COVID-19 and the rise of intimate partner violence

Abstract

Stay-at-home policies have been implemented worldwide to reduce the spread of the SARS-CoV-2 virus. However, there is a growing concern that such policies could increase violence against women. We find evidence in support of this critical concern. We focus on Peru, a country that imposed a strict nationwide lockdown starting in mid-March and where nearly 60% of women already experienced violence before COVID-19. Using administrative data on phone calls to the helpline for domestic violence (Línea 100), we find that the incidence rate of the calls increased by 48 percent between April and July 2020, with effects increasing over time. The rise in calls is found across all states and it is not driven by baseline characteristics, including previous prevalence of violence against women. These findings create the need to identify policies to mitigate the negative impact of stay-at-home orders on women's safety.

Keywords: COVID-19; Domestic violence; Intimate partner violence; Lockdowns; Peru.

Fig. 1

Calls to Línea 100 in 2019 and 2020 Note: Red and blue lines report the number of calls to the helpline Línea 100 (averaged across all states) per 100,000 people for 2020 and 2019, respectively. Shaded areas represent the 95% confidence intervals. The (black) dashed vertical line represents the month when the national lockdown started. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Event study: Calls to Línea 100 in 2019 and 2020 Note: Each circle shows the exponentiated coefficients of a Poisson regression for the number of calls per 100,000 people. The omitted category is the month of February 2020. Controls include state fixed-effects and state-specific trends. Confidence intervals at the 95% are shown by the vertical lines and are calculated using robust standard errors clustered at the state-year level.

Fig. 3

Heterogenous effects Note: Each circle shows the exponentiated coefficients for the months of March, April, May, June and July, respectively, interacted with Year2020 from a Poisson regression for the number of calls to Línea 100 per 100,000 people. Within each figure, a separate regression is estimated for the sample above (blue) and below (red) the median of the baseline characteristic. Controls include those in column 3 of Table 1: year, state and month fixed effects in addition to state-specific and month-specific trends. Confidence intervals at the 95% are calculated using robust standard errors clustered at the state-year level. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)