Policy Brief Reducing Gender Gaps in Management: Experimental Evidence from Peru March 2025 Photo: Cityscape view of Ayacucho, Peru by jkraft5 via canva.com KEY MESSAGES Simple, cost-effective information campaigns, particularly those using behavioral science insights -such as personalized messaging, role models, and motivational cues-, can effectively encourage female teachers to apply for school management positions. Gender disparities are present at multiple stages of Peru's Concurso de Acceso, the process for assigning school management positions. Although women represent a significant portion of the eligible candidates, their participation decreases at each stage, with the most notable decline occurring during the registration process (the first of four steps). While gender gaps in school leadership exist both in urban and rural areas, targeted behavioral interventions are particularly effective in rural settings. By addressing specific barriers faced by women in these areas—such as limited access to information and fewer role models— policies can reduce gender gaps in access to management positions by more than 20%. BACKGROUND Across the world most teachers are female, but The first step for participation is to apply during most school administrators are male. In Latin the registration window. The registration process America, there is a 20-percentage point (p.p.) gap takes place online and requires only basic between the share of school leaders who are female information such as a national identification and the share of teachers who are female (Adelman number and education. and Lemos, 2021). Qualified female candidates often fail to apply for these management positions. Once registered, the second step for participation is to take part in a centralized standardized The fact that women who have the skills, examination called Prueba Unica Nacional. The experience, and ability to be considered for examination tests teachers’ knowledge in reading promotions opt out of competitions for promotion comprehension, public sector management, and is not cost-free. It reduces the overall skillset in the education management. If approved, the third step pool of applicants. And the underrepresentation of is to participate in a decentralized evaluation women discourages future generations of women process by ranking a selected number of regions from pursuing leadership positions. In other and local authorities and being evaluated through settings where women are underrepresented, an interview and review of experience, education women have been found to be more productive, on and other achievements by a school committee average, than men because they had to be far above guides by the Ministry of Education. the bar to overcome the barriers to entering the field (Anzia and Berry, 2011). This suggests that Job offers are jointly determined by the scores equalizing representation of men and women in received during the centralized standardized school leadership positions could raise the average examination and the evaluation conducted by the quality of school leadership, and possibly student school committee. The fourth and final step is to learning outcomes (Adelmann and Lemos, 2021). In accept the job. education, closing gender gaps in management is crucial. Diverse management teams make more Using data from previous Concursos de Acceso held balanced, equitable decisions that benefit a broader in 2014, 2016, and 2018, we calculate the share of range of students. For example, gender diversity in women who make up the total stock of teachers at school leadership can lead to more inclusive each stage of the Concurso de Acceso and the curricula, better representation for female teachers subsequent two years on the job (Figure 1). We and students, and stronger role models for future observe two important patterns. First, only 30 generations of leaders. percent of all eligible teachers (with tenure and seniority) applied to earlier Concursos de Acceso. CONTEXT Second, the largest drop in female participation occurs during the registration window. In Peru, public sector teachers who wish to pursue a school management career must apply to the We also observe large differences in gender gap Concurso de Acceso. This is a centralized system across urban and rural areas. In rural areas, women for allocating school management positions. Eligible make up 49 percent of eligible teachers, but they teachers sign-up for a national-level process make up only 38 percent of teachers who have involving a qualification exam, application to applied to the Concurso de Acceso. While in urban specific positions, and interviews. areas, women make up a larger share of eligible teachers at 65 percent, they only make up 54 There are two eligibility criteria for participation: percent of teachers who have applied. In both rural public sector teachers must be tenured and have a and urban areas, the increase in gender gap at this minimum level of seniority as measured by their stage is 11 p.p. salary scale (third salary scale or above out of eight salary scales). Figure 1: Gender Gap in Stages of Previous Concursos de Acceso Note: This figure uses data from each stage of the Acceso Contest selection process carried out in 2014, 2016, and 2018 and shows (i) the average number of teachers for all regions represented by the bars, and (ii) the average gender gap for all regions as well as urban and rural regions represented by the lines. INTERVENTION During earlier Concursos de Acceso, all eligible For example, the enhanced WhatsApp messages teachers were sent a series of standard text also included elements such as personalization messages (SMS) and emails from the MINEDU with (Castleman and Page, 2015; Madrian, 2014); information on timelines and registration affirming statements (you are one of the