Is there a gender gap in our responses to a gender pay gap? An experiment

ABSTRACT

A significant amount of academic work has gone into trying to identify and quantify the importance of the key sources of the significant and persistent gender pay gap that exists. However, far less attention has been devoted to testing whether the desire or willingness to pursue policies to reduce this gap differs across gender. Using a laboratory experiment, I find, perhaps unsurprisingly, that males are far less likely than females to support such a policy when males are the higher-earning gender. However, the results flip dramatically when the earning positions are reversed along gender lines.

KEYWORDS:

• Gender pay gap
• equality
• fairness
• intergroup conflict

JEL CLASSIFICATION:

• D63
• D64
• D60

Highlights

 In an experiment, subjects can choose to vote to reduce a gender pay gap.

Most males vote against the gender-pay-gap-reducing policy when males are advantaged.

Most females also vote against this ‘redress-pay’ policy when females are advantaged.

Overall, a minority of subjects vote for this ‘redress-pay’ policy.

Acknowledgments

I am deeply grateful to those at the Department of Economics of the Ohio State University who graciously allowed me to use their experimental lab along with their subject pool. Special thanks to Dr Paul Healy and Samantha Stelnicki (student lab manager at the time of the experiment). Thanks also to my wife, Cameisha Williams, and Alexandra Edrington (a student teaching assistant) for helping to proofread the manuscript, and to three graduate assistants (David Lenzini, Samuel Mancini, and Adam Mohamed), who helped with conducting the experiment sessions.

Disclosure statement

There is no potential conflict of interest to report.

Notes

¹ It may be best to think about the pay gap discussed in this paper in the context of the pay gap that exists in many professional sports, such as international tennis, soccer, and cricket. In these real-world settings, a reasonable (though still controversial) argument can be made that males generate greater fan interest and greater revenues than females or that males are required to exert more effort (e.g. males play longer matches in tennis). In spite of these differences in income-generating attributes or efforts across genders, in each of these sports, there is still an earnest push towards ‘equal pay’, or perhaps more accurately, ‘equal earnings’.

² Madison Keys, a twenty-five-year-old female American tennis player, sums up the contentious nature of the discourse over ‘equal pay’ in tennis in this way, ‘I feel like it’s one of those conversations: When you’re friends with someone, you don’t talk about politics; when you’re friends with someone, you don’t talk about equal prize money in tennis.’ (Rothenberg (2016))

³ Background information for the experiment, such as information on location and recruitment, can be found in Appendix 5.1. Also, for samples of the key experiment instructions, see Appendix 5.2.

⁴ Given the standard probit regression model, we can recover the various probabilities using the cumulative normal distribution ($\mathrm{\Phi }$) and the estimated coefficients ($\mathbf{b}$). Suppose $Y_i=1$ when subject i votes in favour of the redress-rate pay option, $X_{2i}=1$ when subject i is a male, $X_{3i}=1$ when subject i is in a male-advantaged treatment session, and $X_{4i}$ is the vector of other regressors. ${\beta }_1$ (constant), ${\beta }_2$, ${\beta }_3$, and ${\beta }_4$ (a vector) are the corresponding estimated coefficients. Given these notations, we can calculate the probability that a male subject in a male-advantaged treatment will vote in favour of the redress-rate policy as $P(Y_i=1|X_{2i}=1,X_{3i}=1)=\mathrm{\Phi }({\beta }_1+{\beta }_2{X}_{2i}+{\beta }_3{X}_{3i}+{{\beta }_4}^{\prime }{X}_{4i})$.

Additional information

Funding

Partial funding for the experiment came from a research grant that was provided by the Dean’s Office of the School of Business of the University of Dayton. Supplemental funding was provided by the Department of Economics and Finance of the University of Dayton. I am greatly appreciative of the funding provided by both departments. However, the role of these two funding sources did not extend in any way beyond providing financial support; University of Dayton [Dean’s Research Grant, School of Business, Department of Economics and Finance].