (e.g., Wright et al., 2005). Thearray of findings extends our understanding of thecontextwithinwhich employment branding signalsare interpreted, shows that they matter beyond in-strumental HR offerings (Lievens & Highhouse,2003), and demonstrates that signaling might dif-fer considerably in effects on job seekers and or-ganizational incumbents.

Overall, results show that BPTW certifications, asthird party employment branding efforts, are asso-ciated with lower collective turnover rates in-dependent of firm size or industry job openings; thatinitial certifications are more strongly associatedwith turnover than repeated instances; and that

TABLE 5Regression Results: Applicant Pool Qualitya

Variables Model 1 Model 2 Model 3 Model 4

Control VariablesSouth region dummy .04 .03 .01 .03West region dummy –.05 –.05 –.05 –.04Northeastern region dummy –.05 –.06 –.06 –.05BPTW-certified as of 2011 –.06 –.07 –.07 –.06Salary index 2011 .09 .09 .07 .07Salary index 2013 –.13* –.13* –.12* –.13*Diversity index 2011 .08 .08 .08 .07Diversity index 2013 .00 .00 –.02 –.01HR practices index 2011 –.03 –.03 –.07 –.06HR practices index 2013 .121 .111 .111 .10Applicant pool quality 2011 .38** .38** .38** .37**Independent VariableBPTW certifications 2011–2012 .02 .04 .03Moderating VariablesOrganization size .14** .33**Industry job openings .03 .161

Interaction TermsBPTW certifications x size –.22*BPTWcertificationsx jobopenings –.151

R2 .20 .20 .21 .22DR2 .20** .00 .01* .01*

Note: n 5 466. This sample is reduced from its original size because 120 companies were from four programs that did not assess 2011applicant pool quality, and 38 companies did not report applicant pool quality in either 2011 or 2013.

a Standardized regression coefficients shown. Dependent variable is 2013 applicant pool quality.1 p , .10* p , .05

** p , .01


certification effects on applicant pool quality areconditioned on firm size and job opening rates. Thestudy also addresses calls to investigate alternativeexplanations for turnover (e.g., Holtom, Mitchell,Lee, & Eberly, 2008), contributes to recent models ofcollective turnover by incorporating branding ef-fects as a potentially important antecedent (e.g.,Hausknecht & Trevor, 2011; Heavey et al., 2013),and indicates that third parties may be importantparticipants in signaling systems.

More specifically, the turnover findings are con-sistent with what we term an identity effect: crediblethird party employment branding may strengthenincumbents’ identification with the firm, furtherembedding employees by sharpening the benefits ofremaining compared with leaving for seemingly less

desirable opportunities. This finding held acrossfirm sizes and industry job opening rates, and ex-tends signaling theory beyond its traditional HR ap-plication to job seeking. Specifically, organizationalincumbents already know about working conditionsat their current companies, so superficially theymight discount third party information in theirturnover deliberations. But they are often unclearabout the relative advantages of alternative oppor-tunities, so that third party generated employmentbranding signals are potentially very useful to jobseekers and incumbents alike for signaling compar-ative information about other opportunities. Wefind this result while controlling for a host of ongo-ing HR practices, another important extension toprior work on the signaling effects of reputation and

FIGURE 2Moderating Effects of Company Size and Industry Job Openings on the Relationship between Certifications and

Applicant Pool Quality: 2011–2013

A4.54.3

4.13.9

3.73.53.3

3.12.9

2.72.5

0 1 2


2013

Ap

pli

can

t P

ool

Qu

alit

y

Smaller Companies(–1 SD = 42 employees)

Larger Companies(+1 SD = 1,120 employees)

4.54.3

4.13.9

3.73.53.3

3.12.9

2.72.520

13 A

pp

lica

nt

Poo

l Q

ual

ity

0 1 2


B

Scarcer Job Openings(–1 SD = 2.90%)

Plentiful Job Openings(+1 SD = 3.80%)

A: Company Size Moderation EffectB: Industry Job Opening Moderation Effect


certification. This suggests that signals can matterbeyond objective job characteristics (Lievens &Highhouse, 2003) and even the personal experi-ences with the organization that incumbents al-ready have.

