MAB-scriptieprijs |
Academic editor: Henk Volberda
© 2024 Kim van Schie, Susanne Preuss.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits to copy and distribute the article for non-commercial purposes, provided that the article is not altered or modified and the original author and source are credited.
Citation:
van Schie K, Preuss S (2024) How do consumers react to LGBTQ+ activism? Evidence from mobile phone geolocation data. Maandblad voor Accountancy en Bedrijfseconomie 98(3): 89-101. https://doi.org/10.5117/mab.98.109769
|
Corporate activism is an important tool to bring attention to societal issues. While it can have many benefits, it also comes with risks, such as alienating a relevant share of a firm’s consumers. In this study, we use mobile phone geolocation data to examine how firms’ LGBTQ+ activism influences consumer behavior. Our results suggest that LGBTQ+ activism decreases consumer store visits in the short term. The effect occurs in both liberal and conservative counties. This is surprising, as liberals generally react favorably to LGBTQ+ activism. One possible explanation is that (liberal) consumers reduce their store visits due to pink washing concerns.
Corporate activism, Disclosure, ESG, LGBTQ+, Twitter, Polarization
Firms are increasingly pressured to take stances on controversial, sociopolitical, topics. Our findings (differences in store visits around LGBTQ+ activism events) are important for firms that consider speaking out on such matters. They are also relevant for investors and auditors because they indicate that sociopolitical engagement needs to be considered when assessing company risks.
Firms are increasingly pressured to speak out on controversial, sociopolitical, topics (
In this study, we examine how LGBTQ+ activism influences consumer behavior. Specifically, we use mobile phone geolocation (‘foot traffic’) data provided by SafeGraph Inc. and examine changes in store visits around LGTBQ+ activism events taking place between 2018 and 2021. To identify LGBTQ+ activism events, we manually search press releases and tweets for statements on LGBTQ+ matters. We focus on the first time a firm takes a stance on LGBTQ+ matters, as first-time activism may have the greatest effect on consumer behavior. Our final sample consists of 70 SafeGraph firms, of which 14 had LGBTQ+ activism events during our sample time frame. After having identified activism events, we use a difference-in-differences research design to compare changes in visits at stores belonging to activist firms to changes in visits at stores belonging to non-activist firms.
LGBTQ+ activism has the potential to both increase and decrease consumer store visits in the short-term. On the one hand, taking stances on LGBTQ+ matters is important to attract younger, more liberal consumers (
Our study presents several important findings: First, only about one-fourth of firms with foot traffic data and Twitter accounts took stances on LGBTQ+ matters between 2010 and 2021, supporting the idea that firms consider LGBTQ+ activism risky (Wettstein and Baur 2016). Second, consumers decrease their store visits at activist firms. To be precise, we find that store visits at activist firms decrease by 2.5% relative to store visits at non-activist firms. Interestingly, we observe an effect at grocery stores and restaurants, but not at gasoline stations. One reason may be that LGTBQ+ activism is easier to observe at stores that can integrate LGBTQ+ messages in their products, for example, by selling LGBTQ+ themed foods or merchandise. Third, we detect a similar relationship between LGBTQ+ activism and consumer visits in (highly) conservative and (highly) liberal counties. This latter finding suggests that the negative effect of LGBTQ+ activism is not exclusively driven by conservative consumers who disapprove of a firm’s stance, but may also be driven by liberal consumers who perceive the stance as dishonest. Overall, our results illustrate that sociopolitical activism is a complex issue that requires firms to carefully consider their consumer base and the costs and benefits that come with it.
This study contributes to the literature in several ways. First, it contributes to the literature on the effects of sociopolitical claims made by firms. While there is some evidence on firm value effects of sociopolitical activities (e.g.,
Second, it adds to the literature on the mismatch between firms’ statements and their actions, commonly referred to as ‘cheap talk’. We find negative effects of LGBTQ+ claims even in liberal counties, which could have to do with stances being perceived as hypocritical. Prior studies caution that firms may face reputational costs from cheap talk (e.g.,
Finally, sociopolitical activism is an emerging topic in the field of accounting (e.g.,
In a 1970 New York Times article, Milton Friedman famously wrote that “the social responsibility of business is to increase its profits” (
The idea that firms use social responsibility to maximize shareholder value still holds today (e.g.,
Because activism poses reputational risk, only a few managers speak out on sociopolitical issues (
A growing number of academic papers study firm value consequences of CEO or firm sociopolitical activism. Benedo and Siming (2020), for instance, examine abnormal returns around CEO resignations from Donald Trump’s presidential advisory council and detect losses of -0.57% around the resignation announcements.
