Hidden Badge of Honor: How Contextual Distinctiveness ......Title: Hidden Badge of Honor: How Contextual Distinctiveness Affects Category Promotion among Certified B Corporations Created

r Academy of Management Journal2017, Vol. 60, No. 6, 2294–2320.https://doi.org/10.5465/amj.2015.0416

HIDDEN BADGE OF HONOR: HOW CONTEXTUALDISTINCTIVENESS AFFECTS CATEGORY PROMOTION

AMONG CERTIFIED B CORPORATIONS

JOEL GEHMANUniversity of Alberta

MATTHEW GRIMESIndiana University

Why would an organization pursue membership in an organizational category, yetforego opportunities to subsequently promote that membership? Drawing on prior re-search, we develop a theoretical model that distinguishes between basic and sub-ordinate categories and highlights how organizations may differ in their promotion ofthe same subordinate category. We hypothesize that a subordinate category’s contextualdistinctiveness within different basic categories increases promotion, and that theseeffects are amplified in relatively larger subordinate category peer groups. To test ourhypotheses, we developed a proprietary web-based software toolset and gathered tex-tual and graphical data regarding B Corporations’ web-based promotion of their certi-fication. We supplemented our statistical analysis with interviews of Certified BCorporation entrepreneurs and executives. Our findings challenge prior assumptionsabout the causes of promotional forbearance, while extending our understanding ofcategory distinctiveness within contexts as well as sources of intra-category variation.

B Corp status has become a badge of honor.

– Entrepreneur magazine

I’mnot sure our clients know that we are a B Corp. It’sjust not something that we bring up.

– 2015 Interview with a B Corp executive

Why would a company obtain membership in anorganizational category—the “meaningful concep-tual systems” that grouporganizations on the basis of

shared attributes (Navis & Glynn, 2010: 440)—butthen refrain from promoting their association withthat category? Although organization theorists haveexamined the strategic value organizations derivefrom their categorymemberships (Granqvist, Grodal,&Woolley, 2012;Weber, Heinze, &DeSoucey, 2008),few scholars have examined category promotion,whichwedefine asmembers’ efforts to champion thelabels or cultural artifacts signifying the category.Instead, researchers have primarily focused on howorganizations “self-categorize” byway of claims thatseek to convey legitimate membership in a categoryand further solidify identity-based understandingsof “who we are” as an organization (Kennedy, 2008;Pontikes, 2012). Given evidence that members mayforego opportunities to promote such memberships,it would appear that the motivations for categorypromotion could differ from those that underpininitial membership claims. Although studies haveshown that membership claims can vary over timedue to category leniency (Pontikes & Barnett, 2015)and category legitimacy (Navis & Glynn, 2010),scholarshave largely ignored the important distinctionbetween membership claims and category promotion.

Third-party certifications provide a context fordisentangling the processes associated with mem-bership claims and subsequent category promotion

The authors contributed equally to this article.We thankJingbin (Ivy) Zhang for her research assistance. Previousversions of this researchwere presented at AOM (includedin Proceedings), ARCS (Peoples Choice Award winner),COSI, EGOS, ISA, SERC, and WAM (Past Presidents BestPaper Award nominee), as well as Alberta, UC-Irvine,Duke, Florida Atlantic, Indiana, NC State, Nottingham(China), and UT-Austin. We thank participants at each ofthese venues for their comments and suggestions, as wellas David Deephouse for his comments on the penultimatedraft. We also benefited tremendously from the commentsof Martine Haas and the anonymous reviewers. This workwas supported in part by grants from the Social Sciencesand Humanities Research Council, Office of the VicePresident of Research at the University of Alberta, Cana-dian Centre for Corporate Social Responsibility, AlbertaSchool of Business, and the Killam Research Fund.

2294

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.

https://doi.org/10.5465/amj.2015.0416

(King, Lenox, & Terlaak, 2005; Reinecke & Ansari,2015). When membership is sanctioned by a thirdparty, an actor’s efforts to promote the categorywould appear to indicate somethingmore thanmereself-categorization. Yet prior research on certifica-tions has similarly struggled to explain promotionalforbearance, wherein an organization voluntarilyrestrains from publicizing associations it is legiti-mately entitled to make. For instance, despitewidespread scholarly agreement that certificationscan offer important legitimacy or promotional ben-efits (Rao, 1994; Terlaak & King, 2006; Wade, Porac,Pollock, & Graffin, 2006), recent research has puz-zled over the fact that companies routinely abstainfrom promoting associations that they have gone togreat efforts to obtain (Carlos & Lewis, 2017; Delmas& Grant, 2014). In sum, in existing research on cer-tifications specifically and categories more broadly,scholars have focused on membership rather thanthe phenomenon of category promotion. As such,extant research offers little theoretical basis for un-derstanding promotional forbearance among cate-gory members.

How might we resolve this puzzle related to situ-ations in which membership and promotion di-verge? One starting point is to consider the differentpotential benefits offered by each. On one hand, ex-tant findings indicate that category membership en-ables organizations to more clearly define aspects oftheir identities (Glynn & Navis, 2013), and thus es-tablish similarities with other organizations (which,bydefinition, are alsomembers). In contrastwith this“fitting in” argument, we argue that the degree towhich categories offer a means for distinctiveness or“standing out,” especially vis-a-vis non-members,drives subsequent decisions regarding categorypromotion. Although overlooked as a basis for un-derstanding promotional forbearance, distinctive-ness lies at the heart of research on categories(Hannan, Polos, & Carroll, 2007; Negro, Hannan, &Rao, 2010). For example, researchers have examinedhow distinctiveness is particularly important withincategories, as organizations strive to differentiatetheir organizations from other category members(Paolella & Durand, 2016; Wry, Lounsbury, &Jennings, 2014). Others note the importance ofinter-category distinctiveness, suggesting that great-er affinity among members can contribute to in-creased contrast between categories (Negro et al.,2010). Yet despite these advances, few scholars haveconsidered how the need for distinctiveness mightaffect category members’ strategic actions (e.g., pro-motional forbearance).

We argue that to better understand a category’sdistinctiveness and its ability to affect members’ ac-tions requires further attention to the contextswithinwhich categories are embedded, ormore particularlythe relationships within category hierarchies. Ourarguments draw inspiration from earlier researchthat has differentiated “basic” categories from “su-perordinate” and “subordinate” categories (Hunn,1975; Rosch, Simpson, & Miller, 1976). Whereasbasic categories are cognitively “in the middle” ofa general-to-specific hierarchy (e.g., dog is the basiccategory in the animal-dog-retriever hierarchy),subordinate categories attempt to further delineatethe referenced object from the basic category(Markman & Wisniewski, 1997). For instance, al-though kitchen chairs share many features withchairs in general thereby challenging their distinc-tiveness, such chairs are characterized by a few at-tributes that are not shared, thereby offering apotential means for distinction (Lakoff, 1987; Roschet al., 1976). To the extent that this contrast betweensubordinate and basic categories is pronounced, thismay create an opportunity for promoting member-ship in that subordinate category.

Viewed in light of this research, we theorize thatmembers’ category promotion is a strategic responseto their need for distinctiveness within basic cate-gories (Deephouse, 1999; Jennings, Jennings, &Greenwood, 2009; Zhao, Fisher, Lounsbury, &Miller, 2017). Specifically, we introduce the con-cept of contextual distinctiveness, defined as thedegree to which a particular subordinate categoryoffers its members technical, material, and/or sym-bolic resources to distinguish themselves from or-ganizations that are not members of the subordinatecategory yet belong to the same basic category. Acategory’s contextual distinctiveness is high, for ex-ample, when the features, criteria, or practices thatdefine a subordinate category are not shared by oth-erwise similar organizations that comprise an orga-nization’s wider regional or industrial contexts.Specifically, we theorize that differences in pro-motional forbearance among subordinate categorymembers are attributable to differences in the sub-ordinate category’s contextual distinctiveness.Namely, we expect increased category promotionwhen the subordinate category provides organiza-tionswith ameansof differentiating themselves frombasic categorymembers. Additionally, we argue thatthese effects are amplified by membership size. Theexistence of more category peers helps increase thefamiliarity and credibility of the subordinate cate-gory within a given basic category, amplifying the

2017 2295Gehman and Grimes

effect of contextual distinctiveness on categorypromotion.

We draw our data from the emerging domain ofsocial entrepreneurship and corporate sustainability(Garud & Gehman, 2012; Grimes, McMullen, Vogus,& Miller, 2013; Wry & York, 2017). Since 2007, the BCorp certification has emerged as a popular way forcompanies to affirm their commitments to positiveenvironmental, social, and governance (ESG) prac-tices. At the time of our study, there were nearly1,700 Certified B Corporations globally. Given thesubstantial effort required to become a B Corp, itstands to reason that companies would seize everyopportunity to promote their certification. Yet priorto conducting this study, we noticed considerablevariation. Whereas many B Corps promoted theircertification, some made little or no mention of it.Because the B Corp certification provides a cross-industry, cross-geography sample of organizations,this setting offers an ideal context for studying howa subordinate category’s contextual distinctivenessmight affectmembers’promotional forbearance. Ouranalysis of promotional forbearance is based ona unique web-scraped dataset of B Corp websites thatenabled us to capture and analyze all text- and image-based promotional efforts from these organizationsacross some 650,000 web pages. To further validateour theorized causal model and associated mecha-nisms, we supplemented our statistical analysis byinterviewing B Corp entrepreneurs and executives.

Collectively, our findings offer several contribu-tions. First, we challenge and extend the currentunderstanding of promotional forbearance amongcategorymembers. Prior research has focused on therelationship between stigma and promotional for-bearance, furthering the assumption that onlymembers of less stigmatized or more celebrated cat-egories activelypromote theirmemberships (Durand& Vergne, 2015; Hudson & Okhuysen, 2009). Ourfindings challenge this assumption, suggesting thatpromotional forbearance is more common thanmight be expected, even among generally celebratedcategories. Notably, our concept of contextual dis-tinctiveness significantly explains this outcome.Second, scholars have highlighted the importanceof identity focus and member similarity in drivingmembership claims and category growth, yet ourstudy suggests that category promotion representsa fundamentally different process characterizedby members’ responses to a subordinate category’scontextual distinctiveness. By highlighting the im-portance of category hierarchies to this process,we extend prior conceptualizations of category

distinctiveness that have viewed suchdistinctivenessas simply the average typicality ofmembers’ features.Finally, researchers have argued that within-categorydifferences can be explained by members’ needs todistinguish themselves from other category membersor by general features of the category (e.g., boundaryleniency) (Navis & Glynn, 2011; Pontikes & Barnett,2015). Extending this research, our findings suggestthat intra-category variation can also result fromcontext-specific differences between basic and sub-ordinate category members.

THEORETICAL BACKGROUND

In this section, we discuss existing scholarship onorganizational categories.Whereas past research hasemphasized the role of membership claims in con-veying organizational similarities relative to othermembers, we highlight the potential role of categorypromotion in conveying distinctiveness relative toorganizations that belong to the broader context. Toaddress the theoretical puzzle of why organizationsmight choose to associate with a particular category,yet forego opportunities to promote it, we developa theoretical framework that highlights the impor-tance of category hierarchies for understanding theextent to which organizations promote their cate-gorical associations.

Disentangling Category Membership from CategoryPromotion

Management scholars have long studied organi-zational categories to better understand how orga-nizations relate to one another (see Porac, Thomas, &Baden-Fuller, 2011; Vergne & Wry, 2014 for re-views). Within this research stream, an organiza-tional category—the“meaningful conceptual systems”that group organizations on the basis of shared at-tributes (Navis & Glynn, 2010: 440)—is understoodas a socially constructed partitioning of organiza-tions based on a “mutual understanding of the ma-terial and symbolic resources that serve as a basis toassess membership” (Vergne & Wry, 2014: 68).Certification, for instance, serves as a salient ex-ample of a symbolic resource that denotes categorymembership.

Recently, Kennedy, Lo, and Lounsbury (2010)proposed that organizational categories vary in their“currency,” or the extent to which they have clearmeaning and positive appeal. They suggested thatorganizations are more likely to become members ofcategories in which both attributes are high. First, as

2296 DecemberAcademy of Management Journal

consensus about the meaning of a category label(what they called coherence) increases, the un-certainty associated with becoming a member of thecategory decreases, making the category more at-tractive to organizations. Second, as the degree ofpositive versus negative appeal of a category (whatthey called valence) increases, its currency in-creases, and with it the legitimacy associated withcategory membership. Membership in an organiza-tional category with high currency is important toorganizations in that it provides a basis for derivinglegitimate organizational identities and makinginter-organizational comparisons that guide com-petitive and cooperative strategies (Porac, Wade, &Pollock, 1999). Similarly, studies of certificationsemphasize the role of third parties in constructingand sanctioning the meaning and appeal of particu-lar categories (Sine, David, & Mitsuhashi, 2007).Some certifications become valuable to organiza-tions by establishing conformity to particular so-cietal standards or authenticating extraordinaryachievements (Grimes, Gehman, & Cao, 2017; Rao,1994; Rindova, Williamson, Petkova, & Sever,2005).

While researchers have considered variation inmembership claims over time as a function of thewaxing andwaning currencyof the category (Navis &Glynn, 2010) or the leniency of the category(Pontikes & Barnett, 2015), scholars have largelyoverlooked the potential for differences in the alter-native act of category promotion across differentmembers. Several studies have highlighted how dif-ferences in categories (Negro et al., 2010; Pontikes &Barnett, 2015) can lead to differences in categorymembers’ interactions with the category (e.g., joiningand then leaving); however, these studies did notsufficiently reveal why organizations would obtainmembership inacategoryandyet subsequently foregopromoting it. This puzzle is rendered more acute byevidence suggesting that category members may notonly differ in their level of category promotion, butmay in some cases forego category promotion alto-gether. For instance, a recent study of eco-labelingstrategies in the wine industry showed that whilemany wineries make significant efforts to achieveorganic certification, some wineries opt not to labeltheir products as such (Delmas & Grant, 2014).