We further develop crystallization and celebrityperspectives to examine the effects of multiple thirdparty employment branding certifications over time.Signaling theory describes signaling systems as dy-namic, with the meaning and strength of signalschanging with repeated signaling (Connelly et al.,2011), a previously unexamined possibility in thecontext of employment branding. Our results aremore consistent with what we term a celebrity per-spective: turnover appears to be lower following aninitial certification, but a repeat certification is notassociated with even lower turnover. In addition toclarifying the role of third party branding efforts, thisfinding addresses the under-researched issue of sig-nal utility over repeated episodes. Our findingssuggest that the novelty of an initial signal is moreimportant than the consistencyprovidedby repeatedsignaling over time; identity-enhancing effects maydiminish or even extinguish after the initial signal.Thus an initial third party-generated signal is suffi-ciently credible. When employees first learn thattheir organization is BPTW-certified, they will haveheightened organizational identification, alignedwith work examining celebrity firms (Pfarrer et al.,2010; Rindova et al., 2006). However, when theylearn that their organization is still certified, the en-hanced reputation through longer-term signal con-sistency (Rindova et al., 2005) has less power tofurther augment their identification or credibilityperceptions.

We also find that BPTW certifications are con-ditionally related with applicant pool quality,extending signaling theory by showing that signalinterpretation depends in part on context. First, wefind BPTW certifications to be associated withbetter applicant pool quality for smaller but notlarger firms, counter to our original prediction.Although we suggested the consistency of BPTWcertifications with prior employer knowledge amongjob seekers assessing larger firms would lead tomore effortful processing of information by thoseseekers and thus enhanced applicant pool quality,it may be that there is a ceiling effect whereby it ismore difficult to change job seeker perceptions offirms they may already be familiar with (Cable &Graham, 2000). Even an external signal consis-tent with this pre-existing knowledge may not beenough to alter job seeker information processing.

Smaller firms, on the other hand, appear to havemore to gain by acquiring additional reputationalsignals, perhaps by better-attracting higher qualityjob seekers who previously only considered largerfirms with more established reputations. Thus, thesignaling strength of small-firm BPTW certifica-tions may be greater.

Similarly, labor market context also affects in-terpretation of third party employment brandingsignals. When alternative job openings are scarce,BPTW certifications are positively related to ap-plicant pool quality, but this relationship is notsignificant when alternative openings are moreplentiful. Thus, when jobs are scarcer, BPTW sig-nalsmay help potential applicants focus their timeand energy on firms more likely to be an appro-priate fit. Regarding the applicant pool qualityfindings relative to turnover findings, it is partic-ularly important to study the incremental effectsof BPTW certifications beyond existing HR prac-tices. This is because those HR practices maynot be as apparent or observable to job seekers asthey are to organizational incumbents, whereas certi-fications should be. Also, it is important to recognizethat increased quality may occur on a relative basis,for reasons including: (1) consistent numbers of qual-ifiedapplicantsanddecreasednumbersofunqualifiedapplicants; (2) consistent numbers of unquali-fied applicants and increased numbers of quali-fied applicants; or (3) increased numbers of both,but a relatively greater increase in qualified ap-plicants. While our results indicate situationswhere BPTW certifications are positively associ-ated with applicant pool quality, and we surmisethat the third scenario above is most likely, weencourage future research to examine how greaterquality specifically manifests. Overall, our resultssuggest the importance of context in influenc-ing how workplace-related signals are interpreted(Belogolovsky & Bamberger, 2014). The resultpattern also suggests that BPTW certificationeffects tend to be qualified among external con-stituents (applicant pool quality) but are moreuniform among internal constituents (incumbentturnover).

Limitations

Our results should be viewed in concertwith somelimitations, many of which derive from the archivalnature of these data. First, organization mortalitylimits the generalizability of our results. Specifically,although we could track our outcome variables


among companies that continued competing inBPTW programs, companies that initially competedbut then stopped provided no more data once theywithdrew. Thus, for example, we could not trackturnover from 2011 to 2013 among companies thatentered competitions in 2011 but not in 2013.Certifiedparticipants: (1) tendedtocontinuecompetingmore frequently than non-certified participants; and(2) tended to be recertified more frequently in sub-sequent competitions (e.g., 83% of 2011-certifiedcompanies were recertified in 2012). A stronger de-sign would have included not only organizationsthat joined the competition and failed to achievecertification, but also organizations that never joinedthe competition at all.