female procedures. teachers with the required experience for the positions advertised) appealing to the teachers’ In preparation for the 2023 Concurso de Acceso, we self-efficacy, that is, beliefs about their ability to partnered with Peru’s Ministry of Education successfully achieved the desired outcome (MINEDU) and designed an intervention to test the (Bandura, 1977; Judge and Bono, 2001); a call for difference between the standard messagesand a action (Participate!); role models and peer effects set of ‘enhanced’ messages that added motivational by indicating that many female teachers have cues for female teachers to participate in the succeeded in past Concursos de Acceso and Concurso de Acceso. We added a third medium to including the link to a video with some of these transmit these messages: WhatsApp. teachers narrating their experience, their doubts when deciding to apply, how they overcame them, Leading up to the registration deadline, all and the personal rewards, such as family pride, for candidates received four SMS messages and four their accomplishments; loss aversion (Don’t let this emails originally designed by the MINEDU as well as opportunity pass you by!) and extrinsic motivation four WhatsApp messages as part of the for the teacher identity (Our students count on intervention – either information-only WhatsApp you!). messages or enhanced WhatsApp messages. All WhatsApp messages were sent on Fridays or The enhanced WhatsApp messages were designed Saturdays (when possible) outside of teaching building on insights from behavioral science hours and/or when recipients could have time to regarding drivers of behavior, intentions formation act on the message’s prompts to increase the and action (Ajzen, 1991; Gollwitzer and Sheeran, likelihood of the messages to be read. 2006; Michie et al., 2011; Oettingen and Mayer, 2002). FINDINGS To evaluate this intervention, we randomized the the centralized examinations, and being offered a treatment at the school level. Eligible female and management position in rural areas, but not in male candidates in schools assigned to the control urban areas. group and male candidates in schools assigned to the treatment group received the information-only Impact on application rates: Figure 2 presents the WhatsApp messages. Eligible female candidates in estimated impact of the intervention on application schools in the treatment group received the rates, which is the first step in the Concurso de enhanced WhatsApp messages. This design allowed Acceso. Overall, sending messages that are us to isolate the effect of the enhanced messaging carefully designed to eligible female candidates to intervention. encourage them to participate in the Concurso de Acceso led to a 1 p.p. increase in application rates We use administrative data provided by the relative to sending shorter, informational text MINEDU to estimate the impact of the intervention messages. However, this effect is only marginally on gender gaps. The treatment effect measures the statistically significant at the 10 percent level. impact of the enhanced WhatsApp messages Furthermore, it masks substantial heterogeneity by relative to the information-only WhatsApp urban and rural location. While the gender gap in messages. We focus on the overall pooled effect as application rates is just over 16 p.p. in both urban well as the effect in urban and rural areas and rural areas, our observed treatment effects are separately. Overall, 26 percent of eligible driven by eligible female candidates in rural areas. candidates are from rural areas and 74 percent are Among this subset, the text messages increased from urban areas. eligible female candidates’ application rates by 3.3 p.p. This effect is both statistically significant at We find that the intervention reduces the gender the 1 percent level and sizable- corresponding to a gap by a range of 20.6 to 32.5 percent in the 20.6 percent reduction in the 16.1 pp regression applying for the Concurso de Acceso, in passing adjusted gender gap in rural areas. Figure 2: Overall, Urban v. Rural Gender Gaps in Application Rates Note: This figure shows the regression adjusted gender gap in application rates in the control and treatment groups, respectively, pooled and separately for urban and rural applicants. Impact on examination pass rates: Once Impact on job offer rates: The final step is to be registered, the second step for participation is to offered a managerial job. Figure 4 shows the take to centralized standardized examination - estimated impact of the intervention on gender Prueba Unica Nacional (PUN). While there is a very gaps in being offered a management position. small gender gap of approximately 1 to 2 p.p. in Again, we observe a small, positive, statistically taking the examination, there is a 10 p.p gender gap insignificant effect in the full sample. in passing the examination. Just under half (49 percent) of applicants to the Concurso de Acceso As with the earlier results, the overall impact on the pass the centralized exam. gender gap is driven by a large and statistically significant effect in rural areas, as opposed to a Figure 3 shows the estimated impact of the statistically insignificant effect in urban ones. The intervention on the gender gap in passing the PUN. intervention increased the relative probability that These estimates are unconditional on registering women from rural areas were offered a position by for the Concurso de Acceso so capture both 1.8 pp, significant at the 5 percent level. This is changes in application rates and pass rates. In the equivalent to a reduction of 32.5 percent in the full sample, the text messages cause a 0.19 p.p. gender differential in the probability of receiving a increase in the proportion of women who pass the job offer. Again, we find smaller, negative, exam