|
Abstract
|
In the global fight against COVID-19, scholars have pointed out the health benefits of various non-pharmaceutical interventions, such as travel ban and home isolation. However, these policies also impose large costs on society (Hsiang et al., 2020). Governments and people are considering whether the health benefits are worth the social and economic costs. Therefore, assessing the costs of implementing the control measures is essential, and this study is a vital component of this aspect.
The study aims to examine whether home isolation policy deteriorates the family relationship in China, increasing domestic violence and divorce. Domestic violence denotes aggressive, hostile, and violent behaviors that occur within families, which may be psychological, emotional, physical, or sexual (Pacheco and Rudas, 2021). Even before the COVID-19 pandemic, China displayed an alarming rate of the prevalence of domestic violence in the general population, which could reach 65% (Zhang, 2020). Scholars reported that domestic violence is closely related to divorce (Bowlus and Seitz, 2006). In 2018, the Supreme People's Court of China reported that since 2003, the divorce rate in China has gradually increased across 15 consecutive years (Zhang, 2020). Therefore, if home isolation can worsen these traditional situations, then it will undoubtedly impose great burdens on society.
With the development of the Internet, people have become accustomed to seeking related information in the face of family relationship problems. This online search behavior generates large-scale rich data, which provides an approach to overcome the limitation of the availability of data on domestic violence and divorce. Therefore, we conducted an analysis based on online search data from Baidu, the biggest search engine in China (see "Materials and Methods" in Supplementary Information for details). We collected daily data from the Baidu Index for "domestic violence" and "divorce" at the city level between January 1, 2020, and March 7, 2020. Information on home isolation, which was defined as a policy that limited residents from leaving their homes for non-essential activities, was obtained from various news reports. To estimate the causal effect, a difference-in-differences (DiD) approach was employed. The treatment group was defined as cities implementing home isolation, whereas cities without this policy served as the control group. Propensity score matching (PSM) was used to reduce bias between the treatment and control groups, which enabled accurate comparisons...
Language: en |