Taken together, this accumulating evidence hasled scholars to realize that “opportunities abound toexplore the ways in which organizations strategi-cally signal their affiliation(s) within an existingcategory system” (Vergne & Wry, 2014: 78). Onepossibility, which we consider in the following

section, is that while membership claims serve as ameans of establishing similarities with categorypeers, category promotion alternatively serves ameans of signaling distinctiveness relative to orga-nizations from the broader context. Additionally,given that category members are likely embedded indifferent contexts, or what Durand and Paolella(2013) called situational circumstances, the “same”membership might offer organizations more or lessdistinctiveness, depending on contextual differences.

The Contextual Distinctiveness of SubordinateCategories

To understand how categories might offer varyingdegrees of distinctiveness in different settings firstrequires attention to the ways that categories canrelate to one another hierarchically. In other words,categories may not only overlap with other cate-gories, but they can also be embedded within andthus subordinate to other categories. Basic cate-gories, therefore, refer to the broad contexts withinwhich organizations are most frequently clusteredand labeled. Such basic categories are also charac-terizedby ahighdegree of “internal distinctiveness,”wherein the attributes that are perceived to charac-terize the category are likely perceived as presentamong all members of the category, but absentfrom non-members (Markman & Wisniewski, 1997).Distinctiveness is, therefore, often self-evident(e.g., dogs versus cats), rendering external efforts topromote suchdistinctionsunnecessary. Subordinatecategories, however, share features of the basic cat-egory yet are additionally characterized by a subsetof features that may distinguish to some degree thesubordinate member from other basic categorymembers (Lakoff, 1987; Rosch et al., 1976), for in-stance, attributes that distinguish golden retrieversfrom dogs in general.

Consequently, given similarities in terms of a basiccategory, promoting one’s membership in a sub-ordinate category would be an effective way ofestablishing distinctiveness within a basic category.Whereas membership in basic categories often pro-vides foundational answers to the question “Whoarewe?,” subordinate categories can additionally pro-vide organizations with a basis for answering thequestion “How are we different?” (Sahlin & Wedlin,2008). In this way, subordinate categories provideorganizations with the cultural resources forsignaling alterity (Czarniawska, 2008; Elsbach &Bhattacharya, 2001; Levinas, 1999) relative tobasic category members.


More specifically, we propose that category pro-motion is driven by a subordinate category’s con-textual distinctiveness, or the degree to whicha particular categorical association provides an or-ganizationwith technical, material, and/or symbolicresources for distinguishing itself from other mem-bers of the basic category.1 However, because sub-ordinate categories can at times span multiple basiccategories, the same subordinate category may offermore or less distinctiveness depending on context.Such is often the case with certifications. Forinstance, depending on its context, the ISO certifi-cation may offer a company the means for dis-tinguishing itself with regard tomaterial or technicalqualities and practices (Delmas & Montiel, 2008;King et al., 2005). In addition to offering a basis formaterial or technical distinctions, subordinate cate-gories might also offer organizations symbolic orcultural distinctiveness (Giorgi, Lockwood, &Glynn,2015; Glynn & Abzug, 2002). Given that values havelong been considered to constitute the core of anorganization’s identity (Albert & Whetten, 1985;Selznick, 1957), a subordinate category can thus of-fer its members critical ways for highlighting thedistinctiveness of those constitutive values and theassociated practices that enact those values(Gehman, Treviño, &Garud, 2013; Giorgi et al., 2015;Wright, Zammuto, & Liesch, 2017). Yet because thevalue practices that characterize a particular sub-ordinate category may factor prominently amongmembers of some basic categories and less so inothers (Lee & Lounsbury, 2015; Navis & Glynn,2011), the capacity to invoke a particular sub-ordinate categorical association as a point of differ-entiation is likely to vary.

Recent scholarship has reinvigorated attention tothe importance of the contexts in which organiza-tions operate (Garud,Gehman, &Giuliani, 2014; Low& Abrahamson, 1997). From a basic categories per-spective, physical (i.e., regions, cities) (Lounsbury,2007; Marquis, Davis, & Glynn, 2011) and virtual

(i.e., industries, fields,professions)contexts (Greenwood,Suddaby, &Hinings, 2002; Grimes et al., 2017;Micelotta& Raynard, 2011) are two important vectors that havebeen shown to influence organizational action. Sub-ordinate categorymembers thenmaydiffer on the extentto which they promote the category depending on thevalues practices salient among members of the broaderbasic categories—such as different industry sectors(e.g., coffee roasting vs. coal mining) and geographiccommunities (e.g., Vermont vs. Texas).When the valuespractices that might otherwise distinguish a particularsubordinatecategoryarealreadybroadlydiffusedamongthe non-members within an organization’s regionaland industrial contexts, the category’s contextualdistinctiveness is low. In this case, an organizationstands to gain little distinctiveness relative to theother basic category members by promoting its as-sociation with the subordinate category. For in-stance, distinguishing commercial farms fromorganic farms is only meaningful to the extentcommercial farms do not share the same practices(Weber et al., 2008). In sum, category promotionprovides a means by which organizations mightdistinguish themselves by region and industry;however, the extent of this distinction may varyacross these contexts. As such, we propose thatdifferences in a subordinate category’s contextualdistinctiveness with regard to the extant regionaland industrial practices among non-members canexplain why members engage in category pro-motion. This leads us to our first and secondhypotheses:

Hypothesis 1. The greater a subordinate cat-egory’s contextual distinctiveness within a par-ticular region, the greater the category promotionby members.

Hypothesis 2. The greater a subordinate cat-egory’s contextual distinctiveness within a par-ticular industry, the greater the categorypromotion by members.

Subordinate Category Peer Groups: The ContextualFamiliarity of Subordinate Categories

So far, we have argued that a subordinate cat-egory’s contextual distinctiveness within a basiccategory explains the extent to which memberspromote their associations with the category. Here,we further develop our understanding of a sub-ordinate category’s contextual distinctiveness bydrawing attention to the role of subordinate categorypeer groups in amplifying or attenuating an

1 In addition to offering resources for differentiationwithin a specific context, organizations may promote theirassociation with a subordinate category as a result of itsoverall legitimacy or currency across multiple contexts(Kennedy et al., 2010; Wry, Lounsbury, & Glynn, 2011).Although the need for contrast and collective legitimationrepresent different potential mechanisms, we expect thatboth are likely to work in concert with one another, en-couraging members’ promotion of the subordinate cate-gory. We investigate this possibility as part of ourrobustness checks.


organization’s responsiveness to these effects. Al-though drawing on subordinate category associa-tions to claim distinction within broader industrialand regional contexts may offer members certainstrategic advantages, this is likely conditioned by theextent to which such associations are contextuallyfamiliar or recognizable.2 Existing research suggeststhat categorical coherence, or consensus ofmeaning,depends partly on the number of category members(Kennedy, 2008; Kennedy et al., 2010). Whilescholars have argued that this is true for a category asa whole, here we argue that such consensus and theassociated familiarity is likely to be contextuallyderived. In other words, to fully understand the ex-tent to which a subordinate category might serve asa credible basis for distinctiveness we must addi-tionally consider the subordinate category’s size andconsequent familiarity within these respective con-texts. A larger subordinate category peer group isthus likely to amplify the effects of a category’scontextual distinctiveness on members’ categorypromotion by increasing the degree to which theassociated subordinate category is deemed credible.

Another mechanism whereby a larger peer groupcan amplify the effects of a subordinate category’scontextual distinctiveness on members’ categorypromotion is by increasing the extent to which thereis a strong collective identity that might compelpromotion (Wry, Lounsbury, & Glynn, 2011). Forinstance, as the collective presence of organizationsassociated with a subordinate category grows, itconstitutes a cultural resource that affords membersof that collective greater discretion to deviate fromother conventions or pressures of conformity thatmight be imposed by the wider basic categories(Navis & Glynn, 2010; Swidler, 1986; Wry et al.,2011). Although such a mechanism is no doubt im-portant in general, it is likely all the more potent inthe case of emerging subordinate categories, wherethe strategic value of such deviance is likely inquestion.Accordingly,we theorize that in a region orindustry with more subordinate category peers, anorganization’s sensitivity to any given level ofa subordinate category’s contextual distinctiveness

increases. In short, as subordinate category mem-bership increases, the effect of a subordinate cat-egory’s contextual distinctiveness on members’promotion is amplified. As such, we hypothesize:

Hypothesis 3. More subordinate category mem-berswithin a region amplify the positive effect ofa subordinate category’s contextual distinc-tiveness on category promotion.

Hypothesis 4. More subordinate category mem-bers within an industry amplify the positive ef-fect of a subordinate category’s contextualdistinctiveness on category promotion.

METHODS

B Lab introduced the Certified B Corporation ac-creditation in 2006 as part of its strategy of fosteringand promoting companies that use market-basedapproaches for addressing social and environmentalproblems (for a review, see Cao, Gehman, & Grimes,2017). As of 2015, over 16,000 businesses had beenassessed and over 1,700 from 42 countries had beencertified.3 This context offers several noteworthyadvantages for addressing the proposed researchquestion. First, the B Corp certification provides anideal setting for disentangling category membershipfrom category promotion. By setting evaluativethresholds and accrediting particular organizationsand not others, B Lab acts as a powerful categorizingagent, granting category membership to organiza-tions that uphold the category’s standards (Grimes,2010). As such, the extent to which a member pro-motes the category can be observed as a means ofdifferentiation from non-certified basic categorymembers. Second, given the effort and positive

2 A category’s familiarity is related to yet different fromits saliency, a concept Verne andWry (2014: 73) defined as“how much attention audiences devote to a particularcategory within the broader classification hierarchy.” Al-though a category’s salience is likely sufficient for pre-dicting an increase in its familiarity, it is not a necessarycause in that other factors might similarly increasefamiliarity.

3 Companies interested in becoming a Certified B Cor-poration start by taking the B Impact Assessment, whichrates their sustainability using environmental, social, andgovernance criteria. To be eligible for certification, a com-pany must score at least 80 out of 200 possible points inthese areas. Documentation is required for B Lab’s reviewand organizations are subject to random audits (10 percentof companies reportedly are visited annually). Certifica-tion is valid for 2 years, at which point a companymust bere-certified. Companies legally organized as corporationsor LLCs (as opposed to sole proprietorships) must amendtheir governing documents “to take into consideration theinterests of all stakeholders, not just shareholders.” Fi-nally, companies must sign the B Corp “Declaration of In-terdependence” and a corresponding legal term sheet, andpay an annual fee ranging from $500 to $50,000 or more,depending on annual sales.


regard associated with becoming a member of thecategory, this setting enables us to rule out the pos-sibility that promotional forbearance is merely in-dicative of a stigmatized category (cf. Vergne, 2012).Third, the B Corp certification offers an example ofa subordinate category that can be found in a host ofdifferent industries and geographies, enabling us toexamine how differences across basic categoriesshape organizations’ promotion of a subordinatecategory. Finally, each of the companies main-tained websites geared toward promoting them-selves to their various stakeholders, providing uswith an extremely conservative setting for testingour ideas. Specifically, promoting the B Corp cer-tification online is virtually costless, adding to thepuzzle of why a firm would forego promoting thiscategory.

Sample Construction

Our sample for this study is comprised of the 526BCorps located in theUnited States that were certifiedas of December 2013 and remained certified as ofNovember 2015. This sample enabled us to examinehow conditions at an earlier date affect members’category promotion at a later date. Using Amazon’sMechanical Turk service (MTurk), we had threeseparate raters provide us with the street addressesand phone numbers for each B Corp. When ratersdisagreed, the co-authors independently verifieda company’s location. Using this information, wethen appended data from Dun and Bradstreet re-garding each company’s primary Standard In-dustrial Classification (SIC) code, annual revenue,and number of employees. Of the 526 companies, wewere able tomatch 484 (or 92%)of our sample toDunand Bradstreet records.

Dependent Variables

As scholars have recently highlighted, “most textsencountered in and around organizations are multi-modal” (Meyer, Jancsary, Hollerer, & Boxenbaum,2017: 8), encompassing both verbal and visual ele-ments. However, “despite a prominent line of re-search that addresses discourse (for an overview, seePhillips & Oswick, 2012), the visual mode of mean-ing construction has remained largely unexplored inorganization and management research” (Meyer,Hollerer, Jancsary, & van Leeuwen, 2013: 490). Thisgap is particularly salient within the context ofstudies on categories, given their emphasis on theconstruction and diffusion of meaning. With the rise

of organizational websites and social media, orga-nizations make prolific use of visual artifacts, andsuch artifacts have been shown to be powerful intheir capacity both to influence meaning construc-tion and diffusion, and to persuade audiences.

Accordingly, we designed our dependent vari-ables to capture differences in category promotion—the extent to which a B Corporation promoted (ornot) its certified status—both textually and graphi-cally. We calculated category text promotion asa count of the number of B Corp terms that appearedwithin the first three “levels” of a company’swebsite(i.e., any page on or within 2 clicks of the homepage).4 To develop the dictionary of terms, we man-ually reviewed the B Lab website. The resultingdictionary includes references to the B Corporationcertification aswell as relateddistinguishingphrasessuch as “The Change We Seek.” Similarly, we cal-culated category image promotion as a count of thenumber of B Corp images that appeared within thefirst three levels of a company’s website. We createdthe library of images bymanually reviewing all pageson the B Lab website. These images included theCertified B Corporation logo, B the Change logo, andDeclaration of Interdependence image (see Figure 1for the dictionary and images).