Second, the archival data prohibited direct as-sessments of some of the proposed underlyingindividual-level processes and mediating mecha-nisms. For example, we suggest that lower turnoveris partially a function of increased individualidentification with and embeddedness in theorganization, that reputation perceptions increasewhen companies are certified, and that actors ex-perience heightened emotions when BPTW signalsare novel and unique. We also provide conceptualextensions to signaling theory by proposing thatthird party-generated signals are more credibleand facilitate standardized comparisons acrosscompanies. Yet we could not specifically assessthose dynamics. Thus, while controlling for laggedoutcomes and HR practices helps better isolatecertification effects, the precise reason for the ef-fects remains uncertain. Also, if we had data fromadditional competition cycles, we might haveexplored potential longer-term mediations. Forexample, BPTW signals might affect long-run turn-over by first influencing the quality of applicantsattracted to firms. Future research that addressesthose possibilities over longer time periods wouldlend additional insight to our theoretical perspec-tive (Wright et al., 2005).

Third, baseline turnover among BPTW entrantsmay have been lower than turnover in the pop-ulation, which could attenuate observed effects. Thecompetition organizer offers an extensive feedbackreport which could entice non-elite companies thatwish to improve, but competition entrants are al-ready likely among the elite in people-managementpractices, or at least they perceive themselves tobe. Thus entrants might have potentially lowerturnover on average compared with typical non-competing organizations. Our design and HRpractice and lagged dependent variable controls

increase our confidence that differences in turn-over rates and applicant pool quality can be at-tributed to BPTW certifications, but we cannotdirectly assert that the certifications caused thesedifferences. As it is not feasible to randomly assignactual firms to BPTW status in the field, future re-search that manipulates third party employmentbranding signals in controlled environments couldshed some light on whether these signals changeindividual perceptions and ultimately emerge asfirm-level outcomes.

Fourth, these data are from a time of sluggish ac-tivity in the U.S. economy and labor markets, whichcould have affected employee mobility perceptions,although certified and non-certified companies insimilar geographic regions faced similar economies.We examined moderating effects of industry jobopenings to partially address the role of labormarketconditions and potential mobility. However, futureresearch replicating the findings in stronger labormarketsmay bewarranted. Fifth, company-supplieddata such as turnover rates and salaries shouldbe relatively straightforward to report, but somecompaniesmay track statistics better than others.Wealso recognize that using representatives to furnishorganizational-level data such as applicant poolquality may have potential drawbacks in that thedata may be perceptual. Others have used similarmeasures (e.g., Collins & Han, 2004), but some argueagainst using company-informant data (e.g., Gerhart,Wright,McMahan, &Snell, 2000). Thismaypartiallyexplain our modest effect sizes, although our turn-over relationships, for example, are in line withmeta-analytic estimates for climate and turnoverrates (Heavey et al., 2013), similar to effect sizebenchmarks for relationships of employee evalua-tionswithvoluntary turnover (Bosco,Aguinis, Singh,Field, & Pierce, 2015), and non-trivial from a practi-cal perspective (Hancock, Allen, Bosco, Pierce, &McDaniel, 2013).

Practical Implications and Future ResearchDirections

Despite these limitations, our study offers severalpractical insights. First, organizations cumulativelyspend millions of dollars in annual BPTW entryfees (Burke, 2014). We show that they are makingtheoretically and practically sound choices by con-sidering third party employment branding in-vestments because it appears they are associatedwith lower turnover rates. Participating compa-nies also receive extensive feedback reports and