Even though some of the results seem conflicting (e.g.,
Also, the credibility of the stances matters. Investors may not react to activities that they do not believe to be true (
This study focuses on changes in consumer behavior around activism events supporting the LGBTQ+ community. This includes (announcements of) selling LGBTQ+ merchandise, celebrating Pride Month, providing an LGBTQ+ inclusive work environment, or opposing anti-LGBTQ+ legislation.
As outlined above, sociopolitical activities such LGBTQ+ activism can have both positive and negative consequences for firms. On the one hand, firms may benefit from speaking out on LGBTQ+ matters, especially when it comes to attracting younger individuals who are more likely to place themselves on the LGBTQ+ spectrum (
Examples of rainbow-washing criticism. Source Panel A (left): https://www.instagram.com/p/CsmPtaTr37w/; Source Panel B (right): https://twitter.com/vexwerewolf/status/1400785796651700230.
That said, we also note that there is some uncertainty whether boycott calls are effective (e.g.,
H1: Consumers do not react to corporate LGBTQ+ activism
We combine data from several sources: We use SafeGraph Inc. to obtain data on consumer store visits, simplemaps.com to link SafeGraph zip codes to counties, and FactSet and Twitter to identify corporate LGBTQ+ activism. In addition, we use data from the MIT election lab to explore differences in effects depending on whether stores are located in conservative or liberal counties.
SafeGraph Inc. provides daily store visits (foot traffic) based on anonymized smartphone geolocation data. These data come from smartphone users who agreed to sharing their locations with location-tracking applications such as mapping or weather services (
We apply several sample selection steps summarized in Table
Sample selection. This table presents the sample selection. It includes the number of firms, brands, stores, and observations. We have one observation per store and year-month. Note that the large difference in brands between steps (1) and (2) is due to stores that do not belong to any brand. Each of these stores is counted as an individual brand.
(1) Firms | (2) Brands | (3) Stores | (4) Obs. | |
---|---|---|---|---|
Panel A: Safegraph observations | ||||
(1) Safegraph observations 01/2018 – 11/2021 | 400,030 | 905,584 | 38,755,100 | |
(2) Excluding stores without stock symbol | 95 | 184 | 196,534 | 8,580,900 |
(3) Excluding stores without counties | 95 | 184 | 186,487 | 8,275,122 |
(4) Excluding stores that are not covered throughout the full sample timeframe | 94 | 180 | 163,584 | 7,688,448 |
Panel B: Store visits and LGBTQ+ activism | ||||
(5) Excluding firms without Twitter accounts | 77 | 130 | 130,886 | 6,151,642 |
(6) Excluding firms with activism events prior to 01/2018 | 70 | 113 | 89,270 | 4,195,690 |
Panel C: Final sample | ||||
(7) Excluding year-months with missing values on Treatment for a three-month event window | 70 | 113 | 89,270 | 2,284,639 |
We define LGBTQ+ activism as stances taken by firms supporting the LGBTQ+ community.
We search for activism events at the parent firm level. 26 of the 94 SafeGraph sample firms operate more than one brand. For instance, Yum! Brands Inc. operates KFC, Pizza Hut, and Taco Bell. Amazon.com Inc. operates Amazon Fresh, Amazon Go, and Whole Foods Market. Darden Restaurants Inc. and Kroger Co. operate eight restaurants and 21 grocery store brands, respectively. The choice to search for activism events at the parent level assumes that parent-firm stances are reciprocated by the individual brands. However, even if individual brands do not take stances themselves, association with the parent firms’ stances may still affect them.
After manually collecting firms’ Twitter handles, we add two more sample selection steps, summarized in Table
Summary statistics. This table presents the summary statistics. Ln(visits) and Ln(visitors) are truncated at the 1st and 99th percentile.
