Does leader gender matter for performance evaluations? Evidence from two experiments

ABSTRACT

Public evaluations of organizations and their leaders are core features of public management used to foster accountability. Yet, prior research suggests that evaluations can be unduly influenced by leaders’ personal attributes and social identities, such as those signalled by gender. We examine these expectations using preregistered vignette experiments that vary the manager’s gender and level of performance in two distinct settings: education and policing. Results across both studies suggest women and men public managers are evaluated similarly overall; however, we find that the organizational context and identity of the evaluator likely play a role in evaluations of public managers.

KEYWORDS:

• Performance evaluations
• gender bias
• experiments
• public management
• women and leadership
• public perceptions
• citizen evaluations

Data availability and transparency statement

Hypotheses and protocols for both experimental studies were preregistered with the Open Science Framework prior to data collection. The preregistration for the first study can be retrieved at: http://bit.ly/2LoJX0j. The preregistration for the second study was identical, except we restricted participants from the first study from participating again. Data and replication code for both studies (in Stata format) can be accessed at: http://bit.ly/2LEfUla

Supplementary material

Supplemental data for this article can be accessed at https://doi.org/10.1080/14719037.2021.2000222.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. We use the word “citizen” to refer to the public broadly, not individuals with a certain citizenship status. We also refer to both gender and sex (e.g., woman and female) because the theories we engage with apply explicitly to cisgender women. Greater research is needed to understand how gender bias affects transgender and nonbinary public managers.

2. High-profile examples abound of top leaders being held accountable for organizational failures. Recently in the U.S., in the wake of a scandal at the U.S. Department of Veteran’s Affairs, General Eric Shinseki (retired) resigned as Secretary of Veterans’ Affairs. More recently, following the January 6^th riots at the U.S. Capital, the Capital Police Chief offered their resignation.

3. For instance, parental feedback in the form of survey responses constitute an important part of a quality review in New York City Schools.

4. We randomly selected one male and one female name from the top 20 names given to males and females born between 1970–1979 according to the U.S. Social Security Administration.

5. Additional information was provided to increase the contextual realism of the scenario. The number of students and personnel were based on national averages for U.S. high schools (National Center for Education Statistics 2010). ‘Lincoln’ was chosen as the name of the high school to avoid priming a geographic (e.g. Central) or an urban/rural connotation (e.g. MLK High).

6. Within each performance condition, the configuration of the arrows and their assignment to specific performance categories was randomly assigned to negate the possibility of the results being influenced by respondents weighting a specific performance category as more important than another.

7. For each of the plots, the effect of the gender treatment is statistically significant and negative, meaning the female principal is evaluated less favourably than the male principal, if the confidence interval falls below zero, and statistically significant and positive, meaning the female principal is evaluated more favourably than the male principal, if the confidence interval falls above zero.

8. See Figure A3 in the supplemental appendix. After including controls, the p-values for the marginal effect of gender in the low performance condition are 0.130 in the principal performance model, 0.116 in the principal adequacy model, 0.131 in the principal competence model, and 0.128 in the school performance model.

9. See supplementary appendix Tables A5, A6, A7, A8 for regression coefficients and Figures A4, A5, A6, A7 for marginal effects. The racial composition of the sample (74% white) precludes analyses of heterogeneous treatment effects across respondent race/ethnicity.

10. There is considerable heterogeneity in the organization of police departments in the U.S. Precincts defined by geographic boundaries are a common way that departments organize themselves hierarchically and assign responsibility for performance. Police captains and police precincts also provide a useful parallel to the school principals and high schools used in Study 1.

11. We use checkmarks and X’s instead of arrows since upward (downward) pointing arrows would convey worsened (improved) performance on some of these metrics (e.g. response times).

12. See supplementary appendix Tables B5, B6, B7, B8 for regression coefficients and Figures B4, B5, B6, B7 for marginal effects.

Additional information

Notes on contributors

Kendall D. Funk

Kendall D. Funk is an assistant professor of political science in the School of Social and Behavioral Sciences in Arizona State University’s New College of Interdisciplinary Arts and Sciences. Her primary research focuses on the causes and consequences of women’s representation in political institutions, especially at the local level of government in Latin America. She also researches issues related to local governance, decentralization, international public management, and government performance.

Ulrich Thy Jensen

Ulrich Thy Jensen is an assistant professor in the School of Public Affairs and a faculty affiliate at the Center for Organization Research and Design at Arizona State University. He is also an affiliate at the Crown Prince Frederik Center for Public Leadership at Aarhus University. His research interests span issues of leadership, motivation, values and performance in public service. His research builds on new and innovative ways to understand the importance of leadership in shaping the attitudes of public service providers and the performance of their organizations.

Angel Luis Molina

Angel Luis Molina, Jr. ([email protected]) is an assistant professor in the School of Public Affairs at Arizona State University. His research interests include the impacts of organizational behaviour on policy outcomes in minority communities, the performance implications of minority access to organizational leadership, public values, education policy, immigration, Latino politics and the salience of race and ethnicity for both public and private sector outcomes.

Justin M. Stritch

Justin M. Stritch ([email protected]) is an associate professor at Arizona State University’s School of Public Affairs and is a Senior Research Affiliate at the Center for Organization Research and Design (CORD). His research focuses on public employee attitudes and motivations, quality of work life, organizational fairness and equity, decision making, and public sector sustainability.