benchmarking data worth $5,000 to $10,000 in con-sulting fees, according to the organizing company.However, entering competitions carries demandsand possible disadvantages. Companies must askemployees to complete surveys; if the company failsto succeed in the competition, employees may havenegative reactions. Likewise, considerable frontedcosts are involved in attaining BPTW status. Com-panies must invest in practices ranging from visioninsurance, to fitness or wellness programs if theyhope to be labeled employee-friendly. Also, beyondjustifying the expense of entering competitions once,it may be even more important to justify the costs ofentering multiple times. For example, HR managersand CEOs need to know whether they should re-enter after being certified, or how many times theyshould fail before they stop trying. Our results sug-gest overall positive results but potentially dimin-ishing returns to multiple certifications: initialcertification associates with lower turnover, butfurther certifications may not be marginally advan-tageous. Of course, BPTW success carries severalother benefits such as the higher applicant poolquality we observed or even possible spillovers toconsumer reactions and purchasing patterns. Spe-cific rankings among certified companies also seemto matter. Moreover, our research identifies com-pany sizes that are likely to benefit and the optimaljob markets that will convey benefits in terms of ap-plicant pool quality. Better-qualified job seekers ap-pear to applymore readily to smaller BPTW-certifiedcompanies or to BPTW companies when job open-ings are scarce. Smaller organizations may havefewer resources to invest in third party employmentbranding efforts, so it is useful to provide evidenceregarding when they may benefit from third partycertifications.

We encourage scholars to pursue several relatedresearch directions. We focus primarily on signal-ing theory as a meso-level theory describing howfirm-level activities influence individual reactionsthat in turn influence firm-level outcomes, butother theoretical perspectives may provide alter-native insights. For example, social informationprocessing theory (e.g., Salancik & Pfeffer, 1978)compellingly explains how individuals interpretwork conditions in relation to social informationregarding others’ interpretations and reactions.Particularly with respect to incumbent reactions toBPTW certifications, scholars might explore howsignal interpretation spreads contagiously amongemployees. Similarly, employee attributions aboutfirmmotivations for pursuing and obtaining BPTW

certification may affect responses, as might per-ceptions of certification rarity; i.e., howmany firmsapply and thus certification prestige. Other relatedresearch would be fruitful, such as attributionspertaining to certification in one year followed byfailure to be certified in the next; attributions aboutpositive third party employment branding, partic-ularly among employees who have unfavorableviews of company employment practices; and na-tional versus regional certification effects. Also,when studying third party branding, scholarsshould continue to account for employment prac-tice effects.

Although BPTW success seems to benefit turn-over and applicant pool quality, future workshould examine other potential outcomes, such aslonger-term financial metrics, product marketingbenefits from one-time or continued third partyemployment branding certifications, and whetherturnover tends to comprise better or worse per-forming employees. Our initial attempt to de-velop theoretical perspectives could be overlaidon additional studies considering, for example,whether current customers might quickly becomemore loyal to certified companies, whereas con-sumers of competing products might require morestable certification patterns before switching prod-uct allegiances.

Future work should also consider the longerterm, more subtle results of third party certifica-tions. For example, even though we believe thesignaling value of being “on” or “off” the list ismoretheoretically interesting than the specific certifica-tion attained, studies might build on our supple-mental results to examine specific changes incertification levels across years (e.g., moving from10th to 1st). Also, as shown in Table 1, our samplecomprised relatively smaller companies, reflectiveof the fact that 99% of businesses are typicallyconsidered small, but still possibly limiting thegeneralizability of our results. It would be inter-esting for future work to examine these dynamicsamong firms with larger size ranges, given thatlarger firms employ over 50% of the total workforce(United States Census Bureau, 2015).

Finally, certifiedorganizationsmight beheldmoreaccountable to uphold their image. McCracken(2000) coined the phrase “let the world watch you”to describe how Deloitte & Touche employed in-dependent panels to hold it accountable for im-proving gender diversity, and similar dynamicsmight be at play with BPTW certifications. Simulta-neously, competing organizations that fail to be


certified might become more aware that they mustimprove their employment practices.4 For example,a control theory perspective (Carver & Scheier, 2000)would suggest that BPTW-successful companiesmight “rest on their laurels” and avoid taking furthersteps to enhance their people-management. Yetother firms might build on their success and seek tofurther differentiate themselves from competitors onemployee-related issues. We view our current re-search as a useful starting point for such continuedinvestigations.