Observations | Mean | Median | Std. Dev. | Min | Max | |
---|---|---|---|---|---|---|
Panel A: Treatment group | ||||||
Ln(visits) | 2,190,717 | 5.50 | 5.56 | 1.03 | 2.56 | 7.82 |
Ln(visitors) | 2,190,717 | 5.11 | 5.19 | 1.02 | 2.30 | 7.32 |
Treatment | 279,666 | 0.50 | 0.50 | 0.50 | 0.00 | 1.00 |
Q1 | 2,187,793 | 0.04 | 0.00 | 0.19 | 0.00 | 1.00 |
Q2 | 2,187,793 | 0.28 | 0.00 | 0.45 | 0.00 | 1.00 |
Q3 | 2,187,793 | 0.40 | 0.00 | 0.49 | 0.00 | 1.00 |
Q4 | 2,187,793 | 0.25 | 0.00 | 0.43 | 0.00 | 1.00 |
Q5 | 2,187,793 | 0.03 | 0.00 | 0.16 | 0.00 | 1.00 |
Panel B: Control group | ||||||
Ln(visits) | 2,004,973 | 5.71 | 5.83 | 1.00 | 2.56 | 7.82 |
Ln(visitors) | 2,004,973 | 5.34 | 5.46 | 0.99 | 2.30 | 7.32 |
Treatment | 2,004,973 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
Q1 | 2,002,112 | 0.04 | 0.00 | 0.20 | 0.00 | 1.00 |
Q2 | 2,002,112 | 0.30 | 0.00 | 0.46 | 0.00 | 1.00 |
Q3 | 2,002,112 | 0.42 | 0.00 | 0.49 | 0.00 | 1.00 |
Q4 | 2,002,112 | 0.22 | 0.00 | 0.41 | 0.00 | 1.00 |
Q5 | 2,002,112 | 0.02 | 0.00 | 0.14 | 0.00 | 1.00 |
Most activism events fall in June, which is to be expected as June is dedicated to celebrating LGBTQ+ pride in the U.S. With respect to the summary statistics, we observe slightly lower foot traffic at activist firms. Treatment (control) group stores receive on average 408 (468) visits and 267 (322) visitors per month, respectively (values in Table
Note that we do not expect a causal link between firm disclosures and consumer visits. Instead, tweets and press releases may just be a signal of general LGBTQ+ activism that can be observed more directly by consumers. Shake Shack, for instance, tweeted about serving a ‘pride shake’. The shake, which was sold in the restaurants, was likely more visible than the tweet. In other words, a difference in store visits at Shake Shack locations may be caused by consumers noticing the shake, rather than by consumers reading the tweet.
We use a difference-in-differences design to test the effect of LGBTQ+ activism on consumer store visits. We estimate the following OLS regression model, where i reflects parent firms, j reflects brands, k reflects store locations, and t reflects year-months:
Ln (store visits)ijkt = α + β1 Treatmentit + ϵijkt
The dependent variable, Ln (store visits) is the natural logarithm of the number of store visits at a brand’s local store in a given year-month. The independent variable Treatment is equal to one in the three months following a parent firm’s LGBTQ+ activism event (the first month is the event month), zero in the three months prior to the event, and missing otherwise. The control group, for which Treatment is always zero, consists of firms that did not engage in LGBTQ+ activism events before or during our sample time frame.
Variable overview. This table presents the variables used. Ln(visits) and Ln(visitors) are truncated at the 1st and 99th percentile.
