Do LGBTQ-Supportive Corporate Policies Affect Consumer Behavior? Evidence from the Video Game Industry - Petr Parshakov

Please enable JavaScript to view the full PDF

Journal of Business Ethics https://doi.org/10.1007/s10551-022-05137-7 ORIGINAL PAPER Do LGBTQ‑Supportive Corporate Policies Affect Consumer Behavior? Evidence from the Video Game Industry Petr Parshakov1 · Iuliia Naidenova1 · Carlos Gomez‑Gonzalez2 · Cornel Nesseler3 Received: 18 November 2021 / Accepted: 26 April 2022 © The Author(s), under exclusive licence to Springer Nature B.V. 2022 Abstract This paper empirically examines how consumers react when a company marks a product with a gay label. The company under scrutiny is one of the largest video game developers in the world, and the labeled product is a popular video game character. We use a regression discontinuity design to exploit the quasi-experimental setting. The main finding was significant drop in demand for this character and a return to previous levels after approximately 3 months. Possible mechanisms and dynamics were explored by analyzing demand for other characters and by surveying consumers. The results are consistent after performing robustness checks based on grid search, subsampling, and placebo tests. Keywords LGBTQ-inclusive policy · Consumer behavior · Discrimination Introduction on outcomes such as firm performance, employee retention, and recruitment. Our study aims to contribute to this litera- Advances in lesbian, gay, bisexual, transgender, and queer ture by analyzing how LGBTQ brand positioning influences (LGBTQ) rights such as marriage, civic participation, and consumer behavior. Specifically, we investigate the change workplace equality set the strategic agenda of most devel- in demand after the implementation of an LGBTQ-inclusive oped countries. Same-sex marriage became a reality in many policy in the video game industry. countries after the Netherlands passed the first law in 2001.1 This industry is growing rapidly and attracts many con- In many places, serving in the military is no longer deter- sumers from an early age. Video games greatly impact the mined by sexual orientation or gender identity; renowned audience. Positive effects are associated with better social brands publicly support LGBTQ projects; and inclusive well-being when playing video games with others (Halbrook workplaces are becoming the rule in organizations world- et al., 2019). However, research often finds harmful effects wide (Stonewall, 2021). regarding addiction (Van Rooij et al., 2010), the objectifica- Firms and organizations implement LGBTQ-inclusive tion of women (Behm-Morawitz & Mastro, 2009; Gestos policies that take many forms, some of which attract research et al., 2018), and violence and aggression (Adachi & Wil- attention (Anteby & Anderson, 2014; Chintrakarn et al., loughby, 2016; Carnagey et al., 2007; Dill & Dill, 1998). We 2020; Creed et al., 2002; Cunningham & Nite, 2020; Day & consider this setting to analyze the consequences of socially Greene, 2008; Kaplan, 2006; King & Cortina, 2010; Pichler responsible behavior of a video game developer. et al., 2018; Theodorakopoulos & Budhwar, 2015). This Specifically, we identify a quasi-experimental setting in research analyzes the impact that LGTBQ-inclusive meas- 2019 after Overwatch, a team-based video game developed ures and anti-discrimination policies in the workplace have and published by Blizzard Entertainment, declared one of the game characters as gay. On January 8, 2019, a spokesperson announced via Twitter that one of the game’s shooting charac- * Petr Parshakov ters, Soldier:76, was gay (Chu, 2019a). This LGBTQ-friendly pparshakov@hse.ru news was actively discussed in social networks and media, 1 HSE University, Perm, Russia 2 University of Zurich, Zurich, Switzerland 1 Research examines the economic impact of same-sex marriage in 3 Norwegian University of Science and Technology, specific states/regions and find positive associations, e.g., Zhu and Trondheim, Norway Smieliauskas (2021). 13 Vol.:(0123456789) P. Parshakov et al. including business news sources such as Business Insider Related Literature and Hypotheses (Webb, 2019). By analyzing the demand for Overwatch characters at 63 Corporate Social Responsibility points in time, we traced the dynamics of the demand (pick rate) for Soldier:76 before and after the LGBTQ labeling. The Corporate social responsibility measures are aimed at con- pick rate is our dependent variable and represents how often necting the lines of action of organizations with desirable a character is chosen. This variable reflects clean consumer outcomes in society (Bowen, 1953). There is no consensus behavior patterns within the video game setting. Using a in the CSR literature as to whether CSR activities should regression discontinuity design (RDD), we found that the pick be aligned with the company’s financial performance rate of Soldier:76 significantly decreases after the announce- (Wang et al., 2020). Good CSR practices have the poten- ment compared to all the other characters. The effect is quite tial to enhance firm performance and, under certain condi- substantial but rather short lived. This change in consumer tions, even compensate for firm weaknesses, such as poor behavior does not have long-term effects, which generates corporate ability (Berens et al., 2007). implications for firms interested in promoting LGBTQ rights Most empirical studies on CSR focus on factors that through product branding. help align CSR activities with financial performance. Hill- To corroborate that the results are consistent, we performed man and Keim (2001) suggest that building relationships a series of robustness checks including grid search, subsam- with the main stakeholders—consumers, employees, and pling, and placebo tests. Firms also have different products suppliers—is key to enhancing firm performance. Research available for LGBTQ labeling and do not use their whole examines the impact of CSR on particular stakeholders collection for this purpose. To understand the mechanisms such as consumers (e.g., Chernev & Blair, 2015; Inoue behind consumer behavior and the decision-making process et al., 2017; Lee & Shin, 2010) and employees (Onkila & of companies, we examine the same-skill characters available Sarna, 2021). Overall, the findings are mixed (Barauskaite for selection and perform further analyses. Additionally, we & Streimikiene, 2020; Cavaco & Crifo, 2014; Tang et al., perform a survey asking Overwatch players about their views 2012). in order to gain further insight into the situation with respect We identify two main theoretical frameworks dealing to Soldier:76. with this issue. From a stakeholder theory perspective, Free- Our empirical study adds to the narrow body of literature man (1984) discusses CSR and the complex relationship concerning the influence that LGBTQ-inclusive policies have between firms and stakeholder groups. The performance of on firm value (Pichler et al., 2018), innovation (Hossain, et al., the firm can be positively affected if, for example, there is 2020), and credit ratings (Chintrakarn et al., 2020). Our study consumer demand for CSR (McWilliams & Siegel, 2001). also contributes to the literature on consumer responses to cor- Consumer-focused studies show that CSR can positively porate social responsibility (CSR). This literature mainly relies affect a company’s reputation and consumer loyalty (Osakwe on consumer satisfaction and purchase intention, but to date, & Yusuf, 2020; Stanaland et al., 2011; Walsh & Bartikowski, has not considered LGBTQ-related factors (Deng & Xu, 2017; 2013), strengthen