Empowering Women through Safety at Work: An Experiment to investigate Sexual Harassment in the Workplace.

Our cluster level intervention (first stage) leverages three motivational levers: (1) Gender-equity: appealing to owners who value gender equality and prioritize safeguarding female employees but likely lacked the tools to do so; (2) Commitment to compliance, targeting those who are motivated by the need to align with existing laws and ensure their practices meet established standards; and (3) Social alignment and peer influence, which resonate with owners who are inspired by observing their peers adopt similar measures. These will be used to study the impact of the pitches on both intention to take up and actual take up of our offered intervention of sexual harassment awareness training.

For the firm level intervention (second stage), we have developed a sexual harassment training intervention in consultation with two organizations with significant experience in this field. The training is designed to be interactive, engaging and inclusive, ensuring it is not only informative but also actively involves participants. It is 30 minutes long and includes videos that recreate real-life scenarios, Q&A sessions, and dynamic interactions, helping to avoid the pitfalls of passive, boring training sessions. The training also incorporates behavioral modeling by using video-based scenarios demonstrating both desired and undesired behaviors.

Our training is tailored to the workplace environment specific to the urban retail sector and aims to integrate the topic into broader discussions about workplace culture, strategy, and customer service. It broadly covers five key topics: (1) recognizing sexual harassment, (2) how to address sexual harassment in the workplace, including bystander intervention strategies for employees, (3) benefits of preventing sexual harassment, including effective strategies for bystander interventions, (4) legal obligations of workplaces under the POSH Act, including maintaining necessary reporting mechanisms for owner training, and (5) rights of workers under the POSH Act as well as numbers of helplines in case of an SH incident for worker training. This will be self-administered by both employers and employees in treatment firms.

We will also provide other tools and materials like a booklet on best practices and a zero tolerance poster to the employer to use in their own shop to raise further awareness. The former provides quick access to crucial information for both employers and employees, making it easy to refer to key points or address harassment-related issues as they arise.

2025-05-27

2025-08-31

We define four domains of outcomes of interest. The first domain addresses take-up, the second domain addresses awareness and attitudes towards sexual harassment, the third domain addresses improved business practices and workplace culture, and the fourth domain addresses female workforce participation.
Domain 1 (Take-up): 1) Intention to take up treatment- Does the owner agree to take up training when pitched and 2) Actual take-up of treatment-Is training administered to at least one worker of the firm/to the employer

Domain 2: : SH awareness, victim blaming, gender and risk attitudes, beliefs
This includes:
1) Information about SH (index): Knowledge of the law against sexual harassment, bystander intervention techniques, perception of prevalence of sexual harassment, and recognizing harassment
2) Attitudes towards SH (index)- Measuring victim-blaming attitudes and perception of the share of false reports

Domain 3: Workplace environment, Sexual harassment, and business practices:
1) Take-up of intervention (index)- Take-up of intervention, including putting up posters and using the booklet provided after training
2) Business practices to reduce sexual harassment (index)-Beliefs about reporting behavior, having conversations with colleagues about SH, confidence to address SH complaints, perception of employer commitment to fight SH, and actions taken to address harassment (e.g., signing a petition to form a local committee)
3) Workplace culture (index)-Workplace satisfaction, collegiality, and perception of supervisor/owner support
4) Reported incidents of sexual harassment-Faced by self and others as measured through surveys, and anonymous report platform.

Domain 4: Labor market outcomes
1) Referral of women in your network to work at the firm: Willingness of employees to refer females in their network to an internship in their firm or to work in the urban retail sector in Delhi
2) Support by employees to hire women-Perception that the workplace is attractive to women, measured via incentivized resume rating exercises comparing male and female interns
3)Openness of employers to hire women-Employers' willingness to hire women, measured by incentivized resume rating exercises comparing male and female interns
4)Female employment-Whether the firm has any female employees

We use a two-stage randomization design in our study, which we describe below.

First stage: In the first stage, to generate exogenous variation in the type of pitch firms receive, we assign treatment pitches to different clusters of firms thereby creating four treatment arms: a control arm and arms for each of the gender equality pitch, legal compliance pitch and peer recommendations pitch. Within each pitched cluster, we will approach 40 firms to join our study. Of these 40 firms, we will randomize 6 firms to be a part of a control group (CP). The control firms will not be offered any training program. These are selected to allow us to measure the presence of spillovers of treatment within the cluster, if any. The remaining 34 firms will be randomized to be offered the pitch treatment that is specific to their cluster (TP1, TP2 and TP3). See experimental design figure in the pre-analysis plan.

We anticipate roughly a 50% rate of acceptance for participation in our study and thus anticipate 20 firms per cluster to ultimately be a part of our study to reach our desired sample size.

Firm-level randomization: Among the firms that decide to take-up our training, we will randomize half of them to receive the training and the other half will be a part of the control group. This allows us to study the impact of the training, while controlling for selection into the training. We will then measure the impact of the training on downstream outcomes such as sexual harassment, employment of female workers, workplace culture and business practices, attitudes and awareness towards sexual harassment.

Timeline
The project spans 18 months. The baseline survey will be launched in May 2025, during which firms' availability for training sessions will be collected.
Shortly after the baseline, treated firms will receive a sexual harassment training intervention over 4 weeks after the baseline for each shop assigned to receive the treatment post take-up.

Post-training, a short survey will assess understanding of the content, and employers will be given informational booklets and QR codes linking to Safecity’s website. A midline survey will be conducted three months after baseline, accompanied by a Zero Tolerance poster and access to training videos for retraining purposes. The endline survey will follow six months later (nine months post-intervention) to measure longer-term effects.

Not available

Done in office by a computer

-First stage randomization is at the cluster level.
-Second stage of randomization is at the firm level.

Yes

250

5000

First stage randomization: 75 clusters per pitch treatment, 25 clusters for the control arm for a total of 250 clusters
Second stage randomization: 956 firms in T and 956 firms in C.