Variable | Description | Source |
---|---|---|
Treatment | Dummy variable equal to one in the three months following an activism event and zero in the three months preceding an activism event | Self-collected via Twitter and FactSet |
Ln(visits) | Natural logarithm of one plus the number of store visits per month | SafeGraph |
Ln(visitors) | Natural logarithm of one plus the number of store visitors per month | SafeGraph |
Q1 | Dummy variable equal to one if the percentage of votes to the republican presidential candidate in a given county is below 20% | MIT Election Lab |
Q2 | Dummy variable equal to one if the percentage of votes to the republican presidential candidate in a given county is greater or equal to 20% and below 40% | MIT Election Lab |
Q3 | Dummy variable equal to one if the percentage of votes to the republican presidential candidate in a given county is greater or equal to 40% and below 60% | MIT Election Lab |
Q4 | Dummy variable equal to one if the percentage of votes to the republican presidential candidate in a given county is greater or equal to 60% and below 80% | MIT Election Lab |
Q5 | Dummy variable equal to one if the percentage of votes to the republican presidential candidate in a given county is greater than or equal to 80% | MIT Election Lab |
Research design. This figure illustrates our regression model using activism events from Kroger Co and Yum Brands! Inc. Our variable Treatment is equal to one at store locations in the three months following a firm’s LGBTQ+ activism event, zero in the three months before the event, and missing otherwise. Note that September 2019 has missing values on Treatment, which means it will be omitted from the regressions. Our model and this figure are based on
In addition, we add store and county-year-month fixed effects. Store fixed effects estimate coefficients at the store-location-level and account for time-invariant store characteristics such as stores’ local popularity. Year-month fixed effects account for time-specific differences in consumer visits. Consumers may, for instance, visit stores less in colder months. By using county-year-month fixed effects, we additionally account for differences in consumer visits in the same year-months in the same county. Such differences include county-specific weather events and local economic wealth (
Our main results are presented in Table
Effects of LGBTQ+ activism on store visits. This table presents the results of regressing Ln(visits) on Treatment. For the treatment group, Treatment is equal to one in the three months after an activism event, zero in the three months before the event, and missing otherwise. For the control group, Treatment is always zero. Standard errors are clustered by county. t-statistics are in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% level, respectively.
Dependent variable | Ln(visits) | |||
---|---|---|---|---|
(1) | (2) | (3) | (4) | |
Store type | All stores | Gasoline stations | Grocery stores | Restaurants |
Treatment | -0.025*** | 0.000 | -0.109*** | -0.011*** |
(-15.41) | (0.04) | (-11.86) | (-4.82) | |
Observations | 2,266,121 | 689,060 | 494,202 | 1,036,512 |
Location FE | Yes | Yes | Yes | Yes |
County-year-month FE | Yes | Yes | Yes | Yes |
Adjusted R2 | 0.892 | 0.914 | 0.908 | 0.859 |
Our base results in column (1) present the effects of comparing LGBTQ+ activist parent firms to all other non-activist control group firms. In columns (2) to (4), we re-estimate effects within the specific store types (Hou and Polquin 2023;
An explanation for the insignificant coefficient at gasoline stations may be that activism is less observable at those stores. In addition to announcing their support, grocery stores and restaurants can offer LGBTQ+ themed products such as Shake Shak’s ‘Pride Shake’ mentioned in Section 3.2. Besides, boycotts may be less effective at gasoline stations, where price and location are the main drivers of consumption choices (
We run several robustness tests. First, we change the event window for Treatment to one and six months, respectively. Second, we use the natural logarithm of the number of individual visitors, Ln(visitors), as an alternative dependent variable. Third, to rule out that our results are driven by one specific activist parent firm, we rerun the test in Table
Stacked events. This figure presents the coefficient estimates from regressions of ln(visits) on Treatment and county-year-month fixed effects in a stacked events dataset. Treatment is equal to one if the firm has an LGBTQ+ activism event during our sample time frame, and zero otherwise. Month zero is the event month. Standard errors are clustered by county.
We find that corporate LGBTQ+ activism is associated with reduced store visits. As described in Section 3.2, this negative effect may be caused by two distinctly different consumer groups: conservative consumers who may boycott firms because they disapprove of LGBTQ+ activism, or liberal consumers who may perceive the stances as hypocritical. To shed more light on the source of the effect, we examine differences in consumers’ political ideologies. If the decrease in store visits is predominantly caused by fewer conservative consumers, we would expect a stronger effect in conservative counties.
To proxy for ideology, we use data on presidential election results by county provided by the MIT Election Lab.
Differences in Democratic and Republican counties. This table presents the results of regressing Ln(visits) on Treatment and on the interaction of Treatment with Qi. For the treatment group, Treatment is equal to one in the three months after an activism event, zero in the three months prior to the event, and missing otherwise. For the control group, Treatment is always zero. Qi represents different cut-offs for county-level election results. A higher value on Qi indicates greater support for the Republican presidential candidates. The main effects of Qi are absorbed by county-year-month fixed effects. Standard errors are clustered by county. t-statistics are in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% level, respectively.