We collected the data necessary to construct thesevariables by developingCULTR, a suite ofweb-basedapplications specially designed to crawl websitesand capture information about category promotion(http://www.cultrtoolkit.com/). TheCULTR text andimage scrapers started on the internet home page ofeach company in our sample and then followed allinternal links, ultimately traversing more than650,000 B Corp web pages. These scrapers capturedand reported each instance of the text and images ona given page, along with metadata such as the sizeand location of these instances. The results reportedare based on category promotion data collected inlateOctober andearlyNovember 2015, atwhich time

4 During a pilot study, we evaluated multiple construc-tions of our dependent variables, ranging from using onlywebsite homepages to using all pages on awebsite.Wealsoused MTurk to manually collect data regarding categorypromotion for comparison andvalidationpurposes.On theone hand,we found that being too restrictive, such as usingthe home page only, resulted in false negatives; weundercounted the true extent of category promotion. Onthe other hand, we found that being overly broad, such asusing the entire website, produced artificially high countsin a few cases. For this project, using the first three levelsprovided the best balance between these two extremes.


http://www.cultrtoolkit.com/

we succeeded in gathering data for 507 of the 526(96%) websites in our sample.5 Figure 2 is a screen-shot of a B Corp showing the one image and nine textpromotions captured by our measures for this par-ticular page.

Independent Variables

First, to assess Hypotheses 1 and 2 regarding con-textual distinctiveness by region and industry, wesought a measure capable of indicating the extent towhich a member’s promotion of the B Corp certifi-cation would distinguish it from the non-B Corps inthese same two contexts. Because companies qualifyfor B Corp certification based on their ESG practices,we looked for a way to measure the extent to whichnon-B Corps were committed to or deviated fromcomparable ESG practices. These considerations ledus to collect data from MSCI ESG KLD STATS(hereafter MSCI STATS), “an annual data set ofpositive and negative environmental, social, andgovernance performance indicators applied toa universe of publicly traded companies” (MSCI,2015: 10).6 Specifically, we used MSCI STATS for2012, providing us with meaningful temporal sepa-ration between this key independent variable andour dependent variables. These data provided us

withESGratingson theMSCIUSAInvestableMarketIndex. With approximately 2,500 constituents, thisindex covers 99% of the US stock market, includinglarge, mid, and small cap segments. Given the sa-lience of publicly traded firms, a measure that ac-counts for the ESG-related practices of these firms byregion and industry provides an ideal means forassessing the contextual distinctiveness afforded bypromoting the B Corp category.

Using these data, we calculated regional contex-tual distinctiveness as the mean number of ESGconcerns among publicly traded companies withina focal organization’s state.7 For instance, our mea-sure of regional contextual distinctiveness forPennsylvaniawas based on ratings of 109 companiesheadquartered in the state, ranging from Air Prod-ucts in Allentown to Urban Outfitters in Phila-delphia. Similarly, we calculated industry contextualdistinctiveness as the mean number of ESG concernsamong publicly traded companies within a focal or-ganization’s industry, as defined by its SIC division,a widely accepted classification of 10 industrialgroups (e.g., https://www.osha.gov/pls/imis/sic_manual.html). Examples of ESG concerns includetoxic emissions and waste, which assesses the sever-ity of controversies related to a firm’s non-greenhousegas (GHG) emissions; product quality and safety,which assesses the severity of controversies related tothe quality and safety of a firm’s products and ser-vices; workforce diversity, which assesses the sever-ity of controversies related to a firm’s workforcediversity; and controversial investments, which as-sesses the severity of controversies related to the so-cial and environmental impact of a firm’s lending,underwriting, and financing activities (MSCI, 2015).In addition to the theoretical appropriateness of ourmeasures, it has been demonstrated in prior re-search that the MSCI STATS concerns data(i.e., negative indicators) have greater historicalaccuracy and better predictive validity than thestrengths data (i.e., positive indicators) (Chatterji,Levine, & Toffel, 2009; Mattingly & Berman, 2006).

FIGURE 1B Corporation Dictionary and images

B Corp* Certified B Corp* B Lab benefit corp* B Impact Assessment B Impact Report Declaration of Interdependence Global Impact Investing Rating System GIIRS The Change We Seek

* denotes a wildcard search character

5 Errors typically resulted whenwebsites did not permitaccess to our web crawler or the website design made textinaccessible (e.g., Adobe Flash-based websites).

6 Over the past 25 years, these data have become themost widely used and critically evaluated source of sus-tainability ratings among management scholars (for as-sessments, see Chatterji, Levine, & Toffel, 2009; Delmas,Etzion, & Nairn-Birch, 2013; Hart & Sharfman, 2015;Mattingly & Berman, 2006).

7 We chose to delimit a B Corp’s region by its state forseveral reasons. First, laws regarding incorporationvarybystate, suggesting this is likely an important institutionalboundary for B Corps when making inter-organizationalcomparisons. Second, state is a common unit of analysis,with sources asdiverse as theFortune500annual rankings,the US Bureau of Labor Statistics andmanagement studies(e.g., Ruef&Patterson, 2009) all aggregating andcomparingdata at the state level. We test the robustness of this choiceusing a narrower definition of region.


https://www.osha.gov/pls/imis/sic_manual.html

https://www.osha.gov/pls/imis/sic_manual.html

In sum, these variables enabled us to assess con-textual distinctiveness between B Corps and non-BCorps by measuring the difference between the av-erage ESG concerns of public companies and theessentially negligible ESG concerns of the pro-totypical B Corp member within particular regionaland industry contexts.

Second, to assessHypotheses 3 and4 regarding theinteractive effects of a category’s contextual famil-iarity on category promotion, we constructed mea-sures of B Corp peer membership by region andindustry. We measured regional category peers asa count of the number of Certified B Corporationswithin a focal organization’s state. Similarly, wemeasured industry category peers as a count of Cer-tified B Corporations within a focal organization’sindustry. Data for these measures were derived from

the Certified B Corporation directory, together withaddress information gathered using MTurk and SICcodes gathered from Dun and Bradstreet. Both vari-ables were observed as of December 2013, providingclear temporal separation from our dependent vari-ables. We mean centered all four independent vari-ables to facilitate interpretation of our interactionterms (Echambadi & Hess, 2007).

Wealso evaluatedwhether these variablesprovideevidence of discriminant validity (i.e., whethermeasures that are supposed to beunrelated are in factunrelated).8 Ourmeasures of regional category peersand regional contextual distinctiveness are corre-lated at 2.01, while our measures of industry

FIGURE 2Example of B Corp Category Promotion coding

Our web crawler searched the text and image filenames of Certified B Corporation websites for thefollowing terms: B Corp*; Certified B corp*; B Lab; benefit corp*; B Impact Assessment;B Impact Report; Declaration of Interdependence; Global Impact Investing Rating System; GIIRS;The Change We Seek. * is a wildcard search character.

8 We thank an anonymous reviewer for suggesting thisevaluation.


category peers and industry contextual distinctive-ness are correlated at 2.25. Additionally, the mea-sures are derived from entirely different kinds ofdata. Taken together, the evidence indicates that ourkey variables are conceptually and empiricallydiscriminant.

Control Variables

To help rule out alternative explanations of cate-gory promotion, we controlled for a number of otherfactors. To control for the effects of company size oncategory promotion, we used data from Dun andBradstreet to measure annual sales as the natural logof sales volume in US dollars, and employees asa company’s total number of employees. To controlfor the possibility that company age affects categorypromotion, we calculated how old a company was,in years, based on its founding date. These data cameprimarily from Dun and Bradstreet, and were sup-plementedwith data from the B Labwebsite,MTurk,and internet searches. To control for differences inopportunities to promote, we used data derived fromour web-based crawler to calculate website words(website images) as the natural log of the total num-ber of words (images) within the first three levels ofa company’s website, thereby controlling for differ-ences in website size and complexity. To control forthepossibility that somecompanies are simply betterat internet marketing than others, we calculated so-cialmedia presence as a variable ranging from 0 to 3,depending on whether a company had active ac-counts on Twitter, Facebook, or LinkedIn. Thismeasure is based on data collected from the B Labwebsite and supplemented by data from MTurk andinternet searches.

Evidence from prior research suggests that cate-gory members can vary in their prototypicality, andthat such differences can affect how members relateto the category (Jones, Maoret, Massa, & Svejenova,2011; Rindova, Petkova, & Kotha, 2007). We controlfor this possibility in three different ways. First, wemeasured category founder as a binary variablecoded as 1 for any companies that were listed asfoundingmembers on the B Labwebsite. Second, wemeasured best in class as a binary variable coded as 1for any companies that obtained B Impact Assess-ment scores in the top 10% in any year until 2013.Third, we measured born B as a binary variablecoded as 1 for any company that was certifiedwithin1 year of its founding. Additionally, to control forinconsistent temporal separation between some in-dependent and dependent variables, we measured

months certified as the number of months elapsedbetween a company’s certification and the date wecollected the data for our dependent variable(i.e., late October to early November 2015).

To control for differences in the presence of publiccompanies by region and industry, we used datafrom MSCI STATS to measure regional public com-panies and industry public companies as a count ofthe number of publicly traded companies withinthese respective contexts.9 Because the correlationbetween regional public companies and regionalcategory peers was .96, we partialed out their com-mon variance by regressing regional category peerson regional public companies and then used the re-siduals in our analysis (Cohen, Cohen, West, &Aiken, 2003; Pollock & Gulati, 2007). Doing so re-duced the correlation between these variables to .17,and enabled us to control for any influence regionalpublic companies had on category promotion thatwas unrelated to regional category peers. We testedthe robustness of this choice in two ways: by ex-cluding the regional public companies controlvariable completely, and by using the original un-transformed control variable instead of the re-siduals. In all cases, we found the same pattern ofresults reported below. To control for broader re-gional and industry differences that might influ-ence category promotion, we coded regionaldummy variables based on US Census regions:midwest-mountain region, south region, and westregion (leaving northeast as the omitted referencecategory); and a services industriesdummyvariablefor any companies with SIC codes 7000–9999. Intwo cases, these variables were strongly correlatedwith other measures: west region and regional cat-egory peers (.74), and services industries and in-dustry category peers (.88). These correlations arenot surprising: approximately 40% of the sample islocated in the west region, and approximately 51%of the sample is categorized in services industries.To determine whether these correlations influ-enced our results, we tested models without theregion and industry control variables and found thesame pattern of results reported below.

Model Estimation

Our dependent variable is a discrete count vari-able. Not only are these data skewed left with longright tails, they also contain excess zeroes, meaning

9 We thank an anonymous reviewer for proposing thesecontrol variables.


there are more zero counts than would be predicted(Cameron & Trivedi, 2013). Hurdle models are nowconsidered one of the foremost methods for dealingwith such data (Cameron & Trivedi, 2013; Hilbe,2014).10 Thismethod assumes that data reflect a two-part process in which positive counts are generatedonly after first crossing a hurdle. Typically, hurdlemodels partition the data into a zero component,generating a binary response (0 or 1), and a countcomponent, generating positive counts (1 andhigher). These models have been used to researchdiverse phenomena ranging from website surfingbehavior (Bucklin & Sismeiro, 2003) to commercialfertilizer demand (Ricker-Gilbert, Jayne, & Chirwa,2011). In short, hurdle models are the preferred ap-proach for dealing with excess zeroes when only the“at risk” population can generate zeroes.

In addition to meeting these general parameters,two other features of our context suggest the appro-priateness of hurdle models. First, because the BCorp accreditation is actively championed by B Lab,hurdlemodels enable us to control for the possibility(without imposing a requirement) that the initialbinary decision regarding whether to promote isdriven by a different process than the subsequentstrategic decision regarding how much to promote.Second, because websites can vary dramatically insize, hurdle models enable us to control for the pos-sibility (again, without imposing a requirement) thatthe opportunity to promote might differentially af-fect these decision processes. Thus, we model cate-gory promotion as a two-part process in which thezero and count generating processes are not con-strained to be the same. In the first part, we modelzeroes using a binomialmodelwith a logit link. In the

secondpart,we test ourhypothesesusing a truncatedPoisson model with a log link. We estimated all pa-rameters based on maximum likelihood using thehurdle command in the R statistical package.

RESULTS

Hurdle Models of Category Text Promotion

In Table 1, we report descriptive statistics andcorrelations for the dependent, independent, andcontrol variables used in this study. In Table 2, wepresent the results of the models that test our hy-potheses related to category text promotion. Model 1is the zero model. It is a binomial regression pre-dicting whether a B Corp promotes its category af-filiation using the control variables, together withour independent variables.

Model 2 is provided for diagnostic purposes. Itsspecification is identical to Model 1, except it is es-timated as a zero-truncated Poisson regression. Itshows that whereas the variables underlying ourhypotheses are non-significant predictors of the bi-nary promotion decision (Model 1), they are signifi-cant predictors of the extent of promotion (Model 2).This result supports our choice to model categorypromotion as a two-part process.

Model 3 is the baseline control model. In the firstpart of the regression, we include the zero modelexactly as described in Model 1. In the second part,we enter the complete array of control variables.They are all highly significant. Annual sales, em-ployees, website words, social media presence, cat-egory founder, best in class, services industries,regional public companies, and industry publiccompanies significantly increase category text pro-motion. By comparison, company age, born B,months certified, midwest-mountain region, southregion, and west region significantly decrease cate-gory text promotion.

