Intimate Partner Violence and Women's Economic Preferences
How to cite
Dam, Anaya; Anderberg, Dan; Cassidy, Rachel; Hidrobo, Melissa; Leight, Jessica; Morsink, Karlijn. 2024. "Intimate Partner Violence and Women's Economic Preferences." AEA RCT Registry. AEARCTR-0014038. https://doi.org/10.1257/rct.14038-3.0
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
This project puts forward and tests the hypothesis that intimate partner violence impacts women’s economic decision-making. We estimate the causal impact of intimate partner violence on women's time preferences, cognitive functioning and risk preferences using a survey experiment. We assign respondents randomly to one of our two experimental arms. Respondents in the treatment arm are primed to think about episodes of IPV they may have experienced, using a survey instrument based on the well-established DHS and WHO instruments. Respondents in our control arm receive a placebo module. After the treatment or control module, respondents participate in incentivised activities to measure risk preferences, time preferences and cognitive functioning.
Experimental design
We leverage the sample of female respondents of an ongoing RCT in rural communities in Ethiopia. We add our survey experiment to the end of the endline survey conducted a year post-intervention. The sample consists of women from households registered and assisted by the Productive Safety Net Program (PSNP) in Amhara and Oromia regions of Ethiopia.
We assign respondents randomly to one of our two survey experimental arms. Respondents in the treatment arm are primed to think about episodes of IPV they may have experienced, using a survey instrument based on the well-established DHS and WHO instruments. Respondents in our control arm receive a placebo module. After the treatment or control module, respondents participate in incentivised activities to measure risk preferences, time preferences, and cognitive functioning.
The interviews were conducted in person by a trained enumerator using computer-assisted personal interviews from October 2023 to January 2024. Each interview took on average two hours to complete: 1.5 hours for the survey modules and half an hour for the survey experiment which was conducted at the end of the interview. Experiment modules consisted of the randomized treatment/control module (Prime 1 -- IPV experiences; Control -- placebo) and then the post-prime incentivized activities. We usethe demographic information collected in the modules before the survey experiment for balance checks and descriptives. Finally, women who had not received Prime 1 were also asked the IPV module at the end of the incentivized activities. This implies that we have data on IPV from all participants, but the risk, time and cognitive tasks were not primed for those women who were not assigned to the treatment arm.
Design details
Cited by
No citation data available for this registration yet.