Dependent variable | Ln(visits) | |||
---|---|---|---|---|
(1) | (2) | (3) | (4) | |
Store type | All stores | Gasoline stations | Grocery stores | Restaurants |
Treatment | -0.023*** | -0.013 | -0.078*** | -0.012 |
(-4.00) | (-1.16) | (-3.27) | (-1.05) | |
Treatment × Q2 | -0.002 | 0.019 | -0.016 | -0.006 |
(-0.24) | (1.54) | (-0.57) | (-0.48) | |
Treatment × Q3 | -0.001 | 0.020* | -0.036 | 0.003 |
(-0.08) | (1.68) | (-1.28) | (0.25) | |
Treatment × Q4 | -0.005 | -0.003 | -0.099*** | 0.006 |
(-0.72) | (-0.26) | (-2.99) | (0.53) | |
Treatment × Q5 | 0.005 | 0.008 | -0.027 | 0.007 |
(0.49) | (0.45) | (-0.58) | (0.35) | |
Observations | 2,263,799 | 688,383 | 493,180 | 1,035,945 |
Location FE | Yes | Yes | Yes | Yes |
County-year-month FE | Yes | Yes | Yes | Yes |
Adjusted R2 | 0.892 | 0.914 | 0.908 | 0.859 |
Note that because we now use interactions of Treatment with Qi, the interpretation of Treatment differs: Treatment presents the effect of LGBTQ+ activism in the highly liberal Q1 counties. The interaction terms Treatment × Qi present the differences in post-activism store visits in Qi relative to the Q1 base group.
Except for one coefficient on Treatment × Q4 in the grocery stores group, none of the interaction terms is significant at conventional levels (p < 0.05). This means that there are no significant differences between highly liberal and more conservative counties. The effects are similar when we use sample splits instead of interactions (untabulated). In these subsamples, we observe a significantly negative coefficient on Treatment for grocery stores and restaurants in both liberal and conservative counties.
These effects are surprising given that liberals generally support LGBTQ+ activism (
Overall, we find that LGBTQ+ activism is associated with decreased store visits. These findings are consistent with prior research documenting firm value losses around activism events (e.g.,
In this study, we test whether LGBTQ+ activism influences consumer behavior. To do so, we use a large dataset of mobile phone foot traffic and estimate changes in store visits at firms that take stances on LGBTQ+ matters. Our results show that consumers decrease their visits at activist grocery stores and restaurants. One surprising finding is that the negative effect of LGBTQ+ activism not only exists in conservative counties, but also in (highly) liberal counties, suggesting that factors such as pink-washing concerns play a role in consumer behavior.
We note that our study faces several important limitations. First, our sample is based on an earlier publicly available dataset of mobile phone foot traffic downloaded in December 2021. The dataset, now available under a different provider, has likely extended its coverage since. Second, in contrast to prior studies that identify activism events in newspaper articles (e.g.,
Finally, while we detect decreases in store visits surrounding the activism events, an inclusive corporate image with meaningful support for minority groups may still increase consumer visits in the short and longer term.
Kim van Schie is an audit trainee at KPMG. She obtained her MSc in Accountancy and Control from the University of Amsterdam.
Susanne Preuss is an Assistant Professor of Accounting at the University of Amsterdam.
We appreciate the helpful comments by Sanne van Duin and Malte Max. This paper is based on Kim van Schie’s master thesis, which makes her one of the winners of the MAB thesis award 2023.
A difference-in-differences design compares the outcome of a treatment group to the outcome of a control group over time. The first difference is the outcome of the treatment group before and after the treatment. It accounts for systematic differences between the treatment and control group. The second difference is the outcome of the treatment group relative to the outcome of the control group. It accounts for factors other than the treatment that may explain differences in the outcome. In our setting, the treatment is LGBTQ+ activism and the outcome is store visits. An important assumption for this design is that without the treatment, store visits at the treatment group would have been identical to store visits at the control group (see e.g., https://dimewiki.worldbank.org/Difference-in-Differences)
For brevity, we use the term pinkwashing hereafter.