the willingness of consumers to engage in Palihawadana et al., 2016; Tian et al., 2011). Our empirical positive word-of-mouth (Lacey et al., 2015), and improve strategy and the quasi-experimental setting allow us to link evaluations of the functional performance of company prod- changes in LGBTQ firm strategies to consumer responses in ucts (Chernev & Blair, 2015). demand and infer a causal relationship that results from a pub- Friedman (1970) frames investment in CSR as an lic company announcement. agency problem: although managers may benefit in terms The paper is organized as follows: Related Literature and of their career goals, the investment does not necessar- Hypotheses reports the findings from previous studies related ily improve the efficiency of the firm itself. Research to LGBTQ-inclusive policies, corporate social responsibility documents overinvestments in CSR by firms that result in measures, firm performance, and consumer behavior. Data and modest financial gain (Barnea & Rubin, 2010) as well as Methods describe the research design and the methodological negative responses of consumers, such as lower behavioral approaches. Results present the findings and discusses possible loyalty in terms of product repurchase (Inoue et al., 2017) interpretations and implications. Concluding Remarks sum up or share of wallet (Ailawadi et al., 2014). Chernev and and conclude the paper. Blair (2015) note that socially responsible firms can be perceived as less competent in their core area of expertise, and emphasize that a company’s prosocial behavior should be aligned with consumers’ moral values. CSR takes many forms, among which corporate vol- unteering, employee fairness, environmental programs, 13 Do LGBTQ‑Supportive Corporate Policies Affect Consumer Behavior? Evidence from the Video… and responsible investment are most commonly consid- 2006; Paul et al., 2011) and they are loyal to products from ered (Ailawadi et al., 2014; Inoue et al., 2017; Wang et al., LGBTQ-inclusive companies (Clermont, 2006; Valenti, 2020). There can be other CSR measures, however, such as 2012). selling locally produced products (Ailawadi et al., 2014) Consequently, a few studies have considered the impact or LQBTQ-friendly and inclusive policies (Brahma et al., of LGBTQ-inclusive policies on the financial performance 2021). Companies may also address CSR and company of firms. Research indicates that LBGTQ-inclusive policies branding strategically (Vallaster et al., 2012). Actions and have a positive effect (Johnston & Malina, 2008; Li & Nagar, public statements that help companies present themselves 2013; Pichler et al., 2018; Wang & Schwarz, 2010). Addi- as open minded, liberal, and supportive regarding the tionally, Chintrakarn et al. (2020) show that companies that rights of minority groups are likely to improve perceived adhere to this type of LGBTQ-inclusive policy enjoy better CSR and, consequently, consumer trust and loyalty (Sta- credit ratings. naland et al., 2011). Among actions to support the LGBTQ Despite the positive association between LGBTQ-sup- movement and enhance perceived CSR, product branding portive policies and performance, establishing a causal rela- is a popular strategy with a strong impact (Friederichsen, tionship is challenging, especially using the firm value of big 2019). companies as the dependent measure. Supportive policies may also have detrimental effects. Chintrakarn et al. (2020) LGBTQ‑Inclusive Policies pointed out that if LGBTQ-inclusive policies are not ben- eficial for shareholders, they may enhance the public image Some brands and firms actively support social movements of managers but be a waste of resources for the organization through publicity and marketing campaigns and product as a whole. Indeed, several studies find that more effective labeling. For example, during soccer Euro 2020, Volkswa- and insider ownership companies invest significantly less gen, Booking, and TikTok showed pitch-side rainbow adver- in CSR (e.g., Cespa & Cestone, 2007; Chintrakarn, et al., tisements and made public statements in favor of LGBTQ 2016; Oh et al., 2011). rights (McCarthy, 2021). We interpret this type of LGBTQ- Additionally, although the public acceptance of LGBTQ friendly initiative from brands that are well positioned in rights has improved in most countries, this group still suffers the market as CSR measures. However, although the public from explicit and implicit discrimination in several social acceptance of the LGBTQ community has improved in some domains (Coffman et al., 2017; Dessel et al., 2017). Chin- countries,2 it does not necessarily mean that the LGBTQ trakarn et al. (2020) find that companies located in regions community or products associated with it enjoy a better posi- with a better attitude toward LGBT communities are more tion (Adamczyk & Liao, 2019). likely to adopt congenial policies. The upshot is that analyz- Previous studies find that LGBTQ-inclusive policies pro- ing only small, local companies can result in biased findings. vide firms and organizations with benefits related to human Beyond the influence of LGBTQ-inclusive policies on resources, marketing, and performance. LGBTQ-inclusive firm financial performance, there are implicit marketing- policies in the workplace, which allow members to be open related benefits in CSR initiatives related to consumer about their sexual orientation, have been linked to better behavior (Brown & Dacin, 1997). CSR initiatives are not employee recruitment, lower employee turnover, and a less always effective, however (Sen & Bhattacharya, 2001). stressful working environment (Clermont, 2006; Metcalf & There is extensive research on the reaction of consumers Rolfe, 2011; Ragins & Cornwell, 2001; Ragins et al., 2007; to different CSR initiatives (e.g., Deng & Xu, 2017; Pali- Stonewall, 2021). hawadana et al., 2016; Tian et al., 2011), but so far we have Research shows that LGBTQ-inclusive companies can only limited evidence on how consumers react to LGBTQ- profit from positive consumer attitudes (Cunningham & inclusive measures. Our paper contributes to this literature Melton, 2014; Patel & Feng, 2021; Pounders & Mabry- by empirically testing how consumers react in terms of Flynn, 2016). In the US, LGBT identification has increased demand when confronted with a new gay-labeled product from 3.5% in 2012 to 5.6% in 2020 (Jones, 2021), which in the video game industry. significantly increases the pool of potential consumers from this group. LGBTQ consumers have higher income levels Product Branding and Consumer Behavior than heterosexual counterparts (Colgan et al., 2007; Iwata, in Computer‑Mediated Environment Product branding is a business strategy that consists of 2 In a 2012 poll directed at members of the LGBT community, 92% describing or designing a product to differentiate it from of the respondents stated that they see society as more accepting other products (Ward & Lee, 2000). For example, car manu- than ten years before (Pew Research Center, 2013). Attitudes toward LGBTQs vary greatly throughout the world and change gradually facturers brand cars as being especially appropriate for fami- (Haerpfer et al., 2020). lies, singles, or adventurous consumers. Airlines brand their 13 P. Parshakov et al. services as safe and reliable. Soft drink companies brand compliance with consumer moral diversity, especially in products as healthy, exciting, exotic, or suited for younger countries where LGBTQ issues face opposition. Blizzard consumers. Fast-food companies try to rebrand their prod- Overwatch League Pride Day is one such example (Venter, ucts as healthy