In Model 4, we add the regional contextual dis-tinctiveness variable, which is positive and signifi-cant (p , .001), providing support for Hypothesis 1.In Model 5, we add the industry contextual distinc-tiveness variable, which is positive and significant(p , .001), providing support for Hypothesis 2.Model 6 includes both contextual distinctivenessvariables simultaneously; they are positive and sig-nificant, providing further support for Hypotheses 1and 2. As the category’s contextual distinctivenessincreases by region and industry, category pro-motion increases. Stated vernacularly, B Corps em-bedded in “dirtier” regional and industrial contextsincrease their category promotion.

10 Zero-inflated models are the other alternative formodeling count data with excess zeroes (Cameron &Trivedi, 2013; Hilbe, 2014). Whereas hurdle models aretwo-part models, zero-inflated models are finite mixturemodels in which two data-generating mechanisms are as-sumed: one generating only zeroes (so-called bad zeroes),and the other generating the full range of counts (includingso-called good zeroes). “To appropriately employ a zero-inflatedmodel on datawith excess zero counts, the analystshould have a theory as to why there are a class of obser-vations having both observed and expected zero counts”(Hilbe 2014: 196). For instance, when counting birdcalls,zero counts can occur because birds were quiet during theobservation period (“good” zeroes), and because the ana-lyst failed to observe birdcalls that actually occurred(“bad” zeroes). In our case, there are not two zero-generating processes; because we scraped each B Corp’sentire website, no promotions (birdcalls) went unobserved.


TABLE1

SummaryStatistics

Variable

Mea

nSD

min

max

12

34

56

78

910

1112

1314

1516

1718

1920

2122

1Categ

orytext

promotion

154.97

460.28

04,92

92

Categ

oryim

agepromotion

137.69

990.07

015

,871

.53

3Reg

ional

contextual

distinctiven

ess

0.00

0.20

20.75

0.88

.07

.06

4Industry

contextual

distinctiven

ess

0.00

0.44

20.40

1.64

.08

–.03

.05

5Reg

ional

catego

rypee

rs0.00

77.31

281

.52

115.48

–.01

–.02

–.01

.00

6Industry

catego

rypee

rs0.00

114.89

214

8.08

125.92

–.08

–.02

–.03

–.25

–.01

7Annual

sales(log

)12

.07

4.68

22.30

18.52

.00

.06

–.05

–.64

–.08

.18

8Employe

es29

.25

100.14

012

00.06

.04

.05

–.05

–.01

–.07

.14

9Categ

oryfounder

0.09

01

.10

.05

.01

–.07

.07

–.01

–.03

.02

10Bestinclass

0.32

01

.04

–.04

.01

.03

–.02

–.05

.02

.03

.04

11W

ebsite

words(log

)9.84

3.08

22.30

16.69

.27

.16

.03

–.05

.00

–.08

.09

.08

.05

–.01

12W

ebsite

imag

es(log

)6.53

3.12

22.30

14.12

.18

.23

.03

–.11

–.05

–.11

.15

.10

.00

–.02

.67

13Soc

ialm

edia

presence

2.32

0.77

03

.08

.04

.03

–.01

.04

.00

.03

.06

–.06

.03

.15

.19

14Mon

thsce

rtified

31.03

21.89

180

.03

.03

–.04

–.04

.09

.00

–.02

.00

.56

.08

.02

.01

.04

15Com

pan

yag

e12

.27

16.41

122

4.04

.02

–.02

–.10

–.01

–.03

.17

.28

.13

.03

.10

.12

–.04

.12

16BornB

0.27

01

–.05

.01

–.02

.10

–.01

–.02

–.14

–.12

.06

.04

–.16

–.18

.08

.20

–.32

17Reg

ional

public

compan

ies(residual)

20.22

1.03

23.30

2.37

–.05

.06

.20

–.02

.17

.07

.00

–.08

.03

–.01

–.02

–.06

–.01

.00

–.15

–.01

18Industry

public

compan

ies

352.00

261.59

492

0.03

.00

.04

–.35

.05

.03

.23

.14

–.11

–.01

.11

.13

.02

–.07

.12

–.14

–.09

19Sou

thregion

0.16

01

–.04

–.04

.15

–.02

–.37

.05

.03

–.05

–.08

.03

.02

–.02

.12

–.04

–.08

.08

.07

–.02

20Midwest-mou

ntain

region

0.15

01

.02

–.01

.28

.05

–.35

–.03

.02

.03

–.09

.06

.03

.04

–.03

–.11

–.01

.02

–.11

.00

–.19

21W

estreg

ion

0.40

01

–.04

–.08

–.30

–.01

.74

–.03

–.07

–.02

.02

–.03

.01

–.02

.05

.08

.00

.02

–.22

.03

–.37

–.35

22Servicesindustries

0.51

01

–.05

–.03

.00

.21

–.01

.88

–.16

–.12

–.03

–.03

–.13

–.19

.00

–.02

–.08

.03

.07

–.20

.05

–.01

–.03


TABLE2

Hurd

leMod

elsof

Categ

oryTex

tPromotion

Mod

el1

Mod

el2

Mod

el3

Mod

el4

Mod

el5

Mod

el6

Mod

el7

Mod

el8

Mod

el9

Con

stan

t23.32

321.29

5***

20.69

7***

20.62

7***

21.06

4***

21.00

2***

21.42

0***

20.03

320.49

3***

(1.781

)(0.050

)(0.036

)(0.036

)(0.038

)(0.037

)(0.040

)(0.057

)(0.059

)Sou

thregion

21.19

2*20.81

9***

20.82

0***

20.78

3***

20.78

0***

20.73

3***

20.68

7***

20.62

7***

20.59

6***

(0.604

)(0.016

)(0.015

)(0.015

)(0.015

)(0.015

)(0.016

)(0.015

)(0.016

)Midwest-mou

ntain

region

0.71

620.31

7***

20.14

9***

20.20

1***

20.19

1***

20.22

9***

20.23

7***

20.15

0***

20.16

6***

(0.685

)(0.014

)(0.013

)(0.013

)(0.013

)(0.013

)(0.014

)(0.013

)(0.014

)W

estreg

ion

0.25

920.14

6***

20.57

5***

20.44

1***

20.57

4***

20.42

4***

0.28

7***

20.37

6***

0.30

1***

(0.733

)(0.016

)(0.011

)(0.011

)(0.011

)(0.011

)(0.021

)(0.012

)(0.021

)Servicesindustries

0.46

60.20

0***

0.19

8***

0.15

5***

0.11

4***

0.06

7***

0.09

9***

20.50

8***

20.42

5***

(1.825

)(0.052

)(0.009

)(0.009

)(0.009

)(0.009

)(0.009

)(0.059

)(0.059

)Employe

es0.01

60.00

0***

0.00

0***

0.00

00.00

0***

0.00

0***

0.00

0***

0.00

0***

0.00

0*(0.010

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)Bestinclass

0.12

40.11

8***

0.20

1***

0.21

3***

0.11

2***

0.11

8***

0.12

5***

0.12

5***

0.13

2***

(0.358

)(0.008

)(0.008

)(0.008

)(0.008

)(0.008

)(0.008

)(0.008

)(0.008

)Soc

ialm

edia

presence

0.34

20.17

1***

0.19

7***

0.17

8***

0.18

8***

0.17

1***

0.17

2***

0.16

1***

0.16

2***

(0.197

)(0.006

)(0.006

)(0.006

)(0.006

)(0.006

)(0.006

)(0.006

)(0.006

)W

ebsite

words(log

)0.49

4***

0.47

4***

0.48

7***

0.48

5***

0.46

9***

0.46

6***

0.47

1***

0.44

8***

0.45

4***

(0.074

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)Annual

sales(log

)20.06

10.06

0***

0.00

9***

0.00

8***

0.05

9***

0.05

8***

0.05

7***

0.06

5***

0.06

4***

(0.058

)(0.001

)(0.001

)(0.001

)(0.001

)(0.001

)(0.001

)(0.001

)(0.001

)Categ

oryfounder

0.47

30.34

9***

0.55

0***

0.47

7***

0.43

9***

0.36

6***

0.39

3***

0.30

7***

0.34

1***

(0.769

)(0.016

)(0.015

)(0.016

)(0.016

)(0.016

)(0.016

)(0.016

)(0.016

)Mon

thsce

rtified

0.00

320.00

6***

20.00

8***

20.00

8***

20.00

7***

20.00

6***

20.00

6***

20.00

5***

20.00

5***

(0.009

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)Com

pan

yag

e20.01

720.00

7***

20.00

6***

20.00

6***

20.00

6***

20.00

6***

20.00

6***

20.00

6***

20.00

6***

(0.013

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)BornB

20.09

20.02

9**

20.13

6***

20.10

7***

20.05

9***

20.03

7***

20.03

9***

20.02

8**

20.02

9**

(0.394

)(0.010

)(0.010

)(0.010

)(0.010

)(0.010

)(0.010

)(0.010

)(0.010

)Reg

ional

public

compan

ies(residual)

0.14

50.06

1***

0.16

2***

0.09

0***

0.14

9***

0.07

0***

0.00

30.01

3*20.04

7***

(0.240

)(0.005

)(0.005

)(0.006

)(0.005

)(0.006

)(0.006

)(0.006

)(0.006

)Industry

public

compan

ies

0.04

10.04

5***

0.05

5***

0.04

5***

0.05

2***

0.04

2***

0.04

7***

0.02

0***

0.02

5***

(0.081

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)Reg

ional

catego

rypee

rs20.00

220.00

2***

20.00

4***

20.00

4***

(0.004

)(0.000

)(0.000

)(0.000

)Industry

catego

rypee

rs0.00

10.00

0*0.00

6***

0.00

6***

(0.008

)(0.000

)(0.000

)(0.000

)H1:

Reg

ional

contextual

distinctiven

ess

21.59

81.14

8***

0.89

4***

0.96

9***

4.19

7***

1.02

4***

4.06

8***

(1.175

)(0.024

)(0.023

)(0.024

)(0.104

)(0.024

)(0.105

)H2:

Industry

contextual

distinctiven

ess

21.05

40.75

2***

0.78

3***

0.79

4***

0.80

0***

6.29

8***

6.03

7***

(0.900

)(0.028

)(0.013

)(0.013

)(0.013

)(0.160

)(0.160

)H3:

Reg

ional

contextual

distinctiven

ess

0.04

1***

0.03

9***

xregion

alca

tego

rypee

rs(0.001

)(0.001

)H4:

Industry

contextual

distinctiven

ess

0.04

4***

0.04

2***

xindustry

catego

rypee

rs(0.001

)(0.001

)Zeroco

untm

odel

N/A

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

AIC

312.4

1229

43.5

1292

06.7

1276

73.5

1252

54.2

1235

50.2

1220

23.6

1221

15.8

1207

24.2

Log

like

lihoo

d213

6.2

261

431.8

264

567.3

263

799.8

262

590.1

261

737.1

260

971.8

261

017.9

260

320.1

Notes:n

546

5;stan

darderrors

inparen

theses.

*p,

0.05

**p,

0.01

***p,

0.00

1


In Model 7, we add the interaction term regionalcontextual distinctiveness x regional category peers,which is positive and significant (p, .001). InModel8, we add the interaction term industry contextualdistinctiveness x industry category peers, which alsois positive and significant (p, .001). Given the non-linearity of our models, in Figure 3 we have plotted

the effects of these interactions graphically. Model 9is a fully saturated model; both interaction terms arepositive and significant. In addition to evaluating thesignificance of the interaction terms, we comparedthe Akaike Information Criterion (AIC) in Models 6and 7, Models 6 and 8, and Models 6 and 9. In allthree cases, the improvements in model fit are

FIGURE 3Interaction Plots

250

600

400

200

0–60 –30 0 30 60

200

150

100

50

–50 0 50

Region Category Peers

Pre

dic

ted

Cat

egor

y P

rom

otio

nP

red

icte

d C

ateg

ory

Pro

mot

ion

Industry Category Peers

Interaction Plot: (Region Category Peers) × (Region Contextual Distinctiveness)

Interaction Plot: (Industry Category Peers) × (Industry Contextual Distinctiveness)

+ SD Region Cont. Distinctiveness

+ SD Ind. Cont. Distinctiveness

– SD Ind. Cont. Distinctiveness

– SD Region Cont. Distinctiveness


considered “very strong” (Raftery, 1995). Addition-ally, because our models are estimated using iden-tical data and nested within each other, they can beformally compared using the deviance statistic(Singer & Willett, 2003). In all three pairs of models,the differences in deviance statistics far exceed thep, 0.001 critical values of a x2 distribution on 2 and4 d.f., respectively. Taken together, these resultsprovide strong support for Hypotheses 3 and 4. Anincrease in regional and industrial category peersamplifies the effect of contextual distinctiveness oncategory text promotion.

Hurdle Models of Category Image Promotion

Table 3 presents the results of the models that testour hypotheses on category image promotion. Thevariables in this table directly mirror those alreadydiscussed, with two exceptions. The dependentvariable is category image promotion, enabling us totest our hypotheses using a completely differentmeasure of category promotion. In parallel with thischange in dependent variable, we now control forwebsite images.

Model 10 is the zero model and Model 11 is thediagnostic count model. Model 12 is the baselinecount model. In Model 13, we add the regional con-textual distinctiveness variable, which is positiveand significant (p , .001). In Model 14, we add theindustry contextual distinctiveness variable, whichis positive and significant (p , .001). In Model 15,both variables are again positive and significant(p , .001). These results provide strong support forHypotheses 1 and 2.

In Model 16, we add the interaction term regionalcontextual distinctiveness x regional category peers,which is positive and significant (p, .001). InModel17, we add the interaction term industry contextualdistinctiveness x industry category peers, which alsois positive and significant (p , .001). Model 18 isa fully saturated model with both interaction terms.As with the category text promotion models, we alsoevaluated improvements in model fit using the AICand by calculating differences in deviance statistics.In all cases, the results provide strong support forHypotheses 3 and 4. An increase in regional and in-dustrial category peers amplifies the effect of contex-tual distinctiveness on category image promotion.