A closely related survey study by
We downloaded the dataset in December 2021, when SafeGraph Inc. still offered free access to researchers. In January 2023, the data moved to deweydata.io (
A disadvantage of using mobile phone geolocation data is that it is limited to in-person store visits, and we cannot test whether consumers substitute these visits by ordering online. However, we believe that this is less of an issue in our setting for at least three reasons. First, our sample firms consist of grocery stores, restaurants, and gasoline stations, whose business models are largely based on in-person visits, or which only launched their e-commerce business during our sample period (e.g.,
Activism can include both firm statements and actions (
In FactSet we use the search term ‘transgender OR lesbian OR lgbt* OR bisexual’. Twitter does not allow wildcards (special characters that allow for all alternative word endings; typically “*”). Hence, we individually search for the words lgbt, lgbtq, lgbtqia, lgbtqia+, lesbian, and bisexual.
There are slightly fewer observations for the variables capturing county-specific ideology, as some zip codes do not have matching presidential election voting data. Also, the variable Treatment has fewer observations, which is in detail explained in Section 3.3.
Our research design slightly differs from earlier studies on differences in store visits that focus on one moment in time (Hou and Polquin 2022; Painter 2023). In contrast to these studies, we examine multiple “staggered” events taking place at different points in time.
We use this design to estimate effects in the immediate aftermath of the activism (i.e., three months after relative to three months before the effect). Consider, for instance, Potbelly Corp, which had its first activism event on June 2, 2018. If we coded all months prior to the event as zero and as one afterward, we would have five pre-event months (January 2018 to May 2018), and 42 post-event months (June 2018 to November 2021).
We do not have sufficient data for a five-month time frame for two events (Winn-Dixie and El Pollo Loco). Hence, we drop these two events from the additional analyses.
Because we now use a stacked event dataset as opposed to the earlier panel, Treatment is one for firms with activism event, and zero otherwise.
We use the 2018 presidential election results for the months preceding November 2020, and the 2020 election results from November 2020 onwards.
LGBTQ+ activism events. This table presents the LGBTQ+ activism events. Panel A presents activism events of firms included in our sample, with events taking place between January 2018 and November 2021 (our sample time frame). Panel B presents activism events of firms excluded from our sample, with events taking place prior to January 2018.
Firm | Type | Tweet or press release title |
---|---|---|
Panel A: Activist firms included in sample | ||
Amazon.com, Inc. | Tweet | RT @amazonnews: We applaud SCOTUS’s decision to protect LGBTQ employees from discrimination. This is a historic win in our nation’s long struggle to ensure fairness & equal treatment for all. As the fight for full equality continues, we stand proudly with our public & private sector allies. (June 15, 2020) |
Chipotle Mexican Grill, Inc. | Press release | Chipotle Celebrates LGBTQ+ Community With ‘Love What Makes You Real’ (June 6, 2019) |
Domino’s Pizza, Inc. | Tweet | Happy #Pride Month! Help us fill in the blank: The LGBTQIA+ nonprofit organization I’d like to see Domino’s support is ______. (June 11, 2021) |
El Pollo Loco Holdings Inc. | Tweet | We’re proud to announce that Michaela Mendelsohn, CEO of Pollo West Group, has been named @LAPride 2018 Grand Marshal! Mendelsohn now sits among a number of respected LGBTQ+ activists and community leaders who have received CSW’s most prestigious honor. (May 23, 2018) |
Exxon Mobil Corp. | Tweet | Through our employee PRIDE network, ExxonMobil is energizing LGBTQ celebrations around the world. Check out a few of our favorite pictures from the Houston Pride Parade. (June 24, 2019) |