and eco-friendly and expand the variety of 2021). their offerings. A more recent phenomenon is branding a Today’s business setting is strongly influenced by con- product as LGTBQ friendly to attract this group of consum- sumers’ use of online platforms to access information, build ers and enhance perceived CSR (Friederichsen, 2019). relationships, and create social networks (Joo et al., 2011). Popular clothing and footwear brands such as Nike and Social media and online platforms function as electronic Converse also support the LGBTQ + and Black communities word-of-mouth and represent a powerful way to influence by creating products with symbols, colors, and logos that consumer attitudes about products and brands (Podnar & reference these movements. Beyond clothing and footwear, Javernik, 2012). Negative comments toward brands and other global brands such as Skittles, Starbucks, Apple, and products in this context are an important threat for firms. Fossil use product branding to align with the LGBTQ move- Additionally, in computer-mediated environments, social ment. There are many examples of companies that launch norms often become less potent due to reduced social pres- LGBTQ-themed products, especially at public events that ence (Li & Li, 2014) and anonymity (Cirucci, 2013). A maximize attention. person is more likely to behave in a socially desirable way During pride month (June), many companies take the when others are around. Playing video games, individuals opportunity to introduce products—with supporting slo- are more likely to openly express their uncensored attitudes gans—focused on LBGTQ consumers. For example, in (Li & Li, 2014). 2014, San Francisco Burger King started to sell the Proud Whopper and use slogans such as “Be Your Way” and “We Are All the Same Inside.” On June 1, 2021, LEGO The Overwatch Case launched its first LGBTQ-themed set with figures and colors that resemble a rainbow. This initiative by LEGO is part Causal analyses of consumers’ reactions to CSR LGBTQ- of a long-term commitment to creating diverse, inclusive inclusive practices in computer-mediated environments are workplaces in partnership with organizations such as Stone- nonexistent in the literature and a possible strand of novel wall, Workplace Pride, and Open for Business. Similarly, research. In our study, we examine the online multiplayer the video game developer and publisher Blizzard honors game Overwatch. It is a team-based, first-person shooter LGBTQ gamers with the Overwatch League Pride Day, game developed by Blizzard Entertainment. Overwatch is which includes special themed features and merchandising one of the world’s most popular games, generating more (Venter, 2021). than a US$1 billion in revenue and played by some 50 mil- LGBTQ-themed product branding aims to improve the lion consumers worldwide since it was launched in 2016. visibility of members of this minority group and challenge In Overwatch, consumers choose one of 32 different social norms, which often shape behavior (Bursztyn & characters (products) each time they play. Each character Jensen, 2017; Holland et al., 2013; Jones et al., 2014; Li & is unique in appearance and skillset. The skills determine a Li, 2014). CSR measures such as LGBTQ-themed product player’s role in a team and the style of play. branding can create more inclusive societies because aware- On the game’s website, the company briefly describes and ness and role models have the potential to change social updates the stories of the characters. Comics, animated short norms, influence attitudes, and reduce inequality (Bertrand videos, and social networks are used to create these stories. & Duflo, 2017). This is a way of marketing that aims at keeping the consum- On the other hand, LGBTQ-supportive policies can be ers interested and attached to the product. The story of a harmful to companies (Chintrakarn et al., 2016). There are character, however, does not affect its skillset—the product several risks that companies face when implementing CSR stays the same (Venter, 2021). measures. In addition to the danger already discussed—man- In January 2019, Blizzard released the short illustrated agerial gain at shareholders’ expense (Friedman, 1970)— story “Bastet” (Chu, 2019b), which revealed Soldier:76’s companies face the risk of losing consumers with certain close relationship with a man in the past. Managers from religious or cultural values. For instance, after Burger King the company publicly announced through different platforms (BK) aligned some of its products with the LGBTQ move- that the character was gay. ment, some consumers voiced their displeasure on social The announcement attracted considerable attention on media: “You can take that pride BS and shove it BK, me social media. At first, consumers seemed to welcome the and my family are going somewhere else!” (Peterson, 2014). announcement: the initial tweet received 26,000 likes. How- Thus, companies that support LGBTQ initiatives may ever, the most popular retweets were negative. One such was also introduce strategic features in order to balance legal “uninstall overwatch.” The story evoked a huge, though short 13 Do LGBTQ‑Supportive Corporate Policies Affect Consumer Behavior? Evidence from the Video… Fig. 1  Google normalized search activity (from 1 to 100) for the term “Soldier:76” in Jan- uary 2019. The dashed vertical line represents the date of the announcement (January 7). The horizontal line represents the maximum pre-announcement search activity level term, interest in both game and character. Figure 1 shows Research identifies several moderators that influence con- how the interest sharply increased after the announcement. sumer response to CSR measures such as CSR demand from Our empirical analysis aims to examine the consumer consumers (McWilliams & Siegel, 2001), trust and aware- response to the announcement. Specifically, we ask whether ness (Tian et al., 2011), ethical ideologies (Palihawadana the character pick rate was penalized after Blizzard took et al., 2016), identification with a company (Deng & Xu, this LGBTQ-friendly step (Venter, 2021). The setting is 2017), and several demographic aspects such as age, income, well suited to test the response of players to the LGBTQ- or education (McWilliams & Siegel, 2001; Tian et al., 2011). supportive measure because the skillset of the characters is Video game users’ youth challenge the features that moder- unaffected by their sexual orientation. ate positive responses to CSR measures (ESA, 2021). Addi- tionally, research extensively discusses the masculinity and Hypotheses manliness stereotypes inherent in the video game industry (Taylor & Voorhees, 2018), especially in first-person shooter Public attitudes about sexual orientation have become stead- games (Voorhees & Orlando, 2018). Thus, we cannot rule ily more positive in most developed countries, and the iden- out that some users will negatively respond to the new gay- tification with the LGBTQ movement has increased in recent labeled character. years (Jones, 2021). Previous research analyzes how con- Over the last thirty years, several countries, including sumers react to CSR. Following the dominant stakeholder the United States, have experienced increasing support for theory in CSR (Freeman, 1984), this research is crucial for LGBTQ; however, there are multiple countries where the firms to understand the impact of their CSR decisions on majority disapprove of homosexuality (Adamczyk & Liao, key stakeholders. Positive associations of consumers with 2019). As our video game setting is global, we can also CSR in terms of awareness, information