The World Health Organization (WHO) declared the novel coronavirus disease 2019 (COVID-19) outbreak, caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), to be a public health emergency of international concern on 30 January 2020 (World Health Organization 2020). Since this time, countries around the world have been affected in unprecedented ways. The number of infections has risen quickly, as has the number of deaths attributed to the virus.
Following the first diagnosed case in late 2019, there have been more than 13 million confirmed cases of COVID-19 and half a million deaths worldwide. An important focus has been on the capacity of the health system in each country to respond to the pandemic. The early experience of China, followed by countries such as Italy, Iran, and more recently the UK and USA, where health systems have quickly become overwhelmed and struggled to cope with the demand from large numbers of infected patients, has motivated countries to take drastic action and implement a wide range of proactive and reactive measures to limit the spread of COVID-19.
Collectively referred to as “containment measures,” these strategies are designed to limit community transmission between individuals. There is, unsurprisingly, a great deal of variation in the ways in which different countries have approached the pandemic, and the types of containment measures they have employed. Despite these differences, each jurisdiction’s response has involved some combination of travel restrictions and border controls, quarantine requirements, social isolation and distancing requirements, and the large-scale closure of various services, business, and educational facilities. While the primary concern of government has been to minimize the health impacts of COVID-19, attention has also turned to the other effects of the pandemic. Significant societal and economic impacts have raised questions about potential increases in organized crime (Global Initiative Against Transnational Organized Crime 2020), cybercrime (Europol 2020), and violent crime (Eisner and Nivette 2020), particularly domestic violence (Fitz-Gibbon and Meyer 2020; van Gelder et al. 2020). Conversely, reductions in opportunistic, property crimes have been anticipated (Farrell and Tilley 2020). The impact of COVID-19 on law enforcement (Lum et al. 2020), courts (Rossner and Tait 2020), and correctional systems (Kinner et al. 2020) have also been the subject of speculation and analysis. Here, the focus has been primarily on the emerging health risks and the impact of containment measures on how these agencies and systems operate.
Particular attention has focused on the need for empirical research into the impact of COVID-19 and associated containment measures on violent crime (Eisner and Nivette 2020; Peterman et al. 2020). Changes in the routine activities of individuals, limited social interaction, the potential strain on relationships, fear and anxiety resulting from isolation, and the financial stress of job losses and reduced incomes each have the potential to influence violent crime in different ways. Understanding these impacts is important to inform how to respond, particularly as the pandemic and containment measures are likely to persist for some time yet.
In this study, we use officially recorded police data from Queensland, Australia, to look for early signs that violent crime has changed in the context of the COVID-19 pandemic. In an international context, Australia is a unique site of analysis because the rapid introduction of strong social distancing regulations has coincided with the transition to winter—a time when seasonal crime rates are expected to fall. Without accounting for these seasonal effects, there is a risk that any observed short-term decline in crime may be overstated or incorrectly attributed as a consequence of COVID-19. In this study, we examine officially recorded violent crime rates for the month of March and April 2020. We conduct this analysis in Queensland because it was the first state or territory in Australia to declare a public health emergency on January 29, 2020. It is also the first jurisdiction to provide open access to up-to-date data.
What restrictions have been introduced in Australia?
Compared with other countries, the infection rate in Australia has remained relatively low. The first Australian case was recorded in January; however, the number of new cases increased rapidly during March (Fig. 1). There have been more than 7700 confirmed cases recorded in Australia (June 30) and more than 100 people have died as a result of the disease (Australian Government Department of Health 2020). Most confirmed cases acquired their infection during overseas travel, or via contact with someone who had travelled overseas, and on-board cruise ships (Australian Government Department of Health 2020).
Importantly, as a federated nation, responsibility for implementing containment measures is shared between the federal and state and territory governments, while states and territories are able to choose how to enforce them. Although there has been a strong emphasis on a consistent national approach, there has also been variation in some of the measures employed in each jurisdiction. This has included strategies such as border controls and school closures.