REFERENCES

Allen, D. G.,Mahto, R. V., &Otondo, R. F. 2007.Web-basedrecruitment: Effects of information, organizationalbrand, and attitudes toward a Web site on applicantattraction. The Journal of Applied Psychology, 92:1696–1708.

Bangerter, A., Roulin, N., & Konig, C. J. 2012. Personnelselection as a signaling game. The Journal of AppliedPsychology, 97: 719–738.

Barney, J. B. 2001. Resource-based theories of competi-tive advantage: A ten-year retrospective on theresource-based view. Journal of Management, 27:643–650.

Becker, B. E., & Huselid, M. A. 2006. Strategic human re-sources management: Where do we go from here?Journal of Management, 32: 898–925.

Belogolovsky, E., &Bamberger, P. 2014. Signaling in secret:Pay for performance and the incentive and sorting ef-fects of pay secrecy. Academy of Management Jour-nal, 57: 1706–1733.

BLS 2014a. Job openings and labor turnover—January2014. http://www.bls.gov/news.release/archives/jolts_03112014.pdf. Accessed on January 30,2015.

BLS 2014b. Job openings and labor turnover archivednews releases. http://www.bls.gov/schedule/archives/jolts_nr.htm. Accessed on January 30, 2015.

Bosco, F. A., Aguinis, H., Singh, K., Field, J. G., &Pierce, C. A. 2015. Correlational effect size bench-marks. The Journal of Applied Psychology, 100:431–449.

Bowen, D. E., & Ostroff, C. 2004. Understanding HRM-firmperformance linkages: The role of the “strength” of theHRM system. Academy of Management Review, 29:203–221.

Burke, P. 2014, August 12. Personal communication.

Cable, D. M., & Graham, M. 2000. The determinants of jobseekers’ reputation perceptions. Journal of Organi-zational Behavior, 21: 929–947.

Cable, D. M., & Turban, D. B. 2001. Establishing thedimensions, sources and value of job seekers’ employerknowledge during recruitment. In K.M. Rowland &G.R. Ferris (Eds.), Research in personnel and hu-man resource management: 115–163. Greenwich,CT: JAI Press.

Cable, D. M., & Turban, D. B. 2003. The value of organiza-tional reputation in the recruitment context: A brandequity perspective. Journal of Applied Social Psy-chology, 33: 2244–2266.

Carlson, K. D., Connerley,M. L., &Mecham, R. L., III. 2002.Recruitment evaluation: The case for assessing thequality of applicants attracted. Personnel Psychol-ogy, 55: 461–490.

Carver, C. S., & Scheier, M. F. 2000. On the struc-ture of behavioral self-regulation. In M. Boekaerts,P. R. Pintrich, & M. Zeidner (Eds.), Handbook ofself-regulation: 41–84. San Diego, CA: AcademicPress.

Chapman, D. S., Uggerslev, K. L., Carroll, S. A., Piasentin,K. A., & Jones, D. A. 2005. Applicant attraction to or-ganizations and job choice: A meta-analytic review ofthe correlates of recruiting outcomes. The Journal ofApplied Psychology, 90: 928–944.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. 2003. Ap-plied multiple regression/correlation analysisfor the behavioral sciences (3rd ed.). Mahwah, NJ:Lawrence Erlbaum.

Collins, C. J. 2007. The interactive effects of recruitmentpractices and product awareness on job seekers’ em-ployer knowledge and application behaviors. TheJournal of Applied Psychology, 92: 180–190.

Collins, C. J., & Han, J. 2004. Exploring applicantpool quantity and quality: The effects of early re-cruitment practice strategies, corporate advertis-ing, and firm reputation.Personnel Psychology, 57:685–717.

Collins, C. J., & Kanar, A. M. 2014. Employer brand equityand recruitment research. In K. Y. T. Yu & D. M. Cable(Eds.) Oxford handbook of recruitment: 284–297.New York, NY: Oxford University Press.

Collins, C. J., & Stevens, C. K. 2002. The relationship be-tween early recruitment-related activities and the ap-plication decisions of new labor-market entrants: Abrand equity approach to recruitment. The Journal ofApplied Psychology, 87: 1121–1133.