processing, trust, expect a negative response in demand to the new gay-labeled and ethical ideologies, anticipate a positive reaction (Deng character, especially in the short term as habituation is a & Xu, 2017; Palihawadana et al., 2016; Tian et al., 2011). slow process and shaping social norms takes time (Groves Based on the empirical evidence on LGBTQ-supportive & Thompson, 1970). Several studies also show that there corporate policies, which also reflects consensus on ben- remains a substantial amount of discrimination and anti-gay efits for companies that engage in such policies (Chintrakarn sentiment that research and society tend to underestimate et al., 2020; Hossain et al., 2020; Johnston & Malina, 2008; (Coffman et al., 2017; Dessel et al., 2017). Li & Nagar, 2013; Pichler et al., 2018; Wang & Schwarz, Additionally, the anonymity that characterizes com- 2010), it is possible that consumers will welcome the puter-mediated environments is likely to overcome social actions taken by the firms to increase diversity, e.g., through norms (Cirucci, 2013; Li & Li, 2014), foster negative atti- LGBTQ-themed products. More specifically, we hypothesize tudes toward LGBTQ-related campaigns and policies by that video game users will pick more often new LGBTQ- firms worldwide (Peterson, 2014), and negatively influence themed products to participate in the contest. We label this the inclusiveness hypothesis. 13 P. Parshakov et al. Fig. 2  Dynamics of the pick rate of Soldier:76 (right panel) and average pick rate for the rest of the characters (left panel). The dashed vertical line represents the date of the announcement (January 7). 95% confidence intervals are in gray consumer behavior. In our setting, several social media announcement. Our sample includes 1777 observations on users shared negative comments and concerns after Blizzard each Overwatch character for 63 points in time from Sep- decided to brand the Overwatch character Soldier:76 as gay.3 tember 2018 until May 2019. This precedent predicts that some video games users may Figure 2 shows the dynamics of the pick rate for Sol- also penalize LGBTQ-inclusive efforts and pick less often dier:76 and the average pick rate for all other characters. The new LBGTQ-themed products. We refer to this hypothesis average pick rates for all characters are stable across the time as the discrimination hypothesis. period. In contrast, we find a sharp decrease in pick rate of Soldier:76 after the announcement. To test whether there is a statistically significant decrease Data and Methods in pick rate after the announcement, we use a regression discontinuity design (RDD). RDD is a popular, quasi-exper- In Overwatch, players are assigned to two teams of six, imental technique in economics research (Angrist & Lavy, which aim to complete map-specific objectives within a 1999; Goldberger, 1972). We use this approach because we given period of time. Each player selects their character from cannot simply compare the treated character to the group of a roster of 32. Characters vary in their ability, appearance, other characters because the characters are probably different and personal details. from each other with respect to observable and unobservable A character’s popularity is reflected in its “pick rate”: how skills. This might potentially bias the estimates of another often it is chosen out of the total set of characters. To access method of treatment effects estimation. Another reason to pick rates, we rely on data provided by Overbuff.com, a con- use RDD is that the character’s in-game abilities did not sumer website that helps players improve. In addition to pick change at all during the analyzed time period. Therefore, the rates, Overbuff publishes tie and win rates per character. only reason for the change in pick rate is the user’s attitude Naturally, characters differ in their pick rates, presumably toward the new information regarding the gay background due to role, strength, ease of play, personal background, and of Soldier:76. so on (Kinkade & Lim, 2015). Additionally, we perform tests to trace consumer demand Because archival data are not available on the Overbuff dynamics for other Overwatch characters. We focus espe- webpage, we collect it via the Wayback Machine, a digital cially on the only other LGBTQ-branded character in the archive that provides data on characters’ historical pick rates. game, Tracer. Finally, we perform a survey among Over- The data are not saved regularly, so there are different inter- watch gamers to gain insights into the underlying mecha- vals between collection points. Fortunately, however, there nisms and explore the social significance of the demand are numerous observations close to the date of the Soldier:76 dynamics. The survey was posted on Reddit for 3 weeks and is retrospective.4 Nevertheless, the input is useful to assess the importance of characters’ stories, the controversy 3 In Twitter, for example, the announcement of Soldier:76’s homo- sexuality generated negative comments and reactions from multiple 4 users and received much attention, e.g., https://​twitt​er.​com/​CDawg​ This means that respondents are not likely to be involved in the VA/​status/​10833​31913​87042​6112?s=​20. demand dynamics that our empirical analysis captured back in 2019. 13 Do LGBTQ‑Supportive Corporate Policies Affect Consumer Behavior? Evidence from the Video… Fig. 3  Regression discontinuity design results for Soldier:76 (a) and for the placebo test (b). Pick rate is represented on Y-axis, time on the X-axis. The dashed vertical line represents the date of the announcement (January 7). 95% confidence intervals are presented by dashed lines around the announcement of Soldier:76’s homosexuality, the the average pick rate for a sample of non-LGBTQ charac- acceptance of LGBTQ-inclusive initiatives, and the interpre- ters with similar styles of play and skillsets. Panel (b) of tation of the results. The questions of the survey are avail- Fig. 3 shows the results. At the time of the announcement, able in the Appendix. the average pick rate of characters other than Soldier:76 was not affected. So, the decrease of the pick rate for Sol- dier:76 is due to the LGBTQ announcement. Results Figure 2 shows the recovery of Soldier:76’s pick rate after the announcement, and we can estimate when the The Effect of LGBTQ Branding on Consumer Demand pick rate returns to its previous level. To test this formally, we performed an RDD with grid search, skipping on each Figure 3 represents the results of the RDD. The vertical iteration an increasing number of days after the treatment line represents the date of the announcement. The other to determine when the effect is no longer statistically lines represent the dynamics of the pick rate for Soldier:76 significant. The analysis shows that starting from April (panel a) and its confidence intervals. We have 16 points in 11, the pick rate again has the same level as before the time before the treatment and 20 after the treatment. There announcement. Thus, it took about 3 months (94 days) for is clear evidence of a decrease in the pick rate. The local demand to return to its previous level. average treatment effect is 0.11 percentage points, which is 5.1% of the average pick rate for Soldier:76. We performed a series of robustness checks to test that the result is reliable. Underlying Mechanisms All show that the effect is statistically significant and varies from 0.10 to 0.19. To better understand consumer dynamics regarding gay This finding provides support for the discrimination products, we performed additional analyses. First, we hypothesis