Containment measures have been introduced incrementally (see Ting and Palmer 2020 for a detailed timeline). In terms of a national response, the entry of foreign nationals from mainland China was banned on February 1, before incremental travel bans on Iran, South Korea, and Italy in early March. This was followed by self-isolation requirements on all travelers arriving in Australia introduced on March 16. Large, non-essential, and organized public gatherings of more than 500 people were also restricted from this date, as were indoor gatherings of more than 100 people. At the same time, social distancing requirements were introduced, which required individuals to maintain a distance of 1.5 m (almost 5 ft) from one another. The Biosecurity (Human Biosecurity Emergency) (Human Coronavirus with Pandemic Potential) Declaration 2020 was announced on March 18, followed by a further announcement that Australian borders were closed to all non-Australian citizens and non-residents effective March 20. The following day, the requirement that there be 4 m2 per person in any enclosed space was introduced. On the 22nd of March, the Prime Minister announced large-scale closures of take-away liquor outlets, licensed premises, restaurants and cafes (except for takeaway food), entertainment venues, and places of worship, which took effect the following day. Further restrictions were imposed on a range of other venues, including indoor and outdoor markets, on March 26, while limits were placed on the number of people who are allowed to attend weddings and funerals. Public gatherings were limited to two people (non-family members) from March 30, and Australians were advised that they were only allowed to leave home for essential shopping, medical needs, exercise, or for work or education.
Queensland became the first Australian state or territory to declare a public health emergency under the Public Health Act 2005 on January 29, 2020, providing the Chief Health Officer with broad powers to make directions regarding the types of restrictions that may be imposed to limit the transmission of COVID-19. Since the non-essential business, activity, and undertaking closure direction was first released on March 23, a series of revisions have been made in line with national requirements, imposing further limits on which venues and businesses may continue to operate. School closures came into effect on March 30, remaining open to the children of essential service workers. Queensland borders were closed effective March 26, with entry limited to Queensland residents, residents of border communities undertaking essential activities, and other exempt persons. Non-residents were initially required to self-isolate for 14 days after crossing the border; however, as of early April, restrictions were tightened further and only Queensland residents could cross the border. These restrictions are enforceable by law.
Why is violent crime likely to be impacted by COVID-19?
Broadly speaking, there are several mechanisms through which the COVID-19 containment measures are likely to influence violence. We limit our discussion here to the most likely short-term situational effects that are best explained by routine activities theory (Cohen and Felson 1979) and the short-to-longer-term financial and emotional pressure (General Strain Theory, Agnew 1992) that might spur a range of negative emotions which, in turn, manifest as violence as the pandemic progresses.
Routine activities
The first, and most obvious of these mechanisms, is the impact on the routine activities of individuals. Routine activities play an important role in understanding when and where crime is more likely to occur. According to Cohen and Felson (1979), crime occurs when there is a convergence of three critical elements—a motivated offender, a suitable target and the absence of a capable guardian. Crime is more common in places where these three things converge on a regular basis. Crime pattern theory combines aspects of the routine activity approach and other environmental criminological theories and focuses on how offenders may come across opportunities for crime in the course of their everyday lives (Brantingham and Brantingham 1993). Brantingham and Brantingham (1995) describe crime as occurring when the activity spaces of offenders—which concentrate around the places they visit most frequently—intersect with the activity spaces of a target, precipitated by some triggering event. Obviously, major disruptions of these routine activities may have a significant impact on when, and where, we might expect crime to occur. Containment measures are likely to mean that much less time is spent interacting with other people in public settings and much more time spent at home.
