Domestic Violence During the Corona Pandemic: A Comparison of Pre-Pandemic and Pandemic Incidents of Domestic Violence in the Netherlands

Abstract

There is a widespread fear that the measures as a consequence of the worldwide corona pandemic have led to a significant increase in domestic violence and child abuse. The current study uses criminal files from all incidents of domestic violence that were reported to the police in a relatively large city in the Netherlands during 3.5 months before the pandemic and the first 3.5 months from the start of the nationwide shutdown, to compare the characteristics of the incident, the suspect and victim, how the case was handled and the involvement of minors. Results show that levels of domestic violence during the Covid-19-related restrictions did not increase compared to before the pandemic. During the pandemic, suspects were relatively more often men, the violence was less often reciprocal, more often took place inside the homes of suspects and victims, and was more often reported to the police by a witness than before the pandemic. Before and during the pandemic, minors were involved in respectively 34% and 43% of the incidents, mainly as witnesses of the violence between their parents. The results of the study may ease the concerns about the potential threat of domestic violence levels going through the roof during isolating measures such as during a shutdown. Limitations of the study are that official data from a single, but large, city were used and that we were not able to control for seasonal effects.

Introduction

Since the beginning of 2020, Covid-19 has been the most significant public health crisis in the world. More than half of the world’s population was under lockdown conditions in early April 2020 (UN Women Headquarters, 2020). As soon as the first measures to combat Covid-19 were announced, many people expressed their concerns about the consequences of the measures for domestic violence and child abuse. Ample media attention was paid to the fact that not everyone was safe at home and that isolation might be disastrous for some. At that time, concerns were based on assumptions about the effects of the measures on domestic violence and child abuse and limited to, mostly anecdotal, literature on the effects of natural disasters in general on levels of domestic violence. By now, almost 2 years after the beginning of the pandemic, we have started to learn more about the real consequences of the measures for domestic violence through empirical studies measuring levels of domestic violence during the pandemic (e.g., Piquero et al., 2021). The current study adds to this literature by comparing levels and nature of domestic violence before and during the early stages of the pandemic.

Previous Studies on the Link Between Disasters and Domestic Violence

Levels of domestic violence seem to increase during or in the aftermath of a natural disaster, like an earthquake, hurricane, tsunami, or flood. Supposed mechanisms behind this increase are all related to the consequences on personal lives of those affected by the disaster. Disasters might lead to personal threats to person’s lives and the lives of their loved ones. Additionally, social security systems are often disordered, social cohesions collapse, and people’s lives destructed (Rezaeian, 2013). Disasters might leave people injured, homeless, or unemployed (Curtis et al., 2000). Many of the usual routines and patterns of life might be disrupted, leaving people behind with increased stress, feelings of helplessness, and frustration (Miller & Kraus, 1994), ultimately leading to psychological problems, such as depression, anxiety disorder, or post-traumatic stress disorder for some. In its turn, mental distress might induce antisocial behavior, such as violence against those who are nearby.

Empirical studies on post-disaster domestic violence or child abuse are scarce, mainly anecdotal, or lack sound methodological designs. For example, some studies lack reliable quantitative data or measurement of violence before the disaster took place (see e.g., Parkinson, 2019; Subedi et al., 2019). Studies in which the relationship between being exposed to a natural disaster and interpersonal violence or child abuse is investigated present mixed results. In some studies, an association was found. For example, (partner) violence against women was found to increase in frequency and/or severity in the aftermath of a natural disaster (Keenan et al., 2004; Rezaeian, 2013). Also, after the 2004 tsunami, intimate partner violence was found to increase, especially for Indian states on the east coast that were directly impacted (Rao, 2020). In the 6 months after Hurricane Katrina in the Gulf Coast region of the United States, psychological as well as physical victimization increased among women (Schumacher et al., 2010). Some studies suggest that children are more at risk of domestic violence after a natural disaster. In the 6 months after hurricane Floyd in North Carolina, for example, the rate of inflicted traumatic brain injury (one of the most severe forms of child abuse) in children under 2 years of age showed a fivefold increase (World Health Organization, 2021). Child abuse was also found to escalate 3 and 6 months after Hurricane Hugo in South Carolina and the earthquake Loma Prieta in the bay area of California. However, such elevated levels of child abuse were not found after Hurricane Andrew in Louisiana (Curtis et al., 2000). A systematic literature review into violence against children in the aftermath of a natural disaster, based on 11 papers, confirms that no evidence of a consistent statistical association exists between natural disasters and violence against children (Cerna-Turoff et al., 2019). Reviews on studies investigating the effect of a disaster on (partner) violence also show that while some studies indeed find an effect, other studies do not or even find a decrease in levels of domestic violence (see e.g., Bell & Folkerth, 2016; Rao, 2020).

Previous Studies on the Link Between the Covid-19 Pandemic and Domestic Violence

Like natural disasters, the Covid-19 outbreak can be a risk factor for increased levels or escalation of domestic violence. Stress and anxiety can come from the virus itself, that is, the fear of infection, from the Covid-19-related restrictions, such as social distancing, and direct consequences of these measures, such as job or income loss. Moreover, due to the so-called lockdown measures, opportunities for housemates to avoid each other were limited or even absent. The pandemic-related restrictions and economic impact of the pandemic influences factors associated with domestic violence, such as stress, increased unemployment, increased financial insecurity, and maladaptive coping strategies (Brooks et al., 2020; Piquero et al., 2021).

The literature on the effects of Covid-19-related restrictions on domestic violence started to grow right after the virus outbreak and is still rapidly increasing. However, many of the studies published up to now have the same disadvantages as mentioned before: no data at all but only assumptions on the relationship between the pandemic and domestic violence (see e.g., Hoseinnezhad et al., 2021) or no comparison data available from before the pandemic (see, e.g, Vandeviver et al., 2020). In other studies, baseline data were available, enabling a reflection on changing levels of domestic violence or child abuse. For example, it was found that domestic violence reports and emergency phone calls to the police have increased since lockdowns started: 200% in Hubei, China, 30% in France and Cyprus, and 25% in Argentina (UN Women Headquarters, 2020). US police departments recorded a 10% to 27% increase in domestic violence calls in the weeks after public schools closed and stay-at-home orders were announced (Boserup et al., 2020). Based on 18 studies from different countries worldwide, a meta-analysis revealed an increase in domestic violence in response to lockdown or stay-at-home orders, with the strongest effect in studies from the United States (Piquero et al., 2021). Overall, an average of 8% increase in domestic violence was found. A systematic review based on 32 studies from several parts of the world, including ten European countries but not the Netherlands, confirms that COVID-19 has caused an increase in domestic violence cases, especially during the first week of the lockdown in each country (Kourti et al., 2021).

In the Netherlands, such effects have not yet been found. Weekly numbers of domestic violence incidents reported to the police in 2020 were not higher than in 2019. Numbers were a bit lower in 2020 (Kruisbergen et al., 2020). The same trend was found for domestic violence reported to Safe at Home (Veilig Thuis): 2020 numbers were not significantly different from 2019. Qualitative interviews with domestic violence and abuse professionals confirm these numbers: most of them did not see an increase, although they did notice more severe violence (Van Gelder et al., 2021). A difference, however, was found in those who reported the violence to Safe at Home. During and after the first lockdown, neighbors reported violence more often than before the first lockdown (Coomans et al., 2021).

Some studies outside the Netherlands confirm the finding that numbers of (reported) domestic violence did not change in the first stage of the pandemic or even found a decrease in the number of reported incidents of