and suggests that anti-gay sentiment is still pre- focused on another LGBTQ-branded product in the same sent in society (Coffman et al., 2017; Dessel et al., 2017) game, the lesbian character Tracer. This character was and can influence consumer behavior. This negative attitude declared as LGBTQ long before the announcement about toward an LGBTQ-friendly policy contrasts with the posi- Soldier:76. We performed another RDD to understand tive perceptions of consumers to LGBTQ-inclusive equality whether the announcement regarding Soldier:76 affected policies, advertising, and imagery (Cunningham & Melton, the pick rate of Tracer. Figure 4 shows that the pick rate for 2014; Patel & Feng, 2021; Pounders & Mabry-Flynn, 2016). Tracer increased, and this effect is statistically significant. In this case, video game users decreased their demand Soldier:76 and Tracer are the same character type, so the for an LGBTQ-branded product by 5%. It should be noted, decrease in Soldier:76 could increase the pick rates for all however that the demand metric is relative and could be same-skill characters. However, we do not find such an affected by the increase in demand for other products effect for the other same-skill characters. (video game characters). We also performed a “placebo” Thus, the announcement regarding Soldier:76 positively test. We tested if the pick rate of the other characters was affected demand for the only other LGBTQ character. affected on this day to understand if there was a com- To some extent, this finding supports the inclusiveness mon shock that day not related to the announcement of hypothesis. However, our data have limitations that prevent Soldier:76 as gay. To do this, we performed an RDD for us from identifying and tracking individual users’ choices 13 P. Parshakov et al. Fig. 4  Regression discontinuity design results for Tracer. Pick rate is represented on Y-axis, time on the X-axis. The dashed vertical line represents the date of the announcement. (January 7). 95% confidence intervals are presented by dashed lines for characters. We know that after the LGBTQ-supportive as gay; therefore, it should be evaluated with caution. The announcement about Soldier:76, there were homophobic sample consists of 263 players. It includes a substantial comments and insults in the game live chat. Several users proportion of female players (16.3%); varies in age, with shared this fact on social media platforms like Twitter. 5 most respondents 18–24 (52.5%) who are mostly from North Therefore, one possibility is that some players did not America (49.4%) and Europe (39.5%), but also Oceania necessarily dislike the gay-branded product; rather, they (6.8%), Asia (3%), South America, and Africa.6 disliked something around the public announcement that In the sample, most (79.8%) players “sometimes” play drove unexpected attention to this character. This expla- with Soldier:76. Of these, 84.9% know that he is gay; among nation is in line with previous research that suggests that players who do not play with Soldier:76, 81.1% know that he consumers in the video game industry are supportive of is gay. Nearly all (98.5%) players know that Tracer is lesbian. LGBTQ-friendly policies (Jones, 2021). We asked players if they remember the controversy that One can assume that the demand recovery for Soldier:76 the announcement of Soldier:76’s homosexuality generated, was due to attracting new players to the game. Unfortunately, and how long it lasted. More than half (57%) remembered. there are no player-level data to investigate the composi- Among these, most thought the controversy lasted between tion of players. However, we can test whether there was a one or 2 weeks (52.3%) or 1 month (25.8%). This timeframe substantial change in Overwatch popularity using country- is in line with the temporary drop in the pick rate found in level data. For the keyword “Overwatch,” we collected a our analysis. cross-sectional dataset for the quarterly periods before and In an open-ended question, players were asked to provide after the announcement. The dataset represents the popu- further details about the controversy. Some stated that the larity of this search term by country. There are limitations controversy was “not a big deal,” “covered other controver- of such use of Google trends, as the two cross sections are sies,” or “was not a controversy at all.” Others shared stories both normalized to be distributed from 1 to 100 (Parshakov, about insults and hostility in social media and chats: “People 2021); that is, they indicate the relative popularity contrast to on Twitter got mad and called my friends and I homophobic absolute. The Spearman’s rank-order correlation coefficient slurs”; “An uptick in expressed homophobia either in the for the search term before and after the announcement is forums or in-game chat.” Still others commented more gen- 0.89 (p < 0.001), indicating that the popularity of the game erally on the “importance of backstories for characters” in did not change much. the Overwatch community, the possibility that it was just a To gain insight into the drivers of the consumers’ reac- “popularity move by Blizzard,” or it led to much “specula- tions, we performed a survey among Overwatch players. Fol- tion about other characters’ sexuality.” lowing Välisalo and Ruotsalainen (2019), we asked players Some respondents did not care much about the issue one on an Overwatch Reddit channel to participate. The survey way or the other, but others found the issue important. Some was performed three years after Soldier:76 was announced 27.4% of respondents said that they would be annoyed by comments in social media concerning the sexual orientation of a hero. Some respondents shared personal details about 5 See, for example, https://​twitt​er.​com/​Brend​enLive/​status/​10824​ 27372​29131​7760?s=​20; https://​twitt​er.​com/​benny___​bee/​status/​ 6 10823​82247​28880​7425?s=​20; or https://​twitt​er.​com/​NotTa​rzan6​66/​ The results of the survey are available from the authors upon status/​10824​27694​75376​3329?s=​20 request. 13 Do LGBTQ‑Supportive Corporate Policies Affect Consumer Behavior? Evidence from the Video… homophobic comments on live chats, forums, and social Causal analyses of consumers’ reactions to LGBTQ- media. inclusive practices are limited in the literature and a possible Companies that adopt LGBTQ-related policies risk pub- strand of novel research. Apart from observable consumer lic controversy and even hostility from some consumers. behavior toward a product, it is also important to investi- Therefore, it is important to know the social significance gate if there are positive changes in the attitude toward the for stakeholders of such initiatives. We also asked players if LGBTQ community. In other words, researchers should try they are in favor of Overwatch increasing sexual diversity to explore whether such CSR actions can promote social among characters. Some 81% were in favor, although many equity. shared concerns about the timing of the change as well as Future research could focus on the identification of con- the lack of updates and other forms of support. sumer segments and their preferences (Salvati et al., 2016, Finally, we provided participants with the results con- 2018). For example, previous studies show that heterosexual cerning the temporary drop of Soldier:76’s pick rate and women are more tolerant than men toward lesbians (Steffens, the increase of Tracer’s. Most players linked the drop in 2005) and gays (Um, 2014). Other demographic and behav- demand for Soldier:76 to homophobia: players didn’t want to ioral characteristics