Robustness Checks

We also performed a variety of robustness checks,which are described below but not otherwise

reported. First, we constructed two alternative de-pendent variables. We measured category page pro-motions as the number of pages within the first threelevels of a company’s website that contained any BCorp terms, and controlled for website pages, cal-culated as the natural log of the total number of pageswithin a website’s first three levels. We measuredcategory pixel promotions as the pixel area (i.e.,graphical area) within the first three levels of a com-pany’s website that contained any B Corp images,and controlled for website pixels, calculated as thenatural log of the total pixel area devoted to images ofany kind within the website’s first three levels. Wecollected data for these measures using the samescraping procedures described earlier. Using thesenew variables, we then repeated our analysis. Withboth dependent variables, the results support all fourhypotheses (p, .001); thus, our results are robust toalternative measures of category promotion.

Second, we introduced lagged versions of our textand image promotion dependent variables. We col-lected the data for these variables in late March andearly April 2014 using the same data collectionprocedures described earlier for our dependent var-iables. By including these variables, we control fortwodifferent alternative explanations. First, wehaveargued that category promotion is a function ofcontextual distinctiveness within specific contexts(e.g., region, industry). However, an alternativepossibility is that such promotion is the result ofcollective legitimation across multiple contexts(Navis & Glynn, 2010; Wry et al., 2011).11 For in-stance, in our case it could be that category pro-motion is simply a function of the national-levellegitimacy of the B Corporation category, whereasthe context specific effects we observed are epiphe-nomenal. A second possibility is that our analysissuffers from an omitted variable problem, such thatcategory promotion observed in late 2015 wasinfluenced by factors unobserved in our analysis. Byincluding these lagged versions of our dependentvariables,we are able to control for bothpossibilities.Aswould be expected by prior research on collectivelegitimation, higher levels of category promotion in

11 Although we would have preferred to test this possi-bility using repeated observations of the total number of BCorpmembers over time, given the limited panel structureof our data, such a test was not possible. Instead, we useprior promotion as an indicator of prior legitimacy. Thatsaid, understanding the dynamics between collective andcontextual category legitimation is an area ripe for futureresearch (Deephouse, Bundy, Tost, & Suchman, 2017).


TABLE3

Hurd

leMod

elsof

Categ

oryIm

agePromotion

Mod

el10

Mod

el11

Mod

el12

Mod

el13

Mod

el14

Mod

el15

Mod

el16

Mod

el17

Mod

el18

Con

stan

t22.53

3*1.45

5***

22.40

1***

22.64

6***

22.27

9***

22.49

3***

22.70

3***

1.48

4***

1.89

4***

(1.165

)(0.084

)(0.044

)(0.047

)(0.045

)(0.048

)(0.050

)(0.074

)(0.077

)Sou

thregion

21.01

5*21.29

9***

21.59

8***

21.56

5***

21.28

8***

21.26

3***

20.95

9***

21.23

0***

20.82

5***

(0.426

)(0.025

)(0.023

)(0.023

)(0.025

)(0.025

)(0.027

)(0.024

)(0.025

)Midwest-mou

ntain

region

0.03

20.71

0***

20.64

1***

20.63

7***

20.60

3***

20.59

5***

20.56

6***

20.64

2***

20.60

8***

(0.431

)(0.017

)(0.016

)(0.016

)(0.017

)(0.017

)(0.017

)(0.017

)(0.017

)W

estreg

ion

20.17

20.41

9***

20.79

3***

20.78

5***

20.47

2***

20.47

0***

0.58

3***

20.63

8***

0.58

5***

(0.433

)(0.027

)(0.016

)(0.016

)(0.019

)(0.019

)(0.031

)(0.018

)(0.032

)Servicesindustries

1.25

427.57

4***

0.76

4***

0.78

9***

0.62

0***

0.64

9***

0.63

9***

25.50

2***

26.27

9***

(1.446

)(0.140

)(0.013

)(0.013

)(0.013

)(0.013

)(0.013

)(0.118

)(0.120

)Employe

es20.00

10.00

1***

0.00

2***

0.00

2***

0.00

2***

0.00

2***

0.00

2***

0.00

1***

0.00

1***

(0.001

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)Bestinclass

0.07

620.24

8***

20.15

5***

20.15

8***

20.14

6***

20.14

9***

20.08

3***

20.22

7***

20.10

4***

(0.241

)(0.012

)(0.012

)(0.012

)(0.012

)(0.012

)(0.012

)(0.012

)(0.012

)Soc

ialm

edia

presence

20.05

70.01

720.06

9***

20.05

0***

20.09

9***

20.08

2***

20.05

8***

0.04

4***

0.07

2***

(0.146

)(0.010

)(0.009

)(0.009

)(0.009

)(0.009

)(0.009

)(0.010

)(0.010

)W

ebsite

imag

es(log

)0.39

0***

0.75

2***

0.75

5***

0.75

9***

0.73

4***

0.73

9***

0.70

4***

0.74

6***

0.70

1***

(0.051

)(0.003

)(0.003

)(0.003

)(0.003

)(0.003

)(0.003

)(0.003

)(0.003

)Annual

sales(log

)0.00

50.09

4***

0.09

2***

0.10

4***

0.09

3***

0.10

3***

0.09

4***

0.09

1***

0.07

5***

(0.034

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)Categ

oryfounder

0.47

220.05

3*0.33

0***

0.38

3***

0.21

1***

0.26

6***

0.38

1***

0.03

50.19

5***

(0.472

)(0.027

)(0.023

)(0.023

)(0.023

)(0.023

)(0.023

)(0.026

)(0.026

)Mon

thsce

rtified

20.00

820.00

8***

20.01

0***

20.01

1***

20.00

9***

20.01

0***

20.01

0***

20.00

9***

20.01

2***

(0.006

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)Com

pan

yag

e0.01

920.01

0***

20.00

8***

20.00

8***

20.00

8***

20.00

9***

20.00

8***

20.01

0***

20.00

7***

(0.011

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)(0.000

)BornB

20.20

50.49

7***

0.55

3***

0.57

0***

0.64

7***

0.65

4***

0.67

0***

0.51

1***

0.51

4***

(0.280

)(0.014

)(0.012

)(0.012

)(0.013

)(0.013

)(0.013

)(0.014

)(0.014

)Reg

ional

public

compan

ies(residual)

0.25

420.35

7***

20.15

1***

20.15

5***

20.31

5***

20.31

5***

20.45

7***

20.37

1***

20.55

0***

(0.152

)(0.008

)(0.007

)(0.007

)(0.008

)(0.008

)(0.009

)(0.008

)(0.009

)Industry

public

compan

ies

0.02

920.11

1***

20.02

7***

20.02

3***

20.03

9***

20.03

5***

20.02

4***

20.10

5***

20.11

2***

(0.056

)(0.003

)(0.002

)(0.002

)(0.002

)(0.002

)(0.002

)(0.003

)(0.003

)Reg

ional

catego

rypee

rs20.00

320.00

2***

20.00

5***

20.00

5***

(0.002

)(0.000

)(0.000

)(0.000

)Industry

catego

rypee

rs20.00

40.03

5***

0.03

3***

0.03

6***

(0.006

)(0.001

)(0.000

)(0.000

)H1:

Reg

ional

contextual

distinctiven

ess

20.46

22.33

5***

0.50

6***

0.43

5***

0.46

4***

10.983

***

11.782

***

(0.743

)(0.046

)(0.028

)(0.029

)(0.028

)(0.301

)(0.311

)H2:

Industry

contextual

distinctiven

ess

0.13

71.87

2***

1.68

1***

1.63

4***

7.14

0***

1.78

6***

9.01

8***

(0.732

)(0.042

)(0.044

)(0.044

)(0.147

)(0.041

)(0.146

)H3:

Reg

ional

contextual

distinctiven

ess

0.07

4***

0.07

3***

xregion

alca

tego

rypee

rs(0.002

)(0.003

)H4:

Industry

contextual

distinctiven

ess

0.06

9***

0.09

9***

xindustry

catego

rypee

rs(0.002

)(0.002

)Zeroco

untm

odel

N/A

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

AIC

546.0

5599

2.4

6250

1.7

6221

2.7

6098

2.3

6077

7.5

5900

0.3

5537

5.1

5255

9.2

Log

like

lihoo

d225

3.0

227

956.2

231

214.9

231

069.4

230

454.2

230

350.7

229

460.1

227

647.6

226

237.6

Notes:n

546

5;stan

darderrors

inparen

theses.

*p,

0.05

**p,

0.01

***p,

0.00

1


early 2014 lead to higher levels of category pro-motion in late 2015. Specifically, the lagged vari-ables are positive and significant predictors of thebinary decision to promote (p, .05 for text; p, .001for images), and of the extent of promotion (p, .001for both text and images). However, even after con-trolling for these effects, our four hypothesized ef-fects remain significant (p , .001) predictors ofcategory text and image promotion. Thus, these re-sults test the alternative explanations describedabove, and provide further support for the robust-ness of our findings and the overall causal logic ofour hypotheses.

Third, to test whether our results are sensitive tothe way in which we defined industries and regions,we constructed alternative measures for both. Werecalculated industry contextual distinctiveness,industry category peers, and industry public com-panies using the industry definitions proposed byWaddock and Graves (1997). This change had theeffect of doubling the number of industry categoriesused in our analysis. We then repeated our originalanalysis. Results show significant support (p, .001)for all four hypotheses, using both our category textpromotion and category image promotion de-pendent variables. Similarly, we geocoded the lati-tude and longitude of each B Corp, and thenrecalculated regional contextual distinctiveness, re-gional category peers and regional public companiesusing a 75-kilometer radius around each Certified BCorporation.We then repeated our original analysis.Results show significant support (p , .001) for allfour hypotheses, using both our category text pro-motion and category image promotion dependentvariables. Thus, our results are robust to alternativedefinitions of industry and regional boundaries;contextual distinctiveness significantly predictscategory promotion whether using coarser or moregranular contextual boundaries.

Finally, we performed several additional tests re-garding the effects of social media on category pro-motion.12 To do so, we began by collecting data onthenumber ofTwitter followers, if any, for all of theBCorps in our sample. We then calculated a firm’sTwitter followers as the natural log of these data.Next, we checked the robustness of our reported re-sults by testing two new specifications: models inwhich we added Twitter followers in addition to so-cial media presence; and models in which we

substituted Twitter followers for social media pres-ence. For both specifications, results show signifi-cant support (p, .001) for all four hypotheses usingboth our category text promotion and category imagepromotion dependent variables.

We also made use of this new variable in a moresubstantive way. Specifically, we substituted thisvariable for our measures of category peer members,enabling us to test the robustness of our moderatingeffects using a more audience-centric measure. Theresults of these new models show that regional con-textual distinctiveness x Twitter followers and in-dustry contextual distinctiveness x Twitter followersare positive and significant predictors (p , .001) ofboth category text promotion and category imagepromotion. Thus, our results regarding the moder-ating effects of the contextual meaningfulness ofa category are robust to a measure of audience at-tention. In sum, we tested the robustness of our re-sults in 14 additional sets of models finding supportfor our four hypotheses using measures of both textand image promotion in all cases.

Qualitative Analysis of B Corp Membership andPromotion

To further understand and validate the mecha-nisms underlying our results, we conducted semi-structured interviews with B Corp entrepreneursand executives.13 Our interviews focused on un-derstanding why organizations opted to becomeCertified B Corporations, yet differed in their pro-motion of the certification. We were concernedprimarily with surfacing whether the individualswithin these organizations who were responsiblefor both membership- and promotion-related de-cisions actually perceived and were influenced bythe subordinate category distinctiveness affordedby B Corp certification relative to non-certifiedcompanies within their basic categories, as well ascontextual familiarity in the form of subordinatecategory peer group size. The analysis below is basedon 49 interviews—29 conducted by the co-authorsbetween July and October 2015, and 20 interviewspreviously published in The B Corp Handbook(Honeyman, 2014). Our interviews averaged 29minutes in length, and ranged from 20 to 48minutes. With one exception, the intervieweesgave us permission to record their interviews,

12 We thank two anonymous reviewers for their re-spective comments, which inspired these two differenttests.

13 We thank associate editor Martine Haas and twoanonymous reviewers for encouraging us to perform theseinterviews as part of the revision process.


which we had transcribed. To identify potentialinterviewees, we relied primarily on theoreticalsampling of B Corps that were significant promotersor non-promoters, and embedded in amix of differentcontexts (Lincoln & Guba, 1985; Patton, 2002). Usingthe 231 pages of transcripts from our interviews, ourinterview notes, and the secondary interviews, weperformed a theme analysis (Miles & Huberman,1994). The themes that emerged are discussed below,andexamplesof supportingquotesare summarized inTables 4 and 5.

Based on our analysis of the interview data, themostcommon reason for pursuing B Corp certification wasbecause it was aligned with the organization’s pre-existing mission, purpose, values, or identity (seeTable 4), a finding that is consistentwith prior literature(Navis & Glynn, 2010; Wry et al., 2011). In addition tothis “internal” alignment, interviewees reported that BCorpmembershipprovided important “external” third-party validation and legitimation of their organizations’sustainability commitments. Quite a few intervieweesalso reported that they were motivated to become andremain B Corp members because they believed the as-sessment process yieldednew innovations andpracticeimprovements. Finally, many interviewees describedmembership in theBCorp community as a benefit in itsownright, enabling themto learn fromandoccasionallytransact with other B Corps. In sum, our intervieweesreported that their companies became BCorpmemberslargely for identity enactment and validation.

