Companies have frequently been targets for responding to the social issues of the day. Some companies, like Target, release annual “Pride Month Collections” to show their support of gay rights. Others, like Nike, institute company-wide policy changes in response to #BlackLivesMatter (Vredenburg et al., 2020). These actions generally receive positive responses from consumers (Afzali & Kim, 2021). While the transition to working from home due to Covid-19 has reinvigorated the conversation on how companies support workers’ mental well being, little research has been done in examining if rewarding a company’s corporate social responsibility (CSR) would translate to support for worker mental wellbeing as well. Beyond that, there is the question of what is the best target of such activism - focusing on the individual by directly caring for their employees, or by caring for the societal issue at large (Jaspal et al., 2016).

In this short report, we test to see if exposing consumers to companies (not) engaging in CSR towards its workers’ mental health leads to increased social and active engagement at the store due to a perceived support of the issue, and if this differs based on the level of activism at the individual or societal levels.

Method

Using a within-subjects design, 160 college students, faculty, and staff (63.1% female) were recruited through email from a small midwestern college in the United States. They completed an experimental study in which they were randomly presented with six mock Instagram ads (GenerateStatus.com, 2021) that described how a faux company (generated by AI - see: https://namelix.com/) engaged with workplace mental health (Overt|Performative). Overt activism included the company listing direct benefits given, while performative activism involved the company stating information about the issue without direct, explicit actions. We also varied whether they saw it as a societal (donating to a global mental health foundation; noting worldwide mental health statistics) or individual (listing workplace mental health benefits to employees including paid mental health days; extending hours for shopping making the employer’s day longer) issue. We also included two neutral ads for distraction. All analyses (including supplemental), experimental materials, and data are hosted at https://osf.io/b6f5t/.

After each ad, participants were asked to rate their engagement with the ad (liking the post; purchasing from the store, visiting the website) using 7-point scales (𝜔’s range from 0.91–0.93) and a single question measuring perceptions of the company’s support towards mental health. Participants then completed the General Health Questionnaire to control for their own mental health (GHQ) (Werneke et al., 2000) and demographic questions. See Table 1 for demographics and correlation matrix.

Table 1 Correlation matrix and descriptive statistics
Full size table

Results

As a manipulation check, A 2 (Level: Individual/Societal) × 2 (Action: Overt/Performative) linear mixed model using Satterthwaite’s degrees of freedom on perceived support showed a significant interactionFootnote 1 (b = −1.11, SE = 0.19, t(477) = −5.7, <.001). Pairwise comparisons on the estimated marginal means showed that overt companies did not differ in amount of perceived support (Mindividual − overt = 5.21, SEindividual − overt = 0.13; Msocietal − overt = 4.92, SEsocietal − overt = 0.13; p = .17, d = 0.23). However, there were significant differences between all other comparisons ps < .001, ds ranging from 0.45 to 1.36 (Mindividual − performative = 3.54; SEindividual − performative = 0.15), Msocietal − performative = 4.37, SEsocietal − performative = 0.14). Overt companies were perceived to show stronger support for mental health than performative companies, suggesting our manipulation worked as intended.

A 2 (Level: Individual/Societal) × 2 (Action: Overt/Performative) linear mixed model using Satterthwaite’s degrees of freedom on engagement exhibited a significant interaction between the total mean scores, b = −.25, se = .12, t(477) = −2.07, p = 0.039 (Fig. 1). Tukey adjusted pairwise comparisons showed that each condition was significantly different from each other (ps < .046; ds ranging from 0.29 to 1.27), with Overt Societal being the highest (M = 3.75, SE = 0.11), Overt Individual (M = 3.52, SE = 0.12), Performative Societal (M = 3.24, SE = 0.12), and Performative Individual (M = 2.77, SE = 0.1) being the lowest.

Fig. 1
figure 1

Average engagement with ad by support and level

Full size image

A mediation analysis of 5000 bootstrapped samples was run using the MEMORE SPSS extension (Montoya & Hayes, 2017) examining the indirect effect of social and active engagement with the store through the perceived support of mental health. We compared the overt conditions against the performative conditions since a multilevel, within-subjects mediation requires pairwise comparisons (Montoya & Hayes, 2017). Overt action (b = 0.63, SE = 0.07, t = 9.54, p < .001) predicted engagement with the ad and perceived support of mental health (b = 1.11, SE = .11, t = 9.95, p < .001). When both were included in the same model (F(2,157) = 61.21, p < .001, r2 = .44), the effect of the overt action condition on engagement reduced (b = 0.19, SE = 0.06, t = 2.96, p = .004), while the effect of perceived support remained (b = 0.40, SE = 0.04, t = 11.02, p < .001). A bootstrap test produced a 95% confidence interval for the overall indirect effect that did not include zero, b = 0.44, SE = 0.06, CI = [0.33, 0.57] (see Fig. 2).

Fig. 2
figure 2

Mediation path of perceived support of worker’s mental health based on the action of the firm on engagement with the firm

Full size image

Discussion

Using a novel design of faux Instagram posts, we found that companies who engaged in overt activism were the most likely to be rewarded with engagement, and this was particularly the case for the global level. Firms’ activism was rewarded by consumers through the perception that these firms truly cared about worker’s mental health. Our design avoids confounds of pre-conceived feelings towards recognizable companies, while also addressing issues around imagining a nameless company.

The effect of global activism being more rewarded was unexpected. It could be that participants were perceiving donations as a larger cost to the company than mental health days, even though the perceived caring of the issue was equal between overt groups (Shnayder et al., 2015). While our results suggest that perceived care of mental wellbeing mediated the relationship between company’s action and consumer engagement, we cannot claim a true mediating pathway due to the cross-sectional nature of our design (Maxwell et al., 2011). Furthermore, while our repeated measures design increased our power, more work could be done with larger longitudinal designs.

Our work expands current knowledge of CSR by adding the new movement of workers’ mental health to the discussion, when most CSR work focuses on issues of diversity, energy, climate, and redistributions of wealth. We also provide evidence that the rewarding of CSR through engagement is in part due to believing the company genuinely holds concern for its workers. Future research should use more salient social justice concerns that may be immaterial to the company’s products, such as LGBTQIA+ rights, women’s rights, or minority rights.