may be important for diversity branding. be associated with gay characters or were the target abusive In our setting, gay branding did not affect other features comments. The argument was that if players wanted to avoid of the character, such as visual appearance, style, and ease abuse, they would choose Tracer instead of Soldier:76. Some of play. In case of deeper incorporation of new features explained that “the two heroes [Tracer and Soldier:76] can into an existing product, consumers’ reactions could differ. be played similarly as a self-sufficient, high mobility flanker, Response may also vary by country. Further analyses incor- so it makes sense that people would switch from one hero to porating country-, product-, and industry-specific controls the other.” Several mentioned the different attitudes toward would be informative. gay and lesbian characters; for example, “Attractive lesbians are more tolerated than gays by hetero men.” The main finding of the present study is a negative short- Concluding Remarks term impact on a newly identified gay character. This is not in line with previous studies (Cunningham & Melton, 2014; LGBTQ-friendly companies engage in supportive initiatives Pounders & Mabry-Flynn, 2016). One explanation is that our that include, for example, policies to improve the working sample includes consumers from all over the world, unlike conditions of LGBTQ employees, advertising campaigns, studies that have used only US consumers (Adamczyk & and product branding. The present study focuses on the Liao, 2019). As the LGBTQ community is a significant part gaming industry to analyze how LGBTQ-supportive prod- of the global video game market, companies should, there- uct branding influences an important stakeholder—consum- fore, consciously use LGBTQ branding, but be ready to deal ers. We used a regression discontinuity design to exploit with possible short-term negative reactions. Blizzard Entertainment’s announcement that a character in From the company’s performance perspective, the main Overwatch was gay. finding here is the reversion of demand for the product after Our empirical analysis shows that the announcement 3 months. The reversion is probably because the spotlight resulted in a significant short-term drop in demand for this shifted to another topic. While short-term drops in demand character, which, however, returned to its previous level after the implementation of the LGBTQ-friendly policy after 3 months. The announcement coincided with a signifi- can negatively impact stakeholders’ interests, the long-term cant increase in demand for the game’s only other LGBTQ financial results may still show benefits (Chintrakarn et al., character. The lack of consumer-level data is a limitation to 2020; Johnston & Malina, 2008; Li & Nagar, 2013; Pichler further explain the mechanisms behind the drop in demand. et al., 2018; Wang & Schwarz, 2010). However, the survey findings open new avenues for research Two different explanations are possible for the return on consumer behavior and LGTBQ-supportive policies. to previous levels. First, a psychological explanation— Our study extends previous efforts to analyze the influ- “habituation”—would be that non-reinforced responses ence of LGBTQ-friendly policies on companies’ outcomes. decrease after several interactions with a stimulus (Groves Rather than a financial outcome, we analyzed pick rate, the & Thompson, 1970). In everyday language, consumers could behavior of a key stakeholder. The announcement of an have simply “gotten used” to the gay-branded character and LGBTQ-supportive policy had a negative impact on pick used it again. Second, a social explanation would be that rate that reverted to normal after 3 months. Public atten- consumers originally avoided the character due to online tion could play a major role in consumer behavior after the (Xu et al., 2012) and particularly gay (Gearhart & Zhang, implementation of LGBTQ-friendly policies. Research using 2014) bullying. If the harsh comments decreased over time, detailed demographic information and a panel structure consumers may have felt safe to return. could significantly contribute to understanding the influence 13 P. Parshakov et al. of LGBTQ-supportive policies on stakeholders’ behavior. Brown, T. J., & Dacin, P. A. (1997). The company and the product: We expect this issue to attract the interest of future research Corporate associations and consumer product responses. Jour- nal of Marketing, 61(1), 68–84. and drive-informed managerial decisions. Bursztyn, L., & Jensen, R. (2017). Social image and economic behavior in the field: Identifying, understanding, and shaping Supplementary Information The online version contains supplemen- social pressure. Annual Review of Economics, 9(1), 131–153. tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 10551-0​ 22-0​ 5137-7. Carnagey, N. L., Anderson, C. A., & Bushman, B. J. (2007). The effect of video game violence on physiological desensitization Acknowledgements Iuliia Naidenova and Petr Parshakov acknowledge to real-life violence. Journal of Experimental Social Psychol- the support of the Basic Research Program at the National Research ogy, 43(3), 489–496. University Higher School of Economics (HSE University). Cavaco, S., & Crifo, P. (2014). CSR and financial performance: Complementarity between environmental, social and business behaviours. Applied Economics, 46(27), 3323–3338. Declarations Cespa, G., & Cestone, G. (2007). Corporate social responsibility and managerial entrenchment. Journal of Economics and Manage- Conflict of interest There is no conflict of interest to report. ment Strategy, 16, 741–771. Chernev, A., & Blair, S. (2015). Doing well by doing good: The Research Involving Human Participants and/or Animals Not applicable. benevolent halo of corporate social responsibility. Journal of Consumer Research, 41(6), 1412–1425. Informed Consent Not applicable. Chintrakarn, P., Jiraporn, P., Kim, Y. S., & Kim, J. C. (2016). The effect of corporate governance on corporate social responsibil- ity. Asia-Pacific Journal of Financial Studies, 45, 102–123. Chintrakarn, P., Treepongkaruna, S., Jiraporn, P., & Lee, S. M. (2020). Do LGBT-supportive corporate policies improve credit References ratings? An instrumental-variable analysis. Journal of Business Ethics, 162, 31–45. Chu, M. (2019a). Thanks for all the messages about ‘Bastet’! Jack Adachi, P. J. C., & Willoughby, T. (2016). The longitudinal asso- and Vincent were in a romantic relationship many years ago. ciation between competitive video game play and aggression Both identify as gay, Twitter, January 8. Available at: https://​ among adolescents and young adults. Child Development, 87(6), twitt​er.​com/​westo​fhouse/​status/​10823​94625​83742​8736. 1877–1892. Chu, M. (2019b). Bastet. Available at: https://​bnetc​msus-a.​akama​ Adamczyk, A., & Liao, Y. C. (2019). Examining public opinion about ihd.​net/​cms/​galle​r y/​rk/​RK25W​EUAJI​PO154​66497​30105.​pdf. LGBTQ-related issues in the United States and across multiple Cirucci, A. M. (2013). First person paparazzi: Why social media nations. Annual Review of Sociology, 45(1), 401–423. should be studied more like video games. Telematics and Infor- Ailawadi, K. L., Neslin, S. A., Luan, Y. J., & Taylor, G. A. (2014). matics, 30(1), 47–59. Does retailer CSR enhance behavioral loyalty? A case for benefit Clermont, B. (2006). Companies stand to benefit from pro-LGBT segmentation. International Journal of Research in Marketing, policies. Insight Business. 