As a novel pandemic officially declared by the World Health Organization (WHO) on March 12, 2020, COVID-19 has spread across the globe with an exponential increase of affected cases (1, 2). Until the end of February 2021, the number of confirmed cases of COVID-19 worldwide reached approximately 112 million, which includes nearly 2.5 million deaths (3). With such a severe situation, many countries declared stringent policies to constrain people’s mobility and flatten the epidemic curve, which holds significant implications for public health. Although ample research has evaluated the necessity and effectiveness of such policies to reduce the infection rate of the pandemic (4, 5), relatively scarce attention has been paid to the effects of such policies on public emotions. This aspect prevents governments and scholars from comprehensively understanding the impacts of policies to enable appropriate adjustments.

Taking a retrospective perspective, prior studies have illuminated that there is a substantial negative association between stringent lockdown policies and public mental health during a public crisis. For instance, Wu et al. compared the mental condition of health workers who have been quarantined during the SARS with those who have not and found that those who endured the quarantine were 2–3 times more likely to suffer from posttraumatic stress than their colleagues (6). Moreover, Jeong et al. investigated the impact of being isolated on people’s mental status during the MERS in Korea. They found that even at 4–6 months after they were released, anxiety symptoms derived from such experience can still be observed (7). Other research is also in line with the finding that there are significant negative effects on mental health conditions due to official quarantine policies and such effects can range from immediate impacts to long-term results (7–9).

Regarding the COVID-19 that we are facing, ample empirical research also backs up this association. Studies reveal that the level of distress or anxiety in population can significantly grow during a pandemic (10). Among the limited discussions, Qiu et al. found that many respondents suffered from anxiety symptoms or psychological distress during the COVID-19 pandemic in China (11). A study conducted in Kolkata of India further found that many people living with depression during COVID-19 pandemic were also experiencing an increased sense of helplessness and hopelessness about the future (12). Herat explored the impact of lockdown using a linguistic approach to analyze newspaper contents and suggested that people felt more anxious during the quarantine (13). Casagrande et al. conducted an online survey based on Italian people affected by lockdown restrictions (14). The study demonstrated that a majority of respondents reported worse quality of sleep and higher levels of distress compared with the time before the quarantine. Similarly, De Vos also called for attention on the research on the negative impact of social distancing policies on personal wellbeing (15). Gupta et al. comprehensively summarized that lockdown programs can cause pervasive anxiety, distress, and other negative emotions by augmenting the sense of aloneness (16). Furthermore, prior researchers have also identified several indicators that make the quarantine experience to become a stressor for people’s emotion and mental health, which include anxiety, depression, hopelessness, and helplessness (17).

This study draws on Cacioppo’s social isolation theory, which argues that social isolation could activate the neurobiological mechanisms that lead to people’s self-preservation in the short-term and undermine their mental health and increase the rates of morbidity and mortality in the long-term (18, 19). Based on a wide range of experimental and observational studies in epidemiology, the social isolation theory argues that social interaction is a crucial condition for human beings’ survival and prosperity. Specifically, substantial research demonstrates that social isolation has large negative impacts of on physical and mental health, which are equivalent to the effects of smoking, sedentary lifestyle, hypertension and obesity (18, 19). In contrast, ample evidence suggests that regular interactions in various social settings such as workplace are shown to reduce mental stress and thus improve people’s health and wellbeing (20, 21).

Therefore, in the present study, we pay particular attention to the impact of social isolation among those factors of making the quarantine experience a stressor for people’s emotion and mental health. That is to say, the effect on people’s psychological situation due to the limitation of mobility in their daily life is directly caused by the quarantine and closely associated to social isolation and loneliness. A plausible analogy for this factor is incarceration, which has been well researched by scholars. Evidence shows that when people are confined in a limited space, such a dramatic change can lead to a negative repercussion for people’s mental health by decreasing their connections with others and increasing their stress and anxiety (22). While compared with incarceration, people do not need to meet violence and lack of privacy under the quarantine, the loss of mobility to the outside still serves as a concern for the public emotion. Thus, we theoretically consider that the stringent policy increases people’s negative emotions, which triggers the concern of public mental health, and the pathway is partially mediated by the reduction of people’s mobility in their daily life (Figure 1).

Using large-scale longitudinal data from 120 countries from January to July 2020 (N = 9,141 country-day observations), the study aims to fill important gaps in the previous research and achieve two research objectives. First, while a large number of existing studies applied the snowball sampling strategy to recruit online respondents, or focused on particular groups of people (23–27), this study aims to analyze the longitudinal and dynamic association between lagged lockdown policy stringency and public negative emotions, thereby improving the external validity, and providing more comprehensive policy implications. Second, while previous studies have primarily focused on the association between lockdown and public emotions but did not explore the mechanisms, this study aims to test population mobility as a mechanism underlying the relationship between the lockdown policy and people’s negative emotions, deepening our understanding of how worldwide lockdown measures affect public emotions.

