Principled consumption: How gender pay gaps affect perpetrators

Building on previous research, we predicted that disclosing a gender pay gap would dampen purchase intentions by negatively affecting perceptions of fairness. We also proposed that the disclosure of pay gaps may not equally affect how a firm is perceived across different consumer segments.

Prior research suggests that policies requiring public disclosure of information can be powerful means of enabling social change and furthering the greater good—but only as long as the information disclosed is meaningful to the recipient. Gender plays a role in how an individual takes fairness into account when assessing the compensation of others: When asked to allocate rewards between themselves and others, women often choose more equal allocations than men, even when the inputs are identical. Prior research also suggests that women generally are more punitive than men in response to business transgressions and are more likely to boycott firms.

Women are also more likely than men to reject sexist behaviour towards women, and therefore, we felt that consumers’ responses to a gender pay gap—which could be considered a form of sexism—may also depend on gender. We therefore proposed that gender moderates the impact of pay gap disclosure on purchase intention, affecting the purchasing behaviour of women more than men.

Finally, we proposed that disclosing a gender pay gap may not always negatively affect consumer purchase decisions. While we expected that consumers (particularly those who are female) would avoid companies with gender pay gaps when they have the opportunity to do so, we also felt that consumers who face personal costs or barriers in the moment might rationalize a decision not to punish an offending firm.

Also read: Women at work: Can we bridge the gender pay gap once and for all?

Our Research

We conducted six studies to examine when, why and for whom revealing gender pay gaps negatively affect consumer purchase behaviour.

In our first study, the objective was to test whether disclosing a gender pay gap affected consumers’ willingness to buy, whether that effect was caused by perceived wage unfairness, and whether this effect differed by gender. We recruited 300 participants (57.7% male) to complete an experiment on Amazon Mechanical Turk.  All were told the following: “Adidas is a multinational corporation that is engaged in the design, development, manufacturing and worldwide marketing and sales of footwear, apparel, equipment, accessories and services.”

In the ‘gender pay gap present’ condition, participants were also told the following: “Adidas recently reported women working for the company have an average hourly rate that is 18 per cent lower than men’s. In other words, comparing average hourly rates, women earn 82 cents for every dollar that men make.” This wording was directly taken from the UK’s Gender Pay Gap Service portal (https://gender-pay-gap.service.gov.uk).

Participants then indicated their willingness to buy from Adidas on a seven-point scale (1: Not at all likely, 7: Very likely) and indicated their perception of wage fairness on a 7-point scale (1: Not at all fair, 7: Very fair). The result: This study provided evidence that when gender pay gaps are revealed about a firm, consumers are indeed less willing to pay for that firm’s products. Furthermore, the effect of pay gap disclosure on purchase intention is related to gender: While men and women both responded negatively to the disclosure of this gender pay gap, women showed a particularly heightened response.

In our second study, we tested whether gender pay gap disclosure affects consumer behaviour in a realistic real-world context. This study used an interactive mock-up of the Google News website, with actual news articles embedded for all to see. Participants (800 MTurkers, 37.9 % male) were randomized to one of two conditions: a ‘control condition’ versus a ‘gender pay gap disclosure condition.’ Everyone was led to the Google News mock-up website where they could read four articles, all of which were related to Uber.

In the gender-pay-gap-disclosure condition, the first and last article headlines referenced Uber’s gender pay gap, while the second and third articles were randomly chosen from a pool of articles about the company, unrelated to the pay gap. In the control condition, four randomly chosen news articles appeared, with all article links leading to published articles from reputable news sources. Participants could click on the articles and read them, or forward them automatically to their e-mail program. When they had finished reading any articles of their choice, participants returned to the original study site and continued with the experiment.

Our finding: When presented with an opportunity to use Uber, participants in the ‘inequality condition’ were significantly less inclined than the control condition to proceed. They were also less likely to do so when we asked them to imagine a 10 per cent chance of rain. However, when they imagined a 90 per cent chance of rain, there were no significant differences between the two conditions.

We followed up to test whether this could be attributed to the increasing likelihood of rain. Indeed, while the interaction between condition and gender remained significant, the three-way interaction in this repeated measure design with 90 per cent rain was significant, showing that the interaction became less strong when the chance of rain increased. Thus, the greater the chance of rain, the more significant the gender-specific effects of the pay gap became.

Also read: For women, early promotion is vital to reducing the gender pay gap

Study 3 aimed to test whether gender pay gap disclosure affects consumer behaviour when the decision is consequential. Comparing Uber’s average hourly rates, women earn 67 cents for every $1 that men make. This stimulus was directly informed by actual self-reported data from the UK’s public portal. In the first condition, participants did not receive that information but only an introduction

to the company (“Uber is a car-for-hire service that relies on smartphone technology to dispatch drivers and manage fees”).

Then, they were endowed with $1.00 and were introduced to a task that asked them how much they would like to bid on a $5 Uber gift card that they would be randomly assigned to a group of nine other participants. The participant with the highest bid in each group would receive the gift card (in reality, they were given the cash value of the gift card). The bidding mechanism worked similar to those on common auction platforms such as eBay, where only the highest bidder pays for their offer. Then, participants answered the same question about wage fairness as in the previous studies.