Our analysis of the interview data also clarifiedthat the motivations underpinning members’ pro-motion of the B Corp certification differed sub-stantially from theirmotivations to become certified.The most common explanation provided by in-terviewees for why their companies touted the BCorporation certification related to differentiatingthemselves from non-certified companies (seeTable 5). For instance, in the case of industries suchas office supplies and business insurance, the BCorpcertificationwas seen as offering companies away ofstanding out in an otherwise commodified land-scape. Alternatively, in the case of cities and stateswhere sustainability considerations were perceivedas being of broad interest, our respondents reportedfeeling that the B Corp certification sometimes “gotlost in the crowd,” thereby mitigating the opportu-nity to achieve distinction by way of promotion.

Additionally, our interviewees revealed that otherB Corp members influenced their category pro-motion activities. Whereas some regions havea growing network of BCorps,whichwas reported asbuoying categorypromotion, other regions haveonlya few B Corps, which was perceived as stuntingcategory promotion. More generally, most of ourinterviewees talked specifically about the pres-ence (or absence) of other B Corps in their states(e.g., Colorado, Vermont, Massachusetts, California,Illinois, Montana were all specifically invoked byour interviewees) and, to a lesser extent, their cities

TABLE 4Drivers of B Corp Membership

Theme Example quotes

Alignment with a company’smission, purpose, values oridentity

“We were founded in 1982 with the sole purpose of incorporating environmental, social andgovernance factors into the investment process. We consider ourselves to have been a B Corp longbefore there was a name for it.” Interview No. 27

“Theprinciple benefit to Patagonia fromBCorpcertification . . . is themechanism it provides toprotectthe company’s core values during succession.” Interview No. 33

Validating and legitimatinga company’s sustainabilitycommitments

“Achieving B Corp certification also provides validation of a company’s efforts to manage forsustainability, and the quality of the B Impact Assessment enhances the legitimacy and value of thisthird-party review.” Interview No. 48

“Our co-founder really believed that we needed third-party criteria to validate our own internalbusinesspractices, sowecould speakconfidently andwith authorityonourmission.At that point intime we certified ourselves.” Interview No. 7

Innovation and practiceimprovement

“It’s been a good tool for us to improve our practices and improve our organization. Honestly, we getmore benefit out of it in thatway thanwehave externally. It’s not something that a lot of people seemto understand.” Interview No. 21

“Internally, the B Corp framework helps us track our progress and hold our feet to the fire. It’s away ofmaking sure we are progressing against the journey we set out for ourselves.” Interview No. 31

Membership as community “At the endof the daywe chose to becomecertified becausewe felt thatwewere going to becomeapartof a really intentional learning community that would just frankly make us better.” Interview No. 3

“[Our biggest surprise has been] the powerful fellowship among certified B Corps.” Interview No. 48


(e.g., interviewees mentioned Santa Barbara andChicago). In terms of industry effects, several of ourinterviewees worked for businesses that providedinternet marketing and related services, and eachwas clearly familiar with other B Corps in that in-dustry. In another case, a business owner reportedthat the number of other B Corps in his industry(coffee roasting) directly influenced his categorypromotion.

In sum, what we have conceptualized as a cat-egory’s contextual distinctiveness provides an aptexplanation for differences in category promotionas reported by our interviewees. Similarly, thepresence and absence of other B Corps, whetherwithin a region or industry, appears to have beensalient and influential to these category promotionactivities in ways that are consistent with our the-orization. Thus, our interviews confirmed the corethesis of this research: decisions regarding categorymembership and promotion are driven by differentprocesses. Category promotion in particular isdriven by the distinctiveness of the subordinatecategory within the wider basic category and sub-ordinate category peers amplify or attenuate suchcategory promotion.

DISCUSSION

Our paper wasmotivated by both a theoretical andan empirical puzzle: Why would organizations be-come members of a category (e.g., obtain B Corp

certification), but then abstain from promoting thatcategory? We theorized that such category pro-motion, and, conversely, promotional forbearance,could be explained as a function of the subordinatecategory’s contextual distinctiveness. Organizationspromote their subordinate category memberships asa way of standing out from organizations that sharetheir basic category. Conversely, we argued thatwhen association with a particular subordinate cat-egory fails to provide such distinctiveness, itsmembers forego promotion opportunities. We alsotheorized that the effects of a subordinate category’scontextual distinctiveness are amplified by the sizeof the subordinate category peer group within thatcontext. By increasing the contextual familiarity ofthe category, a larger peer group increases members’likelihood to leverage the subordinate category’sdistinctiveness by way of promotion. In the case of BCorps, our mixture of statistical and qualitative evi-dence provides strong support for our theoreticalmodel of a subordinate category’s contextual dis-tinctiveness and promotional forbearance, as well asthe micromechanisms underlying this relationship.A further strength of our study was our use of bothtextual and visual data, the latter being especiallynovel relative to prior research on categories andinstitutions (Meyer et al., 2013, 2017). Notably, ourfindings suggest that organizationsmay not perceivestrong qualitative distinctions between verbal andvisual modes, using these different semiotic modesinterchangeably or as substitutes for one another.

TABLE 5Drivers of B Corp Category Promotion

Theme Example quotes

Conveying distinctiveness vs.category non-members

“We’re in the office supply industry. Office supplies just appear on your desk. It’s not something thatpeople are very cognizant about. So having something else to leadwith has been game changing forus. Now I talk about us being a B Corp; office supplies is a secondary conversation. It’s helped us tomake a traditional industry a lot more interesting. To really show the impact that you can have isgreat.” Interview No. 23

“We are different from most traditional mainstream businesses, and proudly wearing that badge.”Interview No. 20

“The certification provides us with positive competitive differentiation . . . If you’re a purpose-centered business, this is a great way to distinguish yourself.” Interview No. 47

Size of membership breedscontextual familiarity

“At the time there were not as many B Corps so it didn’t mean as much. I think there weren’t as manypeople familiarwith it as certainly they are now.Maybe the value thenwasn’t as strong just becauseit’s like, ‘Oh, we’re B Corp,’ and people are like, ‘What is that?’” Interview No. 20

“The size of the B Corp community influences promotion quite a bit. Chicago is probably a bit biggerthan some, butwe’remuch smaller thanNewYork or San Francisco. In general, I don’t think there’sa huge awareness in Chicago. That’s part of what we’re trying to do. I think the biggest differentiatoris a company like Method has the resources to come into a community and really cement change,and really make themselves known.” Interview No. 18

“There are four of us inMontana.Montana’s a big area, so regionally it is hard to have that kind of vibethat I’ve heard other B Corps having in the bigger cities.” Interview No. 28


Below we discuss how our findings contribute toscholarship on categories.

Promotional Forbearance: Beyond Stigmatization

Prior research has noted a relationship betweenstigma and what we have conceptualized as pro-motional forbearance. Hudson and Okhuysen(2009), for example, showed how men’s bathhousespurposefully avoided promoting prototypical cate-gory features as a means of restricting negativeattention to themselves and their partner organiza-tions. Durand and Vergne (2015) reinforced thesefindings, noting that within the arms industry, or-ganizations not only avoided publicizing theirmembership, but also engaged in divestment as ameans of further distancing themselves from a stig-matized category. Given this evidence, the implicitassumption has been that a lack of stigma wouldeliminate promotional forbearance. As a corollary,when a category has broad societal appeal, such as inthe case of the B Corp certification, prior researchsuggestsweshouldexpect ahighdegreeofpromotion.

The findings from this study challenge these as-sumptions. We found that organizations activelypursue the B Corp certification, a time consumingand lengthy process, yet some opt to forego pro-moting it. In other words, the choice to obtainmembership in a particular category and the choiceto subsequently promote that category are driven bydifferent factors. Prior research has noted that cate-gory membership provides a basis from whichorganizations establish their identities while simul-taneously asserting their affinity to other categorymembers, enabling a shared collective identity(Navis & Glynn, 2011; Wry et al., 2011). Our theori-zation and findings suggest that members’ sub-sequent promotion is driven by a different logic:a desire to promote distinction rather than similarity.And, in this case, the primary reference group un-derpinning this choice appears to be basic categorymembers rather than other subordinate categorymembers. In short, members of subordinate cate-gories promote their certification to signal distinc-tiveness relative to those organizations that belong tothe same basic categories (e.g., industry, region), butnot to the same subordinate categories. Thus, even incases wherein a subordinate category is highly cel-ebrated, members may engage in promotional for-bearance if the subordinate category fails to offeropportunities for distinctiveness vis-a-vis the widerbasic category. Additionally, our findings reveal thata limited peer group further diminishes members’

promotion. This suggests that decreases in peergroup size can challenge a subordinate category’sstatus as familiar or recognizable, thereby resultingin reduced promotion. This effect is likely most ap-parent within the context of emerging subordinatecategories.

Taken together, these two findings also comple-ment and extend Delmas and Grant’s (2014) findingsthat wineries with organic certifications often do notpromote such certifications. Additionally, theyfound that while the organic certification increasedproduction costs, the subsequent act of promotingthis certification on the wine label reduced the priceconsumers were willing to pay. In discussing thesefindings, the authors flagged their inability to ex-plain the variation in promotion and called for futureresearch to do so. Our study addresses this puzzle byproviding a theoretical framework that highlightsthe role of a category’s contextual distinctivenessand contextual familiarity in driving category pro-motion. Compared to our multi-region, multi-industry study, Delmas and Grant studied a singlestate (California, in which they nested 160 appella-tions within seven larger regions) and a single in-dustry (approximately 1,500 Wine Spectator-ratedwineries). Viewed through the lens of our study,this context was characterized by low category dis-tinctiveness, a small category peer group, and ex-tremely low category promotion (only 16 wineriesout of 1,495 wineries in their sample included an‘eco-label’). First, in a region such as California, eco-certifications are likely to provide low distinctive-ness, given many firms may be committed to similarpractices whether certified or not. Second, the eco-certifications in question appeared to have very lowfamiliarity (e.g., only 314 out of 13,111 wine-yearobservations had either of the two certifications).Thus, our theoretical framework appears to providea potential explanation for the extremely low cate-gory promotion over which Delmas and Grantpuzzled.

Category Distinctiveness: The Contrast BetweenSubordinate and Basic Categories

To date, organization theorists who have exploredthe notion of category distinctiveness have focusedon the extent towhich similar features and identitiesamong category members enable greater contrast(McKendrick, Jaffee, Carroll, & Khessina, 2003;Negro et al., 2010). These researchers have assumedthat greater consistency among category membersyields stronger category boundaries, and thus that


members differ substantively from non-members intheir core features by definition (Glynn & Navis,2013; Zuckerman, 1999). This assumption, however,ignores the empirical reality that, in many cases,membership and non-membership in categories isnot so clear-cut. For example, in many cases of cer-tifications, organizations that forego certification canfully resemble organizations that pursue certifica-tion. In these cases, category distinctiveness cannotmerely be defined by the similarities betweenmembers, since an evaluation of member to non-member differences would in fact reveal no differ-ences at all.

This study addresses these conceptual de-ficiencies by drawing on earlier work from socialpsychology that highlights the importance of cate-gory hierarchies (Hunn, 1975; Lakoff, 1987; Roschet al., 1976).Whereas theperceived contrast betweenbasic categories is often apparent by definition(e.g., dogs versus cats), the potentially distinctivefeatures underpinning subordinate categories mayor may not be salient outside the subordinate cate-gory. As such, our study suggests that assessments ofa category’s distinctiveness must acknowledge thedifferences between the core features or practicesthat define the subordinate category and the featuresand practices that are already prevalent among otherorganizations within the relevant basic category. Wehave done so in this study, thereby extending exist-ing conceptualizations of category distinctiveness,as well as our understanding of how such distinc-tiveness can affect important organizational actionssuch as category promotion. Moreover, by recog-nizing category distinctiveness not as an internalfeature of the category, but as a situated and re-lational feature that accounts for differences betweena subordinate category and its context (i.e., basiccategory), we contribute to research which has pro-posed a “categorization by association” perspective(Garud, Gehman, & Karnøe, 2010), in which cate-gories and their members are seen as embedded ina matrix of institutional environments (Douglas,1986; Greenwood, Raynard, Kodeih, Micelotta, &Lounsbury, 2011).

We believe these findings regarding the effects ofa subordinate category’s contextual distinctivenesson members’ promotion open potential opportuni-ties for future research.We focused onorganizations’decisions to promote their certification and specifi-cally how characteristics of both non-certified orga-nizations and certified organizations affect thosedecisions. The proposed mechanism, which wasfurther substantiated through our qualitative data

collection, was the extent to which member versusnon-member differences shaped the perceptions oforganizational members responsible for promotiondecisions. In the future, however, researchers mightalso incorporate measures of audience evaluations(i.e., valence or reputation) of the respective sub-ordinate category as an outcome variable, such thatone could then test the actual derived value (versusthe perceived potential for such value) over timefrom promotion or non-promotion. For example, al-though category promotion appears to increase incontexts where there is a noticeable difference be-tween the category and the non-member referencegroup, suchcontextsmaybe the least likely to rewardcategory promotion, given the potential for morenegative valence. In other words, while some cate-gories may provide differentiation, they may not beculturally valued. Investigating the strategies firmsuse to respond to such tensions appears to be anotherfruitful avenue for future research.