Lowe’s Companies, Inc. | Tweet | RT @LowesCareers: At @Lowes, we are proud to celebrate LGBTQA+ Pride month and appreciate our leadership team members, @SeemantiniGodbo, Don Frieson, and Janice Dupre for helping us kick off the month in style! #LowesForAll #Pride #LowesLife (June 4, 2021) |
Noodles & Co. | Tweet | Here’s a sweet way to support #PRIDE; order a colorful PRIDE Crispy–online or in-person–thru June 29. All proceeds for this limited-edition Crispy go to Out & Equal workplace advocates, our partner working exclusively on LGBTQ+ workplace equality http://order.noodles.com (June 27, 2021) |
Papa John’s International Inc. | Tweet | This week, we’re celebrating love and community with Meghan, Analyst and President of Papa John’s internal LGBTQ+ employee resource group for Equity, Advocacy, & Promotion. PapaProfiles (June 14, 2019) |
Phillips 66 Co. | Press release | Phillips 66 Earns Perfect 100 on 2021 Corporate Equality Index (February 2, 2021) |
Potbelly Corp. | Tweet | If you liked it, then you shoulda’ put a ring on it. #PeaceLovePotbelly #LGBTQ #LoveIsLove #Pride (June 2, 2018) |
Shake Shack Inc. | Tweet | We’re proud to stand with the LGBTQ+ community! This month, we’re servin’ up an appsclusive Pride Shake (strawberry shake blended with lemonade + whipped cream + sprinkles). $1 from every shake will benefit @TrevorProject! Sip + shop our Pride swag here: http://bit.ly/shack-pride-collection-2018 (June 1, 2018) |
The Kroger Co. | Tweet | Proud to earn 100% on @HRC’s Corporate Equality Index for LGBTQ-inclusive workplace policies and practices. #CEI2019 (March 28, 2019) |
Winn-Dixie Stores, Inc. | Press release | Southeastern Grocers Champions Belonging, Inclusion and Diversity with Pride Festival Sponsorships (September 14, 2021) |
Yum! Brands, Inc. | Tweet | Yum! Brands believes in ALL people and we celebrate the contributions of the LGBTQ+ community and its allies. Click here to learn about the importance and impact of inclusion and allyship in the workplace and beyond. (June 14, 2019) |
Panel B: Activist firms with events prior to sample start | ||
BP p.l.c. | Tweet | Proud to be recognized by @LGBTBar for our commitment to raising #LGBT diversity awareness in the workplace: http://bit.ly/10r18L2 (May 30, 2013) |
Chevron Corp. | Tweet | Diversity is 1 of r core values. For the 10th yr we r proud to have achieved a perfect score for LGBT equality @HRC http://spr.ly/60110nb3 (February 21, 2015) |
Darden Restaurants, Inc. | Tweet | We’ve been named a Best Place to Work for #LGBT Equality through @HRC, two years in a row. #CEI2014 http://bit.ly/18wuzPV (December 9, 2013) |
Denny’s Corp. | Press release | Denny’s Participates in 2016 Corporate Equality Index |
McDonald’s Corp. | Tweet | Proud to earn 100% on @HRC’s Corporate Equality Index for #LGBTQ- inclusive workplace policies & practices! http://McD.to/6013D6Lac #CEI2018 (November 1, 2017) |
Shell plc | Tweet | We encourage every person to bring their whole self to work. Celebrating (Inter)National #ComingOutDay. http://go.shell.com/2e2uceG #LGBT #NCOD (October 11, 2016) |
Walmart Inc. | Tweet | We’re excited to be recognized as one of the best places to work for #LGBT equality in @HRC’s #CEI2017! http://hrc.org/CEI (December 6, 2016) |
Firm | Brands |
---|---|
Amazon.com, Inc. | Amazon Fresh, Amazon Go, Whole Foods Market |
Chipotle Mexican Grill, Inc. | Chipotle Mexican Grill |
Domino’s Pizza, Inc. | Domino’s Pizza |
The Kroger Co. | Baker’s Supermarkets, City Market, Dillons Supermarkets, Food 4 Less, Foods Co., Fred Meyer, Fry’s Food & Drug Stores, Gerbes Super Markets, Harris Teeter, Jay C, King Soopers, Kroger, Kroger Fuel Center, Mariano’s, Metro Market, Pay Less Super Markets, Pick ‘n Save, QFC (Quality Food Centers), Ralphs, Ruler Foods, Smith’s Food & Drug Stores |
El Pollo Loco Holdings Inc. | El Pollo Loco |
Lowe’s Companies, Inc. | Lowe’s Market |
Noodles & Co. | Noodles & Company |
Potbelly Corp. | Potbelly Sandwich Works |
Phillips 66 Co. | 76, ConocoPhillips, Phillips 66 |
Papa John’s International Inc. | Papa John’s |
Shake Shack Inc. | Shake Shack |
Winn-Dixie Stores, Inc. | Winn Dixie |
Exxon Mobil Corp. | Exxon Mobil, Mobil |
Yum! Brands, Inc. | KFC, Pizza Hut, Taco Bell |