When asked, adult Australians have reported a high degree of compliance with recommended social distancing practices, including keeping distance between people, avoiding public spaces, and avoiding large gatherings (Australian Bureau of Statistics 2020c). Google’s (Google 2020) mobility data highlights how much peoples’ movement and time spent in public and private spaces has changed as a result of the containment measures. Location data reported by Google on community mobility has tracked how often and for how long people travel to different location types, compared with a baseline value (the median value for the same day of the week in January and early February). Figure 2 shows these changes over time and demonstrates that there have been significant reductions in visits to public spaces, including parks (down 43% as of April 30), retail and recreation premises (down 39%), workplaces (down 34%), and transit stations (down 60%). Not surprisingly, the time spent in residential locations has increased by 18%, which may help to increase the opportunity for interpersonal violence. These changes took effect in late March and were sustained for most of the month of April as it was not until the beginning of May that these restrictions started to be loosened (Australian Broadcasting Corporation 2020a).
Figure 3 describes the trend as of the end of April for Australia, overall, and for Queensland and shows that the trends observed nationally were not as pronounced in Queensland. Nevertheless, similar trends in mobility were observed in Queensland in the second half of March, and these changes were sustained until the end of April. That significant changes in mobility were observed in March, seemingly in advance of the introduction of major restrictions, indicating that people changed their behavior in response to the perceived threat from the pandemic before measures were formally implemented. This is not unique to Australian residents; similar patterns have been observed elsewhere as evidenced by Google’s mobility data in Europe for example, where residents—informed by the risk of infection—engaged in social distancing before the measures announced by government were implemented (Midoes 2020). In large part, this was also due to many business implementing work-from-home policies and only limiting on-premise work to essential employees.
Limiting the amount of time that people are permitted to spend outside of the home, and the number of people outside the home with whom people may have contact, reduces opportunities for interactions between potential victims and offenders. Spending significantly less time socializing in recreational areas, such as in entertainment precincts, may lead to less public forms of violence. A sizeable proportion of assaults—between 37 and 70%, depending on which state or territory—occur in public locations (Australian Bureau of Statistics 2019). We know, for example, that a significant proportion of violent crime in Australia is attributable to the night time economy and peaks on weekends during periods of high alcohol consumption, particularly in major urban centers (Miller et al. 2016). Excessive consumption of alcohol, common in these settings, is directly related to physical aggression and is a contributing factor in physical and sexual violence (Graham et al. 1998). Rates of self-reported involvement in physical aggression in and around licensed premises in the previous 3 months among thousands of patrons surveyed in nine different cities ranged from 7 to 17% (Miller et al. 2016).
Entertainment precincts are also a known hotspot for sexual violence.
Conversely, the increase in time spent at home increases the risk of violence in residential locations. According to the latest ABS ( 2019) data, between 30 and 63% of assaults and between 40 and 73% of sexual assaults normally occur in residential settings (with most jurisdictions closer to the upper end of that range). This might be expected to increase during COVID-19. There is particular concern about the increased risk of family and domestic violence, which is already a significant problem in Australia (Fitz-Gibbon and Meyer 2020; Usher et al. 2020). The most obvious risk is from intimate partner violence, which accounts for more than half of all family violence-related assaults (Australian Bureau of Statistics 2019). Former partners may pose less of a threat (at least in terms of physical violence) in periods of social isolation; however, around four in five assaults by an intimate partner are committed by the victim’s current partner (Australian Bureau of Statistics 2019). Adolescent family violence is also an area of concern and, with the closure of schools and many recreational facilities, may increase due to increased contact between young people who use violence and their parents. As well as being at a greater risk of violence in the home due to increased contact with their abuser, victims of family and domestic violence face increasing social isolation, have limited opportunities to seek help from family and friends, may be unable or fearful of seeking help due to surveillance by their partner, or face challenges with seeking assistance from formal support services due to social distancing measures preventing face to face contact (Peterman et al. 2020; van Gelder et al. 2020; Usher et al. 2020). This raises the potential risk of repeat violence.
