Workplace Harassment in India: Prevalence, Workplace Preferences and Social Media–Driven Behavioral Change
How to cite
Dam, Anaya. 2026. "Workplace Harassment in India: Prevalence, Workplace Preferences and Social Media–Driven Behavioral Change." AEA RCT Registry. AEARCTR-0018054. https://doi.org/10.1257/rct.18054-2.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Workplace sexual harassment is pervasive yet poorly measured, with existing evidence largely confined to severe incidents captured in administrative records. This project documents the prevalence of mild-to-severe harassment experiences among women in India's corporate sector and estimates the economic cost women are willing to bear to avoid them. Using a LinkedIn-based survey of women employed in small-to-midsize Indian firms, we employ a discrete choice experiment (DCE) in which respondents choose between hypothetical job offers varying across wages, skill development opportunities, remote work options, and work environment vignettes describing sexist hostility, sexual hostility, and unwanted sexual attention. We estimate women's willingness-to-pay (WTP) for harassment-free workplaces. Findings from the baseline data and DCE will inform the design of an ongoing RCT on whether social media campaigns can affect harassment experiences, well-being and labour market outcomes of women.
Intervention
At baseline and endline, we embed a discrete choice experiment (DCE) to examine how job attributes—especially work environment and harassment—affect women’s job preferences. Each respondent completes three choice tasks, choosing between two hypothetical job offers that differ only in randomly assigned attributes shown side by side. The design follows Eriksson and Kristensen, 2014 and related studies ((Folke and Rickne, 2022; Maestas et al., 2023; Mas and Pallais, 2017; Wiswall and Zafar, 2018)). Jobs vary along four dimensions: (i) wage (−5%, same, +5%, +10% relative to current wage); (ii) skill development (none, some, high); (iii) work environment (no information; women content; sexist hostility; sexual hostility; unwanted sexual attention); and (iv) remote work (none; up to two days; fully flexible).
Experimental design
Each respondent completes three choice tasks, choosing between two hypothetical job offers that differ only in randomly assigned attributes shown side by side. Attribute order is fixed; levels are randomized in Qualtrics. Following Folke and Rickne, (2022), Job A always includes a harassment vignette (randomized across the three types), while Job B’s work environment is drawn from the non-harassment categories. After the DCE, respondents report how they interpret the vignettes, allowing us to assess perceived harassment and shifts in workplace perceptions.
Design details
Cited by
No citation data available for this registration yet.