Combs, J., Liu, Y., Hall, A., & Ketchen, D. 2006. Howmuchdo high-performance work practices matter? A meta-analysis of their effects on organizational perfor-mance. Personnel Psychology, 59: 501–528.4 We thank the associate editor for this suggestion.


http://www.bls.gov/news.release/archives/jolts_03112014.pdf

http://www.bls.gov/news.release/archives/jolts_03112014.pdf

http://www.bls.gov/schedule/archives/jolts_nr.htm

http://www.bls.gov/schedule/archives/jolts_nr.htm

Connelly, B. L., Certo, S. T., Ireland, R. D., & Reutzel, C. R.2011. Signaling theory: A review and assessment.Journal of Management, 37: 39–67.

Dineen,B.R., &Noe,R.A. 2009. Effects of customizationonapplication decisions and applicant pool characteris-tics in a web-based recruitment context. The Journalof Applied Psychology, 94: 224–234.

Dutton, J. E., Dukerich, J. M., & Harquail, C. V. 1994.Organizational images and member identification.Administrative Science Quarterly, 39: 239–263.

Edwards, M. R. 2010. An integrative review of employerbranding andOB theory.Personnel Review, 39: 5–23.

Festinger, L. 1954. A theory of social comparison pro-cesses. Human Relations, 7: 117–140.

Fiske, S.T., &Taylor, S. E. 1984.Social cognition. Reading,MA: Addison-Wesley Pub. Co.

Fulmer, I., Gerhart, B., & Scott, K. 2003. Are the 100best better? An empirical investigation of the re-lationship between being a “great place to work”and firm performance. Personnel Psychology, 56:965–993.

Gardner, T. M., Erhardt, N. L., & Martin-Rios, C. 2011.Rebranding employment branding: Establishinga new research agenda to explore the attributes,antecedents, and consequences of workers’ employ-ment brand knowledge. Research in Personnel andHuman Resources Management, 30: 253–304.

Gerhart, B., Wright, P. M., McMahan, G. C., & Snell, S. A.2000. Measurement error in research on human re-sources and firm performance: How much error isthere and how does it influence effect size estimates?Personnel Psychology, 53: 803–834.

Graffin, S. D., & Ward, A. J. 2010. Certifications andreputation: Determining the standard of desirabil-ity amidst uncertainty. Organization Science, 21:331–346.

Griffeth, R. W., Steel, R. P., Allen, D. G., & Bryan, N. 2005.The development of a multidimensional measure ofjob market cognitions: The Employment OpportunityIndex (EOI). The Journal of Applied Psychology, 90:335–349.

Hancock, J. I., Allen, D. G., Bosco, F. A., Pierce, C. A., &McDaniel, K. 2013. Meta-analytic review of employeeturnover as apredictor of firmperformance. Journal ofManagement, 39: 573–603.

Hausknecht, J. P., & Trevor, C. O. 2011. Collective turnoverat the group, unit, andorganizational levels: Evidence,issues, and implications. Journal ofManagement, 37:352–388.

Heavey, A. L., Holwerda, J. A., & Hausknecht, J. P. 2013.Causes and consequences of collective turnover: Ameta-analytic review. The Journal of Applied Psy-chology, 98: 412–453.

Holtom, B. C., Mitchell, T. R., Lee, T. W., & Eberly, M. B.2008. Turnover and retention research: A glance at thepast, a closer review of the present, and venture intothe future. The Academy of Management Annals, 2:231–274.

Huselid, M. A. 1995. The impact of human resource man-agement practices on turnover, productivity, andcorporate financial performance. Academy of Man-agement Journal, 38: 635–672.

Huselid,M.A., &Becker, B. E. 1996.Methodological issuesin cross-sectional and panel estimates of the humanresource-firm performance link. Industrial Relations,35: 400–422.

Jiang, K., Lepak, D. P., Hu, J., & Baer, J. C. 2012. How doeshuman resource management influence organiza-tional outcomes? A meta-analytic investigation ofmediating mechanisms. Academy of ManagementJournal, 55: 1264–1294.