Of course, not everyone will comply with social distancing measures. Unsurprisingly, there is some overlap between the profile of those people who are less likely to comply and those people who are more likely to be involved in crime. Specifically, Nivette et al. (2020) found that non-compliance was associated with antisocial potential, including delinquent behavior. It is plausible that these non-compliant individuals will be more likely to offend, reducing any effect from the changes to routine activities of compliant individuals.
General strain
The other potential consequence of the Queensland lockdowns is an increase in criminal—especially violent—behavior—that might occur as a result of the emotional and financial pressure of the pandemic. In Australia, there are already signs of the strain on Australian households. One in three Australian households already report being financially worse off (Australian Bureau of Statistics 2020a). Levels of anxiety are also much higher than usually observed (Australian Bureau of Statistics 2020b). The Australian Bureau of Statistics (2020a) reported that 72% of businesses have experienced a reduced cash flow that is expected to harm business activity and employment over the coming months. In fact, throughout March, these challenging economic circumstances have resulted in a 7.5% decrease in employee jobs and an 8.2% reduction in wages paid to employees (Australian Bureau of Statistics 2020c). Unsurprisingly, some employees and sectors have been harder hit than others, with a 33% reduction in employee jobs for the accommodation and food services sector and a 27% in the arts and entertainment sectors. The implication of this financial stress is not insubstantial; violence, including domestic violence, is concentrated in areas with higher socio-economic disadvantage (Hulme et al. 2019), while financial stress has been identified as a risk factor for repeat offending (Dowling and Morgan 2016).
The restrictions on movement are likely to be compounded by the negative psychological impacts of quarantine measures. A recent rapid review by Brooks et al. (2020) concluded that there are likely to be wide ranging and long-lasting psychological impacts, including a high prevalence of stress disorders, irritability, emotional disturbance, anger, and depression, among other symptoms. Recent findings from a cohort study showed increased levels of perceived stress and anger among young people during the pandemic (Shanahan et al. 2020). Data from call centers in the USA, for example, reveals substantial increases in calls to mental health and suicide prevention hotlines (Goodman 2020; Jackson 2020). Although this may not, in and of itself, lead to increased violence, people who experience these symptoms may be more likely to engage in substance misuse (Clay and Parker 2020) which, together with poor mental health, is associated with an increased risk of violence (Elbogen and Johnson 2009). Also, there is evidence of increased alcohol consumption in the home due to on-premise restrictions and increased home delivery (Foundation for Alcohol Research {and} Education, 2020). While debate persists regarding the causal role of substance use and domestic violence (Graham et al. 2011), these are situational stressors that can further exacerbate the risk of family and domestic violence.
One criminological theory that is apt to help explain a potential increase in violence, in particular violence in the home, is Agnew’s (1992) general strain theory. Recognizing that the early macro-level strain perspectives failed to take into account the social psychological processes involved when perceiving strain, Agnew reframed the theory at the micro-level. Specifically, Agnew posited the existence of three broad categories of strain: (1) failure to achieve positive goals, (2) presentation of noxious stimuli, and (3) removal of positive stimuli. Within the context of COVID-19, it is relatively easy to identify examples of each of these types of strain. For example, being furloughed from work is a good example of the first type of strain as people may not be able to pay their rent or mortgage. An example of the second type of strain is the built-up tension that may exist because individuals find themselves in cramped quarters with little room for escape. And finally, with respect to the removal of positive stimuli, lockdowns prevent individuals from visiting restaurants, bars, shopping centers, and gyms—all of which are activities people may enjoy and for which may also offer stress relief.
For Agnew, however, there is more to understanding criminal behavior than just the presence of these types of strain. Most importantly, he contends that these types of strains produce a wide range of negative behaviors within persons including anger, rage, fear, depression, anxiety, and so forth. It is the experience of these negative emotions that, left unchecked by coping mechanisms, are expected to lead to a higher likelihood of criminal behavior—but especially violence.
