The issue of sexual harassment in the workplace remains pervasive, despite being illegal and a violation of human rights. Approximately one in two women experience sexual harassment during their work lives, with half of them experiencing three or more incidents. However, reporting of harassment is often hindered by stigma, psychological costs, and fear of retaliation. There remains a lack of evidence on what works to reduce harassment faced by women in the workplace and the mechanisms behind it.
In this research we ask: To what extent can a social media campaign targeted at changing men’s harassment behavior towards women reduce women’s harassment experiences in the workplace? Does the reduction in harassment due to the intervention affect well–being and labour market outcomes of women? On one hand, social media is a promising tool for intervention given its ability to target specific audiences at a large–scale, and if effective can be a low–cost tool for behavioral change. On the other hand, social media is a ‘crowded’ space with both information and disinformation as well as limited attention span of users. Therefore, although promising its effectiveness remains ambiguous.
Why do men perpetrate harassment towards women in the workplace? One possibility we explore is that men underestimate the cost of sanctions for sexual harassment conditional on being reported. This would mean underestimating the probability of being reporting or the cost of sanctions itself, such as, dismissal, social disapproval from society and peers. We explore this hypothesis with an intervention that provides information on increased reporting and disapproval from society, should update their prior beliefs and ceteris paribus reduce harassment perpetrated. We term this the ‘fear of being caught’ hypothesis. Another possibility is that men are aware and correctly perceive the costs of sanctions conditional on being reported but underestimate the consequences of their action faced by their victims. The second hypothesis we test is that if men are informed about the psychological, monetary and physical cost faced by women after experiencing harassment, this will increase the weight they place on their other–regarding preferences. We term this the ‘empathy’ hypothesis.
A final objective of our study is a descriptive one. That is uncover and address less–severe harassment behaviors which are more malleable to behavioral change interventions but usually not captured in administrative data – How prevalent are mild to severe forms of harassment? What is the acceptability of such forms of harassment? What are the perceived costs and barriers to reporting? Do workplace and societal norms constrain reporting and/or are these norms (mis)perceived by victims and perpetrators?
We answer our research questions using a cluster randomized control trial in India, where around 800,000 male employees in firms that are randomized to the treatment groups are targeted with an as–it–is– consumed social media advertisement via LinkedIn. The content of the advertisement is regarding sexual harassment faced by women in the workplace. Men who are in firms randomized to the control group do not receive the campaign. We test our two mechanisms in a multiple treatment arm design wherein the content of the ad shown is minimally adjusted to the two hypotheses above. By similarly targeting a survey to about 750,000 women in corresponding firms, we will collect novel data on about 4000 women’s workplace harassment experiences, attitudes, workplace and social norms towards harassment, labour market experiences and well–being. We also collect data from 2000 men through the sponsored messaging feature of LinkedIn on attitudes, workplace and social norms towards harassment.