The COVID-19 pandemic, together with many other factors, has led to increased adverse mental health issues in various countries (32, 33). To the best of our knowledge, this is the first study to provide substantial evidence concerning the effects of lockdown policies on people’s emotions while also deepening our understanding of the underlying mechanisms associated with these emotions based on large-scale, longitudinal cross-country data. In conclusion, this study yields two important findings.

Consistent with previous studies in Nepal, Italy, and India that used small and localized samples (16, 32, 34), our study further confirms that stringent lockdown policies could have a strong negative correlation on a range of public emotions, such as anxiety, depression, hopelessness, and helplessness. In addition, this study extends the previous literature on the relationship between lockdown policies and negative emotions in two important ways. First, this study used longitudinal data and lagged variable models to demonstrate that the negative impact of lockdown policies can last about 20 days, revealing the dynamic relationship between lockdown policies and negative emotions. Second, by analyzing data in 120 countries this study extends beyond any particular countries or population groups, and thus can be generalized to other countries and holds significant implications for the global pandemic response-lockdown policies designed to limit the spread of the COVID-19 but may have other negative and unintended consequences on people’s emotions. While tackling the COVID-19 pandemic, future public health policies should take into account the negative spillover effects of lockdown measures (35).

As another contribution to the previous literature, this study revealed a crucial mechanism underlying the negative impact of lock-down policies by demonstrating that the negative effects of lockdown policies can be significantly mediated by increased mobility in residential areas. This suggests that changes in population mobility in residential areas is an important mechanism through which lockdown policies lead to negative emotions among the public. Previous studies in Nordic and South-European countries have reported that the ability and opportunity to spend time outdoors may be especially important under conditions of lockdown and social-distancing, and long-time stay-at-home could lead to mental health problems (36, 37). In the present study, the finding is overall consistent with the previous studies and suggests that people who stay at home or spend a considerable amount of time in residential areas tend to develop depression, loneliness, and other mental health problems (38). However, there is still a sizable proportion of the effect that cannot be mediated by changes in population mobility. This is likely due to the direct effect of lockdown policies on people’s emotions, as well as other potential mediators, such as changes in lifestyles and working environments.

To summarize, this study makes two important contributions to the previous research. First, our study is based on a panel dataset that covers about 120 countries in 4 months, which allows us to expel unobservable confounders and identify the causal relationship. In contrast, a large number of existing studies applied the snowball sampling strategy to recruit online respondents, distributed online surveys on social network, or focused on particular groups of people, such as medical workers, children, and senior citizens in particular countries or regions (23–27). Second, previous studies have primarily focused on the association between lockdown and public emotions but did not explore the mechanisms. This may limit the in-depth and comprehensive understanding of how lockdown measures affect public emotions. We propose a clear mechanism underlying the relationship between the lockdown policy and people’s negative emotions. Specifically, we argue that the negative impacts of lockdown on negative emotions could be partially explained by people’s restricted mobility. Thus, this study fills an important research gap by demonstrating how lockdown measures affect public emotions.

This study also has some limitations that should be further explored in future research. First, the results of this research are based on country-level data and thus cannot be generalized to the individual level for the potential ecological fallacy. It would be advantageous for future research to combine individual- and country-level data in order to examine the health effects of lockdown policies. Second, this study only tested the mediation effects of population mobility. There may be other mechanisms through which lockdown measures could influence public emotions, such as decreased social contact or changes in working or neighborhood environments (39, 40). Third, this study only explored the impact of lockdown policies on people’s emotions. Future research could explore the impact of lockdown measures can be moderated by other variables, such as employment status or participation in leisure activities (41, 42). Finally, while this study only focused on public emotions due to limited space, future research could profitably study how lockdown measures affect suicidal risks, health behavior such as alcohol consumption, sedentary lifestyles, or other mental disorders etc.

The study contributes to the literature in three important ways. First, the study has much higher external validity than previous research and is able to generalize the findings beyond a single country by drawing on a sample of 120 countries. Second, the study helps gain a better understanding of the dynamic relationship between lockdown policies and public emotions by using rich panel data. Third, the study provides important insights into the mechanisms through which lockdown policies affect public emotions by exploring the mediating effects of population mobility. Overall, the findings demonstrate a strong negative effect of stringent lockdown policies on public emotions, thus highlighting the important mediating role of population mobility in public and residential areas. The study calls for further policy attention to the negative unintended consequences of lockdown measures on public emotions.

Publicly available datasets were analyzed in this study. This data can be found here: Data are derived from the Global Database of Events, Language and Tone (GDELT). It is one of the largest open-access spatio-temporal datasets that identifies the people, organizations, themes, and events that drive society in broadcast, print, and web news in over 100 languages on a daily basis. Further enquires can be directed to the corresponding authors.

YC, SW, and WGu led the study and drafted the manuscript. MZ, GJ, SW, WGo, and YC collected, summarized, and analyzed the data. MZ, GJ, SW, and WGu contributed to sections of the manuscript. All authors reviewed and revised the manuscript, read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.