During the period of enforced social distancing associated with COVID-19, it is easy to envision that Agnew would predict an increase in interpersonal, domestic partner-oriented violence because of the anger and rage that may occur because people do not have much physical separation and because the victim may not be able to reach out for help because their aggressor lives with them. With respect to child abuse, there may be reasons to think there would be both an increase and a decrease in real and reported rates of abuse. Child abuse may increase for similar reasons as to partner abuse. A reduction in reported child abuse could be anticipated during school closures given teachers are the ones who are most often able to “see” the after-effects of physical child abuse. At the same time, there is limited evidence of an increase in physical injuries associated with child abuse being reported in some US hospitals due to increased presentations associated with COVID-19(Santhanam 2020). And with respect to property crime, it is expected, within the theory, that many forms of burglaries and thefts would drop because people are home and not out and about, but cybercrime and other fishing-related scams could increase as people use the internet more often.
What trends have been observed around the world?
Scholars have been quick to examine the impact of COVID-19 on violent crime, particularly in large US cities with open source crime data. Mohler et al. (2020) analyzed daily counts of calls for service and recorded crime in Los Angeles and Indianapolis for several major offense categories. They compared the period after stay-at-home orders had been issued (March 20 and March 24, respectively)—regarded as the time period when full social distancing came into effect—to the period prior to school closures. They observed no change in assault/battery calls for service in either city; however, there was a significant increase in domestic violence calls for service of between 13.6 and 15.2 events per day. Domestic violence assaults could not be distinguished from other assaults in recorded crime data. Recorded aggravated assaults (of which domestic violence will account for some) did not change. Mohler et al. (2020) also observed significant declines in burglary and robbery calls for service in Los Angeles and a small increase in vehicle theft, but no change in Indianapolis. In terms of recorded crime, only robbery in Los Angeles was significantly reduced. These findings were reinforced in a second regression that incorporated Google mobility data. Yet, their results are limited because of the short-term analysis and inability to disentangle seasonal trends and more longer-term(yearly) crime patterns.
Campedelli et al. (2020) analyzed daily recorded crime counts over a 3-year period in Los Angeles using Bayesian structural time-series models, incorporating controls for weather (seasonality) and holiday periods. This approach allowed them to produce a synthetic counterfactual—what would have occurred had there been no containment measures—which they compared with observed crime counts during a period of increasingly stringent social distancing measures in March. They found no impact on homicide, assault with deadly weapons, and intimate partner assault, but observed a decrease in robberies and battery (simple assault)—the latter only during the period with more stringent social distancing measures. Property crimes, including shoplifting and thefts, decreased, as did crime overall.
Most recently, Ashby (2020) analyzed data for 16 large US cities. Similar to Campedelli et al. (2020), Ashby forecasted the expected frequency of crime during the pandemic based on data from 2016 to early 2020, but used seasonal auto-regressive integrated moving average models to produce the synthetic counterfactual for the period after January 21 (the first confirmed case of COVID-19 in the USA). There was no change in crime levels between the observed and predicted values before early March, when social distancing measures (like closing schools and then stay-at-home orders) were introduced. In terms of violent crime, there was no impact in any of the cities on serious assaults in public places or in residences. The latter was used as a proxy for domestic violence. There was some evidence of an impact on property crime, including burglary and theft from vehicles; however, Ashby observed significant variation between the cities, with no two cities exhibiting the same trends in crime.
Finally, Piquero et al. (2020) used data from the Dallas, Texas, Police Department in order to examine the extent to which domestic violence incidents increased as a result of the city’s stay-at-home orders. Their analysis showed that there was an immediate spike in the first 2 weeks after the stay-at-home order went into effect followed by a decrease in the 2 weeks that followed. At the same time, their analysis also suggested that domestic violence incidents were trending higher just before the stay-at-home orders were implemented, which likely reflects the fact that many people were starting to stay at home before the formal orders were enacted.