31(2), 156–167. Coffman, K. B., Coffman, L. C., & Ericson, K. M. M. (2017). The Angrist, J. D., & Lavy, V. (1999). Using Maimonides’ rule to estimate size of the LGBT population and the magnitude of antigay the effect of class size on scholastic achievement. Quarterly sentiment are substantially underestimated. Management Sci- Journal of Economics, 114(2), 533–575. ence, 63(10), 3168–3186. Anteby, M., & Anderson, C. (2014). The shifting landscape of LGBT Colgan, F., Creegan, C., McKearney, A., & Wright, T. (2007). organizational research. Research in Organization Behavior, 34, Equality and diversity policies and practices at work: Lesbian, 3–25. LGBT, and bisexual workers. Equal Opportunities Interna- Barauskaite, G., & Streimikiene, D. (2020). Corporate social respon- tional, 26, 590–609. sibility and financial performance of companies: The puzzle of Creed, W. D., Scully, M. A., & Austin, J. R. (2002). Clothes make the concepts, definitions and assessment methods. Corporate Social person? The tailoring of legitimating accounts and the social Responsibility and Environmental Management, 28(1), 278–287. construction of identity. Organization Science, 13, 475–496. Barnea, A., & Rubin, A. (2010). Corporate social responsibility as a Cunningham, G. B., & Nite, C. (2020). LGBT diversity and inclu- conflict between shareholders. Journal of Business Ethics, 97, sion, community characteristics, and success. Journal of Sport 71–86. Management, 34, 533–541. Behm-Morawitz, E., & Mastro, D. (2009). The effects of the sexualiza- Cunningham, G. B., & Melton, E. N. (2014). Signals and cues: tion of female video game characters on gender stereotyping and LGBT inclusive advertising and consumer attraction. Sport female self-concept. Sex Roles, 61(11), 808–823. Marketing Quarterly, 23(1), 37–46. Berens, G., Van Riel, C. B., & Van Rekom, J. (2007). The CSR-quality Day, N. E., & Greene, P. G. (2008). A case for sexual orientation trade-off: When can corporate social responsibility and corpo- diversity management in small and large organizations. Human rate ability compensate each other? Journal of Business Ethics, Resource Management, 47, 637–654. 74(3), 233–252. Deng, X., & Xu, Y. (2017). Consumers’ responses to corporate social Bertrand, M., & Duflo, E. (2017). Field experiments on discrimination. responsibility initiatives: The mediating role of consumer– In A. V. Banerjee & E. Duflo (Eds.), Handbook of field experi- company identification. Journal of Business Ethics, 142(3), ments (Vol. 1, pp. 309–393). North-Holland Publishing. 515–526. Bowen, H. R. (1953). Social responsibilities of the businessman. Dessel, A. B., Goodman, K. D., & Woodford, M. R. (2017). LGBT Harper & Row. discrimination on campus and heterosexual bystanders: Under- Brahma, S., Gavriilidis, K., Kallinterakis, V., Verousis, T. and Zhang, standing intentions to intervene. Journal of Diversity in Higher M. (2021). LGBTQ and finance. Available at: https://p​ apers.s​ srn.​ Education, 10(2), 101–116. com/​sol3/​papers.​cfm?​abstr​act_​id=​39844​15. 13 Do LGBTQ‑Supportive Corporate Policies Affect Consumer Behavior? Evidence from the Video… Dill, K. E., & Dill, J. C. (1998). Video game violence: A review Kinkade, N. & Lim, K. (2015). Dota 2 win prediction. Available at: of the empirical literature. Aggression and Violent Behavior, http://​jmcau​ley.​ucsd.​edu/​cse258/​proje​cts/​fa15/​018.​pdf. 3(4), 407–428. Lacey, R., Kennett-Hensel, P. A., & Manolis, C. (2015). Is corporate ESA. (2021). Essential facts about the video game industry. Avail- social responsibility a motivator or hygiene factor? Insights into able at: https://​www.​theesa.​com/​resou​rce/​2021-​essen​tial-​facts-​ its bivalent nature. Journal of the Academy of Marketing Science, about-​the-​video-​game-​indus​try/. 43(3), 315–332. Freeman, R. E. (1984). Strategic management: A stakeholder Lee, K. H., & Shin, D. (2010). Consumers’ responses to CSR activities: approach. Pitman. The linkage between increased awareness and purchase intention. Friederichsen, P. (2019). Guide to LGBTQ brand marketing. Brand- Public Relations Review, 36(2), 193–195. ing Strategy Insider. Available at: https://​w ww.​b rand​i ngst​ Li, F., & Nagar, V. (2013). Diversity and performance. Management rateg​yinsi​der.​com/​guide-​to-​lgbtq-​brand-​marke​ting/#.​YMeJk​ Science, 59(3), 529–544. vn7QuU. Li, Z., & Li, C. (2014). Twitter as a social actor: How consumers evalu- Friedman, M. (1970). The social responsibility of business is to ate brands differently on Twitter based on relationship norms. increase its profits. The New York times Magazine, 13, 122–126. Computers in Human Behavior, 39, 187–196. Gearhart, S., & Zhang, W. (2014). Gay bullying and online opinion McCarthy, J. (2021). Euro 2020 sponsors on lighting up stadiums expression: Testing spiral of silence in the social media environ- with Pride ads despite UEFA’s muddled messaging. The Drum. ment. Social Science Computer Review, 32(1), 18–36. Available at: https://​www.​thedr​um.​com/​news/​2021/​06/​29/​euro-​ Gestos, M., Smith-Merry, J., & Campbell, A. (2018). Representation of 2020-​spons​ors-​light​ing-​up-​stadi​ums-​with-​pride-​ads-​despi​te-​ women in video games: A systematic review of literature in con- uefas-​muddl​ed. sideration of adult female wellbeing. Cyberpsychology, Behavior, McWilliams, A., & Siegel, D. (2001). Corporate social responsibil- and Social Networking, 21(9), 535–541. ity: A theory of the firm perspective. Academy of Management Goldberger, A. S. (1972). Selection bias in evaluating treatment effects: Review, 26(1), 117–127. Some formal illustrations. University of Wisconsin-Madison. Metcalf, H., & Rolfe, H. (2011). Barriers to employees in developing Groves, P. M., & Thompson, R. F. (1970). Habituation: A dual-process lesbian, LGBT, bisexual and transgender-friendly workplaces. theory. Psychological Review, 77(5), 419–450. National Institute of Economic and Social Research. Halbrook, Y. J., O’Donnell, A. T., & Msetfi, R. M. (2019). When and Oh, W. Y., Chang, Y. K., & Martynov, A. (2011). The effect of own- how video games can be good: A review of the positive effects ership structure on corporate social responsibility: Empirical of video games on well-being. Perspectives on Psychological evidence from Korea. Journal of Business Ethics, 104, 283–297. Science, 14(6), 1096–1104. Onkila, T., & Sarna, B. (2021). A systematic literature review on Haerpfer, C., Inglehart, R., Moreno, A., Welzel, C., Kizilova, K., Diez- employee relations with CSR: State of art and future research Medrano, J., Lagos, M., Norris, P., Ponarin, E., Puranen, B., agenda. Corporate Social Responsibility and Environmental et al. (2020). World values survey: Round seven—country pooled Management. https://​doi.​org/​10.​1002/​csr.​2210 datafile. JD Systems Institute. Osakwe, C. N., & Yusuf, T. O. (2020). CSR: A roadmap towards cus- Hillman, A. J., & Keim, G. D. (2001). Shareholder value, stakeholder tomer loyalty. Total Quality Management & Business Excellence, management, and social issues: What’s the bottom line? Strategic 32(13), 1424–1440. Management Journal, 22, 125–139. Palihawadana, D., Oghazi, P., & Liu, Y. (2016). Effects of ethical ide- Holland, L., Matthews, T. L., & Schott, M. R. (2013). “That’s so gay!” ologies and perceptions of CSR on consumer behavior. Journal Exploring college students’ attitudes toward the LGBT popula- of Business Research, 69(11), 4964–4969. tion. Journal of Homosexuality, 60(4), 575–595. Parshakov, P. (2021). How to use Google Trends data for empirical Hossain, M., Atif, M., Ahmed, A., & Mia, L. (2020). Do LGBT work- research: A short note. Available at SSRN: https://s​ srn.c​ om/a​ bstr​ place diversity policies create value for firms? Journal of Busi- act=​39432​87. ness Ethics, 167(4), 775–791. Patel, P. C., & Feng, C. (2021). LGBT Workplace equality policy Inoue, Y., Funk, D. C., & McDonald, H. (2017). Predicting behavio- and customer satisfaction: The roles of marketing capability ral loyalty through corporate social responsibility: The mediat- and demand instability. Journal of Public Policy & Marketing, ing role of involvement and commitment. Journal of Business 40(1), 7–26. Research, 75, 46–56. Paul, A. K., McElroy, T., & Leatherberry, T. (2011). Diversity as an Iwata, E. (2006). More marketing aimed at LGBT consumers. USA engine of innovation. Deloitte Review, 8, 109–121. Today. Peterson, E. H. (2014). People are furious over Burger King’s gay pride Johnston, D., & Malina, M. A. (2008). Managing sexual orientation whopper. Business Insider. Available at: https://​www.