Jones, D., Willness, C., & Madey, S. 2014. Why are jobseekers attracted by corporate social performance?Experimental and field tests of three signal-basedmechanisms. Academy of Management Journal, 57:383–404.

Kanfer, R., Wanberg, C. R., & Kantrowitz, T. M. 2001. Jobsearch and employment: A personality-motivationalanalysis and meta-analytic review. The Journal ofApplied Psychology, 86: 837–855.

Katzenbach, J. R., Steffen, I., & Kronley, C. 2012. Culturalchange that sticks. Harvard Business Review, 90:110–117.

Kristof-Brown, A., Zimmerman, R., & Johnson, E. 2005. Con-sequencesof individuals’ fit at work: Ameta-analysisof person-job, person-organization, person-group,and person-supervisor fit. Personnel Psychology,58: 281–342.

Lau, R. S. M., & May, B. 1998. A win-win paradigmfor quality for work life and business perfor-mance.Human Resource Development Quarterly,211–226.

Lievens, F. 2007. Employer branding in the Belgian army:The importance of instrumental and symbolic beliefsfor potential applicants, actual applicants, and mili-tary employees. Human Resource Management, 46:51–69.

Lievens, F., & Highhouse, S. 2003. The relation of in-strumental and symbolic attributes to a company’sattractiveness as an employer.Personnel Psychology,56: 75–102.

March, J. G., & Simon, H. A. 1958. Organizations. NewYork: Wiley.

McCracken, D. M. 2000. Winning the talent war forwomen: Sometimes it takes a revolution. HarvardBusiness Review, 78: 159–167.


McElroy, J. C., Morrow, P. C., & Rude, S. N. 2001. Turnoverand organizational performance: A comparativeanalysis of the effects of voluntary, involuntary, andreduction-in-force turnover. The Journal of AppliedPsychology, 86: 1294–1299.

Merton, R. K. 1968. Social theory and social structure.New York, NY: The Free Press.

Mitchell, T. R., Holtom, B. C., Lee, T. W., Sablynski, C. J., &Erez,M.2001.Whypeople stay:Using jobembeddednesstopredict voluntary turnover.AcademyofManagementJournal, 44: 1102–1121.

Nishii, L. H., Lepak, D. P., & Schneider, B. 2008. Employeeattributions of the “why” of HRpractices: Their effectson employee attitudes and behaviors, and customersatisfaction. Personnel Psychology, 61: 503–545.

Petty, R. E., & Cacioppo, J. T. 1986. Communication andpersuasion: Central and peripheral routes to atti-tude change. New York: Springer-Verlag.

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.

Piening, E. P., Baluch, A. M., & Salge, T. O. 2013. The re-lationship between employees’ perceptions of humanresource systems and organizational performance: Ex-amining mediating mechanisms and temporal dynam-ics. The Journal of Applied Psychology, 98: 926–947.

Rindova, V. P., Pollock, T., & Hayward, M. 2006. Celebrityfirms: The social construction of market popularity.Academy of Management Review, 31: 50–71.

Rindova, V. P., Williamson, I. O., Petkova, A., & Sever,J. M. 2005. Being good or being known: An empiricalexamination of the dimensions, antecedents, andconsequences of organizational reputation. Acad-emy of Management Journal, 48: 1033–1049.

Roehling, M. V., Boswell, W. R., Caligiuri, P., Feldman, D.,Graham,M.E.,Guthrie, J. P.,Morishima,M.,&Tansky,J. W. 2005. The future of HR management: Researchneeds and directions. Human Resource Manage-ment, 44: 207–216.

Roselius, T. 1971. Consumer rankings of risk reductionmethods. Journal of Marketing, 35: 56–61.

Rousseau, D. M. 1995. Psychological contracts in orga-nizations: Understanding written and unwrittenagreements. Thousand Oaks, CA: Sage.

Rucker, D. D., & Petty, R. E. 2006. Increasing the effectivenessof communications to consumers: Recommendationsbased on elaboration likelihood and attitude certaintyperspectives. Journal ofPublicPolicy&Marketing, 25:39–52.