In Sweden, where containment measures have been less stringent than elsewhere, Gerell et al. (2020) compared the weekly crime numbers in 2020 with the median from three previous years. There were differences between the observed and expected values for indoor and outdoor assault; however, these differences were relatively small and, in the case of outdoor assault, not consistent for all of the weeks examined. Gerell et al. (2020) also observed declines in residential burglary, commercial burglary, and pickpocketing.
Overall, there is mixed support for the suggestion that COVID-19 and associated containment measures has had an impact violent crime, and in the more robust studies using longer-term time series analysis, there was no evidence of an impact on domestic violence (Ashby 2020; Campedelli et al. 2020; though see Piquero et al. 2020). Importantly, each of these studies has focused on the initial weeks following the introduction of containment measures, meaning that it might still be too early to have detected any meaningful impact on crime. The impact on simple assault in Los Angeles only emerged in the most recent iteration of the analysis (Campedelli et al. 2020), when social distancing measures had become more stringent. The limitations of using police recorded crime data to measure certain crime types when reporting rates might also be affected, particularly for domestic violence, were also acknowledged.
How has COVID-19 impacted crime in Australia?
Most of the evidence around the impact of crime impacts of COVID-19 in Australia is based on media reports, which draw on information supplied by police agencies. Limited detailed analyses have been conducted, although as the pandemic continues, it is likely that more data will become available and studies will be published. Further, it is recognized that many law enforcement agencies and other data custodians will be conducting their own in-house analyses of crime to help inform their respective government’s response.
There is emerging evidence of the impact of restrictions on family violence and, to date, that evidence has produced something of a mixed picture. Some police jurisdictions have seen an increase in family violence incidents, others have not. Western Australian Police has reported a 5% increase (Shepherd 2020), but South Australia recently reported that their numbers have not changed (Australian Broadcasting Corporation 2020b). Victoria Police have reported 14% of incidents as being related to COVID-19 and the need for victims and perpetrators having to stay at home together (Davey 2020). But there has been no increase in recorded assaults. Most recently, and in the only published analyses that we are aware of, the New South Wales (NSW) Bureau of Crime Statistics and Research concluded there had been no increase in recorded domestic violence (Freeman 2020). Freeman (2020) acknowledges that social isolation may have impacted the willingness or ability of victims to report to police; however, trends in serious forms of violence, less susceptible to changes in reporting rates, were also stable. Still, the study did not involve any time series analysis; rather, the frequency of recorded offenses in March 2020 was compared with March 2019.
Data from other sources has provided a slightly different picture. Fifty percent of domestic violence service providers in NSW reported an increase in demand following the onset of the pandemic in February (Women’s Safety NSW 2020). But elsewhere, demand has not increased. In regional NSW, services have reported a decline in calls, raising concerns isolated victims cannot reach out for help (Tate 2020). Similarly, there have been reports of a decrease in calls to domestic violence helplines—by as much as 30% to 1800 RESPECT have been reported (Talwar 2020)—but these numbers are climbing again (Pearson 2020). Conversely, there has been an increase in calls to men’s helplines from perpetrators reaching out for help to stop them from being violent (Tuohy 2020). Finally, trauma surgeons in the Northern Territory have reported a 15% increase in surgeries related to domestic violence (Davey 2020). However, none of these data sources have been subjected to rigorous analysis.
Current study
The problem with some of the published studies conducted to date—Australian and overseas—is that they have tended to rely on short time series. The benefit of shorter-term studies is that they have used daily offense counts and because of this they are able to detect immediate, short-term changes in crime. However, analyzing short time periods does not account for longer term trends and seasonal variation that might occur, leading to masked or exaggerated effects. The same is true of studies that compare data for March 2020 with the same month in the previous year (or years). Our study overcomes these limitations by using officially recorded violent crime rates to model the series trend over 6 years between February 2014 and 2020. We then compute 6-month ahead forecasts for the various crime types and compare these with the observed rates for March and April 2020, when COVID-19 social distancing regulations were in place.