Our finding: Revealing the pay gap resulted in a lower bidding price, and this result held even when controlling for participants’ Uber usage. Participant gender marginally moderated the effect of pay gap disclosure on bidding price. The contrast between women across the two  conditions was significant, while the contrast between mean across the two conditions was non-significant. This aligned with our prediction: Female participants bid less when they learned about Uber’s wage gap.

In our fourth study, we tested how consumer responses to gender pay gap disclosure compare with disclosure of other transgressions such as child labour, illegal dumping, and price-fixing. We found that consumers responded differently across transgression types. They were less willing to buy when the gender pay gap was disclosed relative to two other types of transgressions: having an anti-family culture and wasting energy. Participants in the price-fixing, dumping waste, and harassment conditions were not significantly different from those in the gender wage gap condition. Those in the minority discrimination condition were marginally less likely to purchase relative to disclosing the gender wage gap. Only participants in the child labour condition were significantly less likely to purchase relative to disclosing the gender wage gap.

Next, we analyzed how participant gender differentially affected the responses to various business transgressions. The difference between the male and female participants in the child labour condition was marginally significant, as was the difference for those in the price-fixing condition, the discrimination condition, and the waste-dumping condition. The difference between male and female participants in the energy, antifamily, and harassment conditions was not significant. These results demonstrated that for the gender pay gap condition, the effect on purchase intention was particularly strong for women relative to men compared to at least a subset of other business transgressions.

This study provides evidence that the presence of a gender wage gap matters (on average) as much as—if not more than—other substantive business transgressions in affecting consumers’ willingness to buy. Moreover, supporting prior research, we demonstrated that women generally are more punitive than men in response to business transgressions across the types we examined.

In Study 5a we used field data to explore the relationship between gender pay gap disclosure and consumer sentiment, examining discussions taking place on social media. The interaction between pay gap and disclosure date showed that after April 3, 2018, a larger gender pay gap led to more negative emotional tweets referencing the offending company on Twitter. After the wage gap was published, there was a strong negative correlation between the gap and valence of the collected tweets. Put simply, people responded negatively to a disclosed gender pay gap, and the larger that gap was, the more negative their responses were on Twitter.

Also read: Will demand for women executives finally shrink the gender pay gap?

This study indicates that the disclosure of gender pay gaps is associated with negative consumer sentiments on social media. Specifically, consumers articulated much more negative sentiments

about companies that reported greater wage gaps, suggesting that a systematic publication of wage gaps would be necessary to ignite consumer responses. This finding is particularly meaningful since prior work has demonstrated that negative consumer sentiments on Twitter can adversely affect businesses.

A key drawback of Study 5A was that we were not able to observe the gender of those authoring the tweets. Thus, while the study measured consumer sentiments independent of their gender—due to the lack of information—we conducted a follow-up study in a similar setting to test one of our key predictions that women would respond more negatively to a gender pay gap than men.

For Study 5b, we obtained a second set of 500 tweets by searching for tweets that either mentioned ‘gender gap’ or ‘pay gap’.

A research assistant coded the perceived gender (i.e., male vs. female vs. unidentified/unsure) of the tweets’ authors based on their name and profile pictures, resulting in a sample of 348 (56.6 per cent female).

Next, we again constructed a measure for user sentiment by subtracting the number of words related to negative emotions from the number of words related to positive emotions that were mentioned in each of the tweets. Study 5b added further nuance to the findings of Study 5a: The gender pay gap was not only discussed more among women, but women also responded more negatively to this topic.

Key Takeaways

Our research indicates that gender pay gap disclosure does indeed affect consumer purchase intentions and that this effect is mediated by perceptions of wage fairness. Some might argue that many consumers are likely to engage in ‘cheap talk’ and won’t follow through and actually punish companies in real life. However, we found that disclosing a gender gap materially affects different consequential outcomes.

This area of inquiry is particularly timely given that firms are under increasing pressure to disclose gender pay gap information that was previously kept confidential. Currently, since much of the publicly available data on pay gaps is hosted on government-run websites, such data are not salient to customers at the point of purchase. However, upon disclosure, a firm’s gender pay gap could ostensibly be publicized by the news media, activist groups, politicians, or competitors. Our research suggests that firms whose competitors have pay gaps could benefit from making gender pay gaps more salient to consumers, as could activist groups. Consumer-facing firms with gender pay gaps could directly lose customers as a result of greater customer awareness, and particularly when their customers are women.

We believe these results have societal implications for the greater good: pay gap disclosure can lead to changes in consumer responses, thereby putting pressure on firms to reduce gender inequality.

Katherine DeCelles is a Professor of Organizational Behaviour and Academic Director of PhD Program at the Rotman School of Management, where she holds the Secretary of State Professorship in Organizational Effectiveness. She is also cross-appointed to the Centre for Criminology and Sociolegal Studies at the University of Toronto. She is the co-author of “Consumers—Especially Women—Avoid Buying From Firms With Higher Gender Pay Gaps” with Tobias Schlager (University of Lausanne), Bhavya Mohran (University of San Francisco) and Michael Norton (Harvard Business School). Their paper was recently published in the Journal of Consumer Psychology.