Rynes, S. 1991. Recruitment, job choice, and post-hireconsequences: A call for new research directions.In M. Dunnette & L. Hough (Eds.), Handbook of

industrial and organizational psychology, vol. 2(2nd ed.): 399–444. Palo Alto, CA: ConsultingPsychologists.

Salancik, G. R., & Pfeffer, J. 1978. A social informationprocessing approach to job attitudes and task design.Administrative Science Quarterly, 23: 224–253.

Schwab, D. P., Rynes, S. L., & Aldag, R. J. 1987. Theoriesand research on job search and choice. Research inPersonnel and Human Resources Management, 5:129–166.

Shaw, J. D. 2011. Turnover rates and organizational per-formance: Review, critique, and research agenda.Organizational Psychology Review, 1: 187–213.

Spence, A. M. 1973. Job market signaling. The QuarterlyJournal of Economics, 87: 355–374.

Swider, B. W., Boswell, W. R., & Zimmerman, R. D. 2011.Examining the job search-turnover relationship: The roleof embeddedness, job satisfaction, and available alterna-tives. The Journal of Applied Psychology, 96: 432–441.

Tajfel, H., & Turner, J. 1979. An integrative theory of in-tergroup conflict. InW. G. Austin & S. Worchel (Eds.),The social psychology of intergroup relations:94–109. Monterey, CA: Brooks-Cole.

Tews, M. J., Michel, J. W., & Allen, D. G. 2014. Fun andfriends: The impact of workplace fun and constituentattachment on turnover in a hospitality context. Hu-man Relations, 67: 923–946.

Turban, D. B., & Cable, D. M. 2003. Firm reputation andapplicant pool characteristics. Journal of Organiza-tional Behavior, 24: 733–751.

Ulrich, D., Younger, J., Brockbank, W., & Ulrich, M. 2012.HR from the outside in. New York, NY: McGraw-Hill.

United States Census Bureau. 2015. Statistics of U.S.Businesses (SUSB) Main. http://www.census.gov/econ/susb/. Accessed on February 10, 2015.

Van Hoye, G., Bas, T., Cromheecke, S., & Lievens, F. 2012.The instrumental and symbolic dimensions of orga-nisations’ image as an employer: A large-scale fieldstudy on employer branding in Turkey. AppliedPsychology, 62: 543–557.

Van Hoye, G., & Lievens, F. 2009. Tapping the grapevine: Acloser look at word-of-mouth as a recruitment source.The Journal of Applied Psychology, 94: 341–352.

Weiner, B. 1986. An attributional theory of motivationand emotion. New York, NY: Springer-Verlag.

Williamson, I. O., Cable, D. M., & Aldrich, H. E. 2002.Smaller but not necessarily weaker: How small busi-nesses can overcome barriers to recruitment. In J. Katz& T. M. Welbourne (Eds.) Managing people in en-trepreneurial organizations: Learning from themerger of entrepreneurship and Human ResourceManagement: 83–106. Amsterdam: JAI Press.


http://www.census.gov/econ/susb/

http://www.census.gov/econ/susb/

Wright, P. M., Gardner, T. M., Moynihan, L. M., & Allen,M. R. 2005. The relationship between HR practicesand firm performance: Examining causal order. Per-sonnel Psychology, 58: 409–446.

Zorn, T. E., Roper, J., & Richardson, M. 2014. Positiveemployment practices or reputational capital?Tensions inherent in third-party legitimation pro-cesses. Management Communication Quarterly,28: 347–374.

Brian R. Dineen ([email protected]) is an AssociateProfessor of Management in the Krannert School of

Management at Purdue University. He received his PhDfrom The Ohio State University. His research interests in-clude recruitment and retention processes, employmentbranding, and job seeker behavior.

David G. Allen ([email protected]) is Professor in theSchool of Management & Labor Relations at Rutgers Uni-versity and Distinguished Research Environment Professorin the Warwick Business School at the University ofWarwick. He earned his PhD fromGeorgia State University.His research interests focuson the flowofhumancapital intoand out of organizations and global talent management.


mailto:[email protected]

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.

third party employment branding: human capital …...folio of methods to self- promote and sustain...

Documents