These findings from our study also highlight aninteresting theoretical tension between the diffusionof subordinate categories and those categories’ con-textual distinctiveness to be considered in futureresearch. On the one hand, prior studies of diffusionsuggest that in later stages of a category’s maturity,increases in membership size may be accompaniedby increases in isomorphic pressures that compelnon-members to adopt similar affiliations or prac-tices (Guler, Guillen, & Macpherson, 2002). Thus, itmay be at these later stages that an accompanyingincrease in isomorphic pressure will reduce thecontextual distinctiveness of the subordinate cate-gory relative to the other organizations within therespective basic category. In such cases, we mightexpect an inverse U-shaped relationship if consid-ered longitudinally, such that a growing member-ship size (and hence contextual familiarity) wouldinitially bolster the effects of category distinctive-ness on members’ promotion (as predicted and evi-denced in this paper), yet over time attenuate theseeffects when coupled with isomorphic pressure.Although this might be the expected pattern overtime and across contexts, a number of recent studiesquestion the implied causal relationship betweenpractice diffusion and isomorphism (Ansari, Fiss, &Zajac, 2010; Colyvas & Jonsson, 2011; Raffaelli &Glynn, 2014). In line with these studies, our modelsuggests the possibility of an extreme situation inwhich a subordinate category sustains a high degreeof distinctiveness relative to other organizations inthe respective basic category despite the presenceof an exceptionally high degree of contextual


familiarity. In such cases, our model would predicta very high degree of promotion. These extremecases might offer an intriguing context for futurequalitative research to disentangle the diffusion andinstitutionalization of categories (Colyvas & Jonsson,2011).

Enriching our Understanding of Intra-CategoryVariation

By conceptualizing subordinate categories as in-tersecting with potentially divergent basic cate-gories, we also contribute to and extend existingresearch on organizational differences within cate-gories. To date, intra-category variation (i.e., differ-ences in the identities and actions of the memberscomprising aparticular category) has been explainedeither as a function ofmembers’desire to distinguishthemselves from other category members (Navis &Glynn, 2011) or more generally as a function of therespective category’s leniency (Pontikes & Barnett,2015). For example, an organization that is focusedon using commercial methods to address socialproblems might initially seek to establish alignmentwith the codes that typify the “social entrepreneur-ship” label, yet subsequently seek ways to differen-tiate from other organizations using that same label.Here our research extendsprior researchon “optimaldistinctiveness” (Deephouse, 1999; Zhao et al., 2017)to the domain of categories, showing how distinc-tiveness within a particular category contributes todifferences in category members’ strategic actions.For categories such as social entrepreneurship,which currently have less well-defined boundaries,existing research findings would suggest that suchdiffuse boundarieswould naturally attract and allowformembers that differ substantively in their features(Pontikes & Barnett, 2015). In that case, any observedintra-category variation is attributable not to organi-zations’ efforts to achieve optimal distinctivenessbut an outcome of a loosely defined category.

Our findings and arguments, however, provide animportant extension to both positions, establishingthat organizational variation is driven not only byintra-category rivalry and structural characteristicsof the category, but also by the intersecting re-lationships between subordinate and basic cate-gories. Specifically, our findings demonstrate howorganizations can be simultaneously embedded inmultiple categories (Garud et al., 2010). For instance,a subordinate category may offer opportunities fordistinction within one basic category (e.g., the orga-nization’s regional context), while offering no such

opportunities within another basic category (e.g.,its industrial context). We show then that these dif-ferent intersecting and hierarchical category re-lationships are important in that they help to explainintra-category variance.

In this study, we focused on intra-category varia-tion in terms of promotional forbearance, explainingthis outcome as a function of the category’s con-textual distinctiveness within different regions andindustries. Yet in doing so we open several oppor-tunities for future research that might consider theeffects of a category’s contextual distinctiveness onother types of intra-category variation. For instance,how does a category’s contextual distinctiveness af-fect the extent to which organizations seek to aligntheir identities to category codes or engage in cate-gory spanning? Moreover, since the B Corp categoryis highly value-laden, we operationalized the dis-tinctiveness of this category in terms of the differ-ences between the values practices emphasized bythe category and those that defined the salient non-members. As such, future research could alsocompare intra-category variation in less normativecategories, wherein contextual distinctiveness isdefined more in terms of technical and material dif-ferences rather than symbolic distinctions (e.g.,Delmas & Montiel, 2008).

Practical Implications

Although the contributions of this study extendwell beyond the context of B Corporations, socialentrepreneurship, and certifications more generally,we see an opportunity to clarify how policymakersand practitioners might apply our insights. First,within the context of Certified B Corporations andsocial entrepreneurship, organizations are oftencharacterized by multiple motivations (Miller,Grimes, McMullen, & Vogus, 2012) and identities(Wry & York, 2017). Together scholars have notedhow these factors can create potential organizationaltensions (Besharov & Smith, 2014; Jay, 2013; York,Hargrave, & Pacheco, 2016). Certifications (such asthe B Corp) are one way to codify the practices androutines associated with otherwise ambiguous andcomplexundertakings, thus providing legitimacy forcertified organizations. Indeed, as the B Corp entre-preneurs and executives we interviewed attested,becoming certified helped them to resolve many ofthe perceived tensions associated with trying toemploy commercial methods to solve social andenvironmental problems. While certification helpedthese companies address one set of challenges, in its


wake was a more generic challenge of achievingcontextual distinctiveness relative to non-certifiedfirms embedded in their various regional and in-dustrial contexts. In other words, our research sug-gests that an exclusive emphasis on internal tensionsbetween purpose and profits may be misplaced. Forpolicymakers and advocates of sustainable organiza-tions, the challenge uncovered by our research has todo with the increasing competitive dynamics be-tween social enterprises and more traditional orga-nizations that have begun to similarly employsustainable valuespractices.Given social enterprises’presumed raisond’etreof creating social value, futureresearchshould seek to furtherunderstandhowsocialenterprises respond to the potential opportunity orchallenge associated with increased competition.

Second, for organizations providing certifications(e.g., B Lab, FairTrade) our findings suggest a para-dox. Whereas certification providers may find iteasier to attract new members in contexts that sharethe values of the certification, our theory and find-ings suggest such members are least likely topromote the certification, owing to the lack of dis-tinctiveness to be gained by promoting the certifi-cation in such contexts. Instead, our researchsuggests that certification providers may do well tofocus on recruiting and supporting new memberswithin regions and industries wherein the corre-sponding labels provide greater cultural value astools for differentiation (Garud, Schildt, & Lant,2014; Lounsbury & Glynn, 2001).

CONCLUSION

Category promotion is a critical means by whichmembers assert their distinctiveness. This study isa first attempt at understanding variance inmember-based category promotion and helps explain why anorganization might opt into a category and yet sub-sequently forego opportunities to promote it. Ourtheory and findings contribute to and extend schol-arship on promotional forbearance, the contextualdistinctiveness of categories, and intra-categoryvariance. We hope our study fosters future work inthis area. For instance, we believe that social media(e.g., Twitter, Facebook) offers a particularly richcontext for examining the dynamics of categorypromotion, as organizations use these new channelsto convey their distinctiveness. Perhaps more nota-bly, opportunities exist to study how differences incategory promotion affect important outcomes suchas resource acquisition, organizational survival, andsocial impact. Given the growth in organizations’

efforts to directly engage and influence stakeholdersthroughamultitudeofdifferent informationchannels,studying how and why organizations exert that influ-ence is clearly important and largely underexplored.This study on category promotion lays a foundationfor advancing such research.

REFERENCES

Albert, S., & Whetten, D. A. 1985. Organizational identity.Research in Organizational Behavior, 7: 263–295.

Ansari, S. M., Fiss, P. C., & Zajac, E. J. 2010. Made to fit:How practices vary as they diffuse. Academy ofManagement Review, 35: 67–92.

Besharov, M. L., & Smith, W. K. 2014. Multiple in-stitutional logics in organizations: Explaining theirvaried nature and implications. Academy of Man-agement Review, 39: 364–381.

Bucklin, R. E., & Sismeiro, C. 2003. A model of web sitebrowsing behavior estimated on clickstream data.JMR, Journal of Marketing Research, 40: 249–267.

Cameron, A. C., & Trivedi, P. K. 2013.Regression analysisof count data (2nd edn). New York: Cambridge Uni-versity Press.

Cao,K.,Gehman, J., &Grimes,M.G. 2017. Standingout andfitting in: Charting the emergence of Certified B Cor-porations by industry and region. In J. A. Katz &A. C. Corbett (Eds.), Hybrid ventures: Perspectivesandapproaches to blended value entrepreneurship.Advances in Entrepreneurship, Firm Emergenceand Growth, vol. 19: 1–38. Bingley, UK: Emerald.

Carlos, W. C., & Lewis, B. W. 2017. Strategic silence:Withholding certification status as a hypocrisyavoidance tactic. Administrative Science Quarterly,1839217695089.

Chatterji, A. K., Levine, D. I., & Toffel, M. W. 2009. Howwell do social ratings actually measure corporate so-cial responsibility? Journal of Economics & Man-agement Strategy, 18: 125–169.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. 2003. Ap-plied multiple regression/correlation analysis forthebehavioralsciences (3rdedn).Mahwah,NJ:Erlbaum.

Colyvas, J. A., & Jonsson, S. 2011. Ubiquity and legitimacy:Disentangling diffusion and institutionalization.Sociological Theory, 29: 27–53.

Czarniawska, B. 2008. Alterity/identity interplay in imageconstruction. InD.Barry&H.Hansen (Eds.),TheSagehandbook of new approaches in management andorganization: 49–62. Thousand Oaks, CA: Sage.

Deephouse, D. L. 1999. To be different, or to be the same?It’s a question (and theory) of strategic balance. Stra-tegic Management Journal, 20: 147–166.


Deephouse, D. L., Bundy, J., Tost, L. P., & Suchman, M. C.2017. Organizational legitimacy: Six key questions. InR. Greenwood, C. Oliver, T. B. Lawrence &R. E.Meyer(Eds.), The Sage handbook of organizational institu-tionalism (2nd edn): 27–54. ThousandOaks, CA: Sage.

Delmas, M. A., Etzion, D., & Nairn-Birch, N. 2013. Tri-angulating environmental performance: What docorporate social responsibility ratings really capture?The Academy of Management Perspectives, 27:255–267.

Delmas, M. A., & Grant, L. E. 2014. Eco-labeling strategiesand price-premium: The wine industry puzzle. Busi-ness & Society, 53: 6–44.

Delmas, M., & Montiel, I. 2008. The diffusion of voluntaryinternational management standards: Responsiblecare, ISO 9000, and ISO 14001 in the chemical in-dustry. Policy Studies Journal: the Journal of thePolicy Studies Organization, 36: 65–93.

Douglas, M. 1986. How institutions think. Syracuse, NY:Syracuse University Press.

Durand, R., & Paolella, L. 2013. Category stretching: Reor-ienting research on categories in strategy, entrepre-neurship, and organization theory. Journal ofManagement Studies, 50: 1100–1123.

Durand, R., & Vergne, J.-P. 2015. Asset divestment as a re-sponse to media attacks in stigmatized industries.Strategic Management Journal, 36: 1205–1223.

Echambadi, R., &Hess, J. D. 2007.Mean-centering does notalleviate collinearity problems inmoderated multipleregression models. Marketing Science, 26: 438–445.

Elsbach, K. D., & Bhattacharya, C. B. 2001. Defining whoyou are by what you’re not: Organizational dis-identification and the national rifle association. Or-ganization Science, 12: 393–413.

Garud,R., &Gehman, J. 2012.Metatheoretical perspectiveson sustainability journeys: Evolutionary, relationaland durational. Research Policy, 41: 980–995.

Garud, R., Gehman, J., & Giuliani, A. P. 2014. Contextual-izing entrepreneurial innovation: A narrative per-spective. Research Policy, 43: 1177–1188.

Garud, R., Gehman, J., & Karnøe, P. 2010. Categorizationby association: Nuclear technology and emission-free electricity. In W. D. Sine & R. J. David (Eds.),Institutions and entrepreneurship. Research in theSociology of Work, vol. 21: 51–93. Bingley, UK:Emerald.

Garud,R., Schildt,H.A., &Lant, T.K. 2014. Entrepreneurialstorytelling, future expectations, and the paradox of le-gitimacy.Organization Science, 25: 1479–1492.

Gehman, J., Treviño, L. K., & Garud, R. 2013. Values work:A process study of the emergence and performance oforganizational values practices. Academy of Man-agement Journal, 56: 84–112.

Giorgi, S., Lockwood, C., & Glynn, M. A. 2015. The manyfaces of culture: Making sense of 30 years of researchon culture in organization studies. The Academy ofManagement Annals, 9: 1–54.

Glynn,M. A., &Abzug, R. 2002. Institutionalizing identity:Symbolic isomorphism and organizational names.Academy of Management Journal, 45: 267–280.

Glynn, M. A., & Navis, C. 2013. Categories, identities, andcultural classification: Moving beyond a model ofcategorical constraint. Journal of ManagementStudies, 50: 1124–1137.

Granqvist, N., Grodal, S., & Woolley, J. L. 2012. Hedgingyour bets: Explaining executives’ market labelingstrategies in nanotechnology. Organization Science,24: 395–413.

Greenwood, R., Raynard,M., Kodeih, F., Micelotta, E. R., &Lounsbury, M. 2011. Institutional complexity and or-ganizational responses. The Academy of Manage-ment Annals, 5: 317–371.

Greenwood, R., Suddaby, R., & Hinings, C. R. 2002. Theo-rizing change: The role of professional associations inthe transformation of institutionalized fields. Acad-emy of Management Journal, 45: 58–80.

Grimes, M. 2010. Strategic sensemaking within fundingrelationships: The effects of performancemeasurementon organizational identity in the social sector. Entre-preneurship Theory and Practice, 34: 763–783.

Grimes, M. G., Gehman, J., & Cao, K. 2017. Forthcoming.Positively deviant: Identity work through B Corp cer-tification. Journal of Business Venturing.

Grimes, M. G., McMullen, J. S., Vogus, T. J., & Miller, T. L.2013. Studying the origins of social entrepreneurship:Compassion and the role of embedded agency.Academy of Management Review, 38: 460–463.

Guler, I., Guillen, M. F., & Macpherson, J. M. 2002. Globalcompetition, institutions, and the diffusion of organi-zational practices: The international spread of ISO9000 quality certificates. Administrative ScienceQuarterly, 47: 207–232.

Hannan, M. T., Polos, L., & Carroll, G. 2007. Logics of or-ganization theory: Audiences, codes, and ecologies.Princeton, NJ: Princeton University Press.

Hart, T. A., & Sharfman,M. 2015. Assessing the concurrentvalidity of the revised Kinder, Lydenberg, andDominicorporate social performance indicators. Business &Society, 54: 575–598.

Hilbe, J. M. 2014. Modeling count data. New York: Cam-bridge University Press.

Honeyman, R. 2014. The B Corp handbook. San Fran-cisco, CA: Berrett-Koehler.

Hudson, B. A., & Okhuysen, G. A. 2009. Not with a ten-footpole: Core stigma, stigma transfer, and improbable


persistence of men’s bathhouses. Organization Sci-ence, 20: 134–153.

Hunn, E. 1975. Ameasure of the degree of correspondenceof folk to scientific biological classification.AmericanEthnologist, 2: 309–327.

Jay, J. 2013. Navigating paradox as a mechanism of changeand innovation in hybrid organizations. Academy ofManagement Journal, 56: 137–159.

Jennings, J. E., Jennings, P. D., & Greenwood, R. 2009.Novelty and new firm performance: The case of em-ployment systems in knowledge-intensive service orga-nizations. Journal of Business Venturing, 24: 338–359.

Jones, C., Maoret, M., Massa, F. G., & Svejenova, S. 2011.Rebels with a cause: Formation, contestation, and ex-pansion of theDeNovo category “modern architecture,”1870–1975.Organization Science, 23: 1523–1545.

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

Kennedy, M. T., Lo, J., & Lounsbury, M. 2010. Category cur-rency: The changing value of conformity as a function ofongoing meaning construction. In G. Hsu, G. Negro, &O. Koçak (Eds.), Categories in markets: Origins andevolution.Research in theSociologyofOrganizations,vol. 31: 369–397. Bingley, UK: Emerald.

King, A. A., Lenox, M. J., & Terlaak, A. N. N. 2005. Thestrategic use of decentralized institutions: Exploringcertification with the ISO 14001 management standard.Academy of Management Journal, 48: 1091–1106.

Lakoff, G. 1987. Women, fire, and dangerous things:What categories reveal about the mind. Chicago, IL:University of Chicago Press.

Lee,M.-D. P., & Lounsbury,M. 2015. Filtering institutionallogics: Community logic variation and differentialresponses to the institutional complexity of toxicwaste. Organization Science, 26: 847–866.

Levinas, E. 1999. Alterity and transcendence (M. B.Smith, Tran.). New York: Columbia University Press.

Lincoln, Y. S., & Guba, E. G. 1985. Naturalistic inquiry.Beverly Hills, CA: Sage.

Lounsbury, M. 2007. A tale of two cities: Competing logicsandpracticevariation in theprofessionalizationofmutualfunds.Academy ofManagement Journal, 50: 289–307.

Lounsbury, M., & Glynn, M. A. 2001. Cultural entrepre-neurship: Stories, legitimacy, and the acquisition of re-sources. Strategic Management Journal, 22: 545–564.

Low, M. B., & Abrahamson, E. 1997. Movements, band-wagons, and clones: Industry evolution and the en-trepreneurial process. Journal of Business Venturing,12: 435–457.

Markman, A. B., & Wisniewski, E. J. 1997. Similar anddifferent: The differentiation of basic-level categories.

Journal of Experimental Psychology. Learning,Memory, and Cognition, 23: 54–70.

Marquis, C., Davis, G. F., & Glynn, M. A. 2011. Golfingalone? Corporations, elites, and nonprofit growth in100 American communities. Organization Science,24: 39–57.

Mattingly, J. E., & Berman, S. L. 2006. Measurement ofcorporate social action: Discovering taxonomy in theKinder Lydenburg Domini ratings data. Business &Society, 45: 20–46.

McKendrick, D. G., Jaffee, J., Carroll, G. R., & Khessina,O. M. 2003. In the bud? Disk array producers asa (possibly) emergent organizational form. Adminis-trative Science Quarterly, 48: 60–93.

Meyer, R. E., Hollerer, M. A., Jancsary, D., & van Leeuwen,T. 2013. The visual dimension in organizing, organi-zation, and organization research: Core ideas, currentdevelopments, andpromising avenues.TheAcademyof Management Annals, 7: 489–555.

Meyer, R. E., Jancsary, D., Hollerer,M.A., &Boxenbaum, E.2017. The role of verbal and visual text in the processof institutionalization. Academy of ManagementReview, amr.2014.0301.

Micelotta, E. R., & Raynard,M. 2011. Concealing or revealingthe family? Corporate brand identity strategies in familyfirms. Family Business Review, 24: 197–216.

Miles, M. B., & Huberman, A. M. 1994. Qualitative dataanalysis (2nd edn). Thousand Oaks, CA: Sage.

Miller, T. L., Grimes, M. G., McMullen, J. S., & Vogus, T. J.2012. Venturing for others with heart and head: Howcompassion encourages social entrepreneurship.Academy of Management Review, 37: 616–fdff 640.

MSCI 2015. MSCI ESG KLD STATS: 1991–2014 data sets.MSCI, New York.

Navis, C., &Glynn,M.A. 2010.Hownewmarket categoriesemerge: Temporal dynamics of legitimacy, identity,and entrepreneurship in satellite radio, 1990-2005.Administrative Science Quarterly, 55: 439–471.

Navis, C., & Glynn, M. A. 2011. Legitimate distinctivenessand the entrepreneurial identity: Influence on in-vestor judgments of new venture plausibility. Acad-emy of Management Review, 36: 479–499.

Negro, G., Hannan, M. T., & Rao, H. 2010. Categoricalcontrast and audience appeal: Niche width and criti-cal success in winemaking. Industrial and CorporateChange, 19: 1397–1425.

Paolella, L., & Durand, R. 2016. Category spanning, eval-uation, and performance: Revised theory and test onthe corporate law market. Academy of ManagementJournal, 59: 330–351.

Patton, M. Q. 2002. Qualitative research and evaluationmethods (3rd edn). Thousand Oaks, CA: Sage.


Phillips, N., & Oswick, C. 2012. Organizational discourse:Domains, debates, and directions. The Academy ofManagement Annals, 6: 435–481.

Pollock, T. G., & Gulati, R. 2007. Standing out from thecrowd: The visibility enhancing effects of IPO relatedsignals on alliance formationbyentrepreneurial firms.Strategic Organization, 5: 339–372.

Pontikes, E. G. 2012. Two sides of the same coin how am-biguous classification affects multiple audiences’ eval-uations.AdministrativeScienceQuarterly, 57: 81–118.

Pontikes,E.G.,&Barnett,W.P.2015.Thepersistenceof lenientmarket categories.OrganizationScience, 26: 1415–1431.

Porac, J. F., Thomas, H., & Baden-Fuller, C. 2011. Com-petitive groups as cognitive communities: The case ofScottish knitwear manufacturers revisited. Journal ofManagement Studies, 48: 646–664.

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

Raffaelli, R., & Glynn, M. A. 2014. Turnkey or tailored? Re-lational pluralism, institutional complexity, and the orga-nizational adoption of more or less customized practices.Academy ofManagement Journal, 57: 541–562.

Raftery, A. E. 1995. Bayesian model selection in social re-search. Sociological Methodology, 25: 111–163.

Rao, H. 1994. The social construction of reputation: Certifi-cation contests, legitimation, and the survival of organi-zations intheAmericanautomobile industry:1895–1912.Strategic Management Journal, 15(S1): 29–44.

Reinecke, J., & Ansari, S. 2015.What is a “fair” price? Ethicsas sensemaking.Organization Science, 26: 867–888.

Ricker-Gilbert, J., Jayne, T. S., & Chirwa, E. 2011. Subsidiesand crowding out: A double-hurdlemodel of fertilizerdemand in Malawi. American Journal of Agricul-tural Economics, 93: 26–42.

Rindova, V. P., Petkova, A. P., & Kotha, S. 2007. Standingout: How new firms in emerging markets build repu-tation. Strategic Organization, 5: 31–70.

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

Rosch, E., Simpson, C., & Miller, R. S. 1976. Structuralbases of typicality effects. Journal of ExperimentalPsychology, 2: 491–502.

Ruef, M., & Patterson, K. 2009. Credit and classification: Theimpact of industry boundaries in nineteenth-centuryAmerica.AdministrativeScienceQuarterly,54:486–520.

Sahlin, K., &Wedlin, L. 2008. Circulating ideas: Imitation,translation and editing. In R. Greenwood, C. Oliver, R.

Suddaby, & K. Sahlin (Eds.), The Sage handbook oforganizational institutionalism: 218–242. ThousandOaks, CA: Sage.

Selznick, P. 1957. Leadership in administration. NewYork: Harper & Row.

Sine, W. D., David, R. J., & Mitsuhashi, H. 2007. From planto plant: Effects of certification on operational start-upin the emergent independent power sector. Organi-zation Science, 18: 578–594.

Singer, J. D., & Willett, J. B. 2003. Applied longitudinaldata analysis. New York: Oxford University Press.

Swidler, A. 1986. Culture in action: Symbols and strate-gies. American Sociological Review, 51: 273–286.

Terlaak, A., & King, A. A. 2006. The effect of certificationwith the ISO 9000 Quality Management Standard: Asignaling approach. Journal of Economic Behavior &Organization, 60: 579–602.

Vergne, J.-P. 2012. Stigmatized categories and public dis-approval of organizations: A mixed-methods study ofthe global arms industry, 1996–2007. Academy ofManagement Journal, 55: 1027–1052.

Vergne, J.-P., & Wry, T. 2014. Categorizing categorizationresearch: Review, integration, and future directions.Journal of Management Studies, 51: 56–94.

Waddock, S. A., & Graves, S. B. 1997. The corporate socialperformance–financial performance link. StrategicManagement Journal, 18: 303–319.

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

Weber, K., Heinze, K. L., & DeSoucey, M. 2008. Forage forthought: Mobilizing codes in the movement for grass-fed meat and dairy products.Administrative ScienceQuarterly, 53: 529–567.

Wright, A. L., Zammuto, R. F., & Liesch, P. W. 2017. Main-taining the values of a profession: Institutional workand moral emotions in the emergency department.Academy of Management Journal, 60: 200–237.

Wry, T., Lounsbury,M., & Glynn,M. A. 2011. Legitimatingnascent collective identities: Coordinating culturalentrepreneurship.OrganizationScience, 22: 449–463.

Wry, T., Lounsbury, M., & Jennings, P. D. 2014. Hybridvigor: Securing venture capital by spanning categoriesin nanotechnology. Academy of Management Jour-nal, 57: 1309–1333.

Wry, T., & York, J. G. 2017. An identity-based approach tosocial enterprise. Academy of Management Review,42: 437–460.

York, J. G., Hargrave, T. J., & Pacheco, D. F. 2016. Convergingwinds: Logic hybridization in the Colorado wind energyfield.Academy of Management Journal, 59: 579–610.


Zhao, E. Y., Fisher, G., Lounsbury, M., & Miller, D. 2017.Optimal distinctiveness: Broadening the interface be-tween institutional theory and strategic management.Strategic Management Journal, 38: 93–113.

Zuckerman,E.W. 1999.The categorical imperative: Securitiesanalysts and the illegitimacydiscount.AmericanJournalof Sociology, 104: 1398–1438.

Joel Gehman ([email protected]) is an assistant pro-fessor in the Alberta School of Business at the Universityof Alberta. He holds a PhD degree in management andorganization from Pennsylvania State University. Hiscurrent research examines strategic, technological,and institutional responses to sustainability and valuesconcerns.

Matthew G. Grimes ([email protected]) is an as-sistant professor in the Kelley School of Business atIndiana University. He holds a PhD degree in organiza-tion studies from Vanderbilt University. His current re-search focuses on linkages between entrepreneurship,identity, and categories.

APPENDIXCompanies Interviewed

Actuality Media Lightspan DigitalBen and Jerry’s MightyBytesBeyond Green Partners Mills Office ProductivityCabot Creamery Net BalanceCascade Engineering New Belgium Brewing CompanyThe Change Creation, Inc. Orbit MediaChannel Islands Outfitters PatagoniaCompass(x) Strategy PreserveCook Trading The Redwoods GroupDansko Seventh GenerationEcovations South MountainEtsy Southern Energy ManagementGoLite StoryStudio ChicagoGreen Engineer TerraCarbonGrowers Secret TMI ConsultingThe Ian Martin Group TriCiclosJuhudi Kilimo TrilliumJustNeem Body Care Veris Wealth PartnersKing Arthur Flour West Paw DesignLarry’s Coffee Zullos

Our analysis is based on 49 interviews with entrepreneurs andexecutives from44 companies; the authors conducted 29 interviews,and 20 interviews were published in The B Corp Handbook(Honeyman, 2014). The companies listed above either consented tobe mentioned, or were mentioned in Honeyman (2014).


mailto:[email protected]

mailto:[email protected]

Hidden Badge of Honor: How Contextual Distinctiveness ......Title: Hidden Badge of Honor: How Contextual Distinctiveness Affects Category Promotion among Certified B Corporations Created

Documents