relative to men. This is a 1.9 percent statistically insignificant impact for candidates for reduction in the 10 p.p. gender gap in rates of urban areas. passing the PUN and it is not statistically significant. Figure 4: Overall, Urban and Rural Gender Mirroring the results observed during the Gaps in Being Offered a Managerial Job registration stage, among candidates from rural areas, the enhanced WhatsApp messages reduced the gender gap by 2.4 pp, or 27.3 percent in rural areas. There is a negative, but statistically insignificant, impact on the gender gap for candidates from urban areas. Figure 3: Overall, Urban and Rural Gender Gaps in Passing Centralized Examinations Note: This figure shows the regression adjusted gender gap in application rates in the control and treatment groups, respectively, pooled and separately for urban and rural applicants. Note: This figure shows the regression adjusted gender gap in application rates in the control and treatment groups, respectively, pooled and separately for urban and rural applicants. POLICY IMPLICATIONS The findings from this study underscore the importance of targeted interventions in addressing gender disparities in school management. Simple, cost-effective information campaigns—especially those leveraging behavioral insights such as personalized messaging, role models, and motivational cues—can significantly increase female participation in managerial positions. However, the impact is more pronounced in rural areas, suggesting that policies to close gender gaps in school management must be tailored to regional contexts. The larger effect in rural areas may be due to several factors. Rural teachers often face greater informational barriers, as well as fewer role models and support networks. Therefore, a well-targeted intervention like this one, which includes motivational cues and peer success stories, could have a more pronounced impact on their decision to apply for management positions. In urban areas, where women are already more engaged, additional strategies may be needed to tackle other barriers, such as implicit biases or institutional structures that disproportionately affect women’s advancement. These results advocate for the integration of behavioral science into educational policies to boost female representation in management roles, particularly in underserved areas, thereby fostering more inclusive decision-making in the public sector. Photo: Cusco, Peru by Sisoje via canva.com REFERENCES Adelman, Melissa and Renata Lemos (2021). Managing for Learning: Measuring and Strengthening Education Management in Latin America and the Caribbean: The World Bank. Ajzen, Icek (1991). “The theory of planned behavior,” Organizational Behavior and Human Decision Processes, 50(2), 179–211. Anzia, Sarah F and Christopher R Berry (2011). “The Jackie (and Jill) Robinson effect: why do congresswomen outperform congressmen?” American Journal of Political Science, 55(3), 478–493. Bandura, Albert (1977). Self-efficacy: The exercise of control. W.H. Freeman and Company. Castleman, Benjamin L and Lindsay C Page (2015). “Summer nudging: Can personalized text messages and peer mentor outreach increase college-going among low-income high school graduates?” Journal of Economic Behavior & Organization, 115, 144–160. Gollwitzer, Peter M and Paschal Sheeran (2006). “Implementation intentions and goal achievement: A meta-analysis of effects and processes,” Advances in Experimental Social Psychology, 38, 69–119. Madrian, Brigitte C. (2014). “Applying Insights from Behavioral Economics to Policy Design,” Annual Review of Economics, 6(1), 663–688. Michie, Susan, Maartje M Van Stralen, and Robert West (2011). “The behaviour change wheel: a new method for characterising and designing behaviour change interventions,” Implementation Science, 6(1), 1–12. Niederle, Muriel and Lise Vesterlund (2007). “Do women shy away from competition? Do men compete too much?” The Quarterly Journal of Economics, 122(3), 1067–1101. ACKNOWLEDGEMENTS This note was prepared by Ciro Avitabile, Ana Maria This work has been funded by the Umbrella Facility Munoz Boudet, Jonathan M.V. Davis, and Renata for Gender Equality (UFGE), which is a multidonor Lemos. It received useful comments from Jacobus trust fund administered by the World Bank to de Hoop and Raquel Melgar Calderón. advance gender equality and women’s empowerment through experimentation and The LACGIL supports impact evaluations and knowledge creation aimed at helping governments inferential research to generate evidence on what and the private sector focus policies and programs works in closing gender gaps in human capital, on scalable solutions with sustainable outcomes. economic participation, social norms, and agency. The UFGE is supported with generous contributions Additionally, the lab disseminates findings to from Australia, Canada, Denmark, Germany, improve operations and policy making in the design Iceland, Latvia, the Netherlands, Norway, Spain, of cost-effective interventions that tackle gender Sweden, Switzerland, the United Kingdom, the inequalities and drive change. To accomplish this, United States, and the Bill and Melinda Gates the LACGIL works in partnership with World Bank Foundation. The findings, interpretations, and units, aid agencies and donors, governments, conclusions expressed in this brief are entirely nongovernmental organizations, private sector those of the authors. They do not necessarily reflect firms, and researchers. the views of the World Bank, its affiliated organizations, the Executive Directors of the World Visit the LACGIL website for more information Bank, or the governments they represent. Email: lacgenderlab@worldbank.org 1818 H. St NW Washington, DC 204 This material should not be reproduced or distributed without the World Bank's prior consent.