​busin​essin​ diversity: The impact on firm value. Group & Organization Man- sider.​com/​burger-​kings-​gay-​pride-​whopp​er-​2014-7?​r=​USR=T. agement, 33(5), 602–625. Pew Research Center. (2013). A Survey of LGBT Americans. Atti- Jones, J. M. (2021). LGBT identification rises to 5.6% in latest U.S. tudes, experiences and values in changing times. Available at: estimate. Gallup. https://​www.​pewso​cialt​rends.​org/​2013/​06/​13/a-​survey-​of-​lgbt-​ Jones, K. N., Brewster, M. E., & Jones, J. A. (2014). The creation ameri​cans/. and validation of the LGBT ally identity measure. Psychology of Pichler, S., Blazovich, J., Cook, K., Huston, J., & Strawser, W. (2018). Sexual Orientation and Gender Diversity, 1(2), 181–195. Do LGBT-supportive corporate policies enhance firm perfor- Joo, Y. H., Kim, Y., & Yang, S. J. (2011). Valuing customers for mance? Human Resource Management, 57, 263–278. social network services. Journal of Business Research, 64(11), Podnar, K., & Javernik, P. (2012). The effect of word of mouth on con- 1239–1244. sumers’ attitudes toward products and their purchase probability. Kaplan, D. M. (2006). Can diversity training discriminate? Backlash Journal of Promotion Management, 18(2), 145–168. to lesbian, LGBT, and bisexual diversity initiatives. Employee Pounders, K., & Mabry-Flynn, A. (2016). Consumer response to gay Responsibilities and Rights Journal, 18, 61–72. and lesbian imagery: How product type and stereotypes affect King, E. B., & Cortina, J. M. (2010). The social and economic impera- consumers’ perceptions. Journal of Advertising Research, 56(4), tive of lesbian, LGBT, bisexual, and transgendered supportive 426–440. organizational policies. Industrial and Organizational Psychol- Ragins, B. R., & Cornwell, J. M. (2001). Pink triangles: Antecedents ogy, 3, 69–78. and consequences of perceived workplace discrimination against 13 P. Parshakov et al. LGBT and lesbian employees. Journal of Applied Psychology, Vallaster, C., Lindgreen, A., & Maon, F. (2012). Strategically leverag- 86, 1244–1261. ing corporate social responsibility: A corporate branding per- Ragins, B. R., Singh, R., & Cornwell, J. M. (2007). Making the invis- spective. California Management Review, 54(3), 34–60. ible visible: Fear and disclosure of sexual orientation at work. Van Rooij, A. J., Meerkerk, G. J., Schoenmakers, T. M., Griffiths, M., Journal of Applied Psychology, 92, 1103–1118. & van de Mheen, D. (2010). Video game addiction and social Salvati, M., Ioverno, S., Giacomantonio, M., & Baiocco, R. (2016). responsibility. Addiction Research & Theory, 18(5), 489–493. Attitude toward gay men in an Italian sample: Masculinity and Venter, E. (2021). Multinational corporate social responsibility and sexual orientation make a difference. Sexuality Research and diversity: Blizzard Entertainment’s Overwatch, the Overwatch Social Policy, 13(2), 109–118. league and LGBTQ pride month. In A. L. Hutchins & N. T. Tin- Salvati, M., Pistella, J., Ioverno, S., Giacomantonio, M., & Baiocco, dall (Eds.), Public relations and online engagement (pp. 52–63). R. (2018). Attitude of Italian gay men and Italian lesbian women Routledge. towards gay and lesbian gender-typed scenarios. Sexuality Voorhees, G., & Orlando, A. (2018). Performing neoliberal masculin- Research and Social Policy, 15(3), 312–328. ity: Reconfiguring hegemonic masculinity in professional gam- Sen, S., & Bhattacharya, C. B. (2001). Does doing good always lead to ing. In N. Taylor & G. Voorhees (Eds.), Masculinities in play (pp. doing better? Consumer reactions to corporate social responsibil- 211–227). Palgrave Macmillan. ity. Journal of Marketing Research, 38(2), 225–243. Walsh, G., & Bartikowski, B. (2013). Exploring corporate ability and Stanaland, A. J., Lwin, M. O., & Murphy, P. E. (2011). Consumer social responsibility associations as antecedents of customer sat- perceptions of the antecedents and consequences of corporate isfaction cross-culturally. Journal of Business Research, 66(8), social responsibility. Journal of Business Ethics, 102(1), 47–55. 989–995. Steffens, M. C. (2005). Implicit and explicit attitudes towards lesbians Wang, H., Gibson, C., & Zander, U. (2020). Is research on corporate and gay men. Journal of Homosexuality, 49(2), 39–66. social responsibility undertheorized? Academy of Management Stonewall. (2021). Stonewall top 100 employers. London: Stonewall. Review, 45(1), 1–6. Tang, Z., Hull, C. E., & Rothenberg, S. (2012). How corporate social Wang, P., & Schwarz, J. L. (2010). Stock price reactions to GLBT non- responsibility engagement strategy moderates the CSR-financial discrimination policies. Human Resource Management, 49(2), performance relationship. Journal of Management Studies, 49(7), 195–216. 1274–1303. Ward, M. R., & Lee, M. J. (2000). Internet shopping, consumer search Taylor, N., & Voorhees, G. (2018). Introduction: Masculinity and gam- and product branding. Journal of Product & Brand Management, ing: Mediated masculinities in play. In N. Taylor & G. Voorhees 9(1), 6–20. (Eds.), Masculinities in play (pp. 1–19). Palgrave Macmillan. Webb, K. (2019). “Overwatch” hero Soldier 76 is gay and fans love it. Theodorakopoulos, N., & Budhwar, P. (2015). Diversity and inclusion Business Insider. Available at: https://w ​ ww.b​ usine​ ssins​ ider.c​ om/​ in different work settings: Emerging patterns, challenges, and soldi​er-​76-​gay-​queer-​overw​atch-​2019-1. research agenda. Human Resource Management, 54, 177–197. Xu, J. M., Jun, K. S., Zhu, X. & Bellmore, A. (2012). Learning from Tian, Z., Wang, R., & Yang, W. (2011). Consumer responses to cor- bullying traces in social media. In E. Fosler-Lussier, E. Riloff, & porate social responsibility (CSR) in China. Journal of Business S. Bangalore (eds.), Proceedings of the 2012 Conference of the Ethics, 101(2), 197–212. North American chapter of the Association for Computational Um, N. H. (2014). Does gay-themed advertising haunt your brand? The Linguistics: Human language technologies (pp. 656–666). Asso- impact of gay-themed advertising on young heterosexual con- ciation for Computational Linguistics. sumers. International Journal of Advertising, 33(4), 811–832. Zhu, J. Y., & Smieliauskas, W. (2021). Evidence on the economic con- Valenti, C. (2012). Companies begin marketing to LGBT market. ABC sequences of marriage equality and LGBT human rights. Journal News. of Business Ethics, 11, 1–14. Välisalo, T. & Ruotsalainen, M. (2019). “I never gave up”: Engagement with playable characters and esports players of Overwatch. In Publisher's Note Springer Nature remains neutral with regard to S. Deterding et al. (eds.), Proceedings of the 14th International jurisdictional claims in published maps and institutional affiliations. Conference on the Foundations of Digital Games (pp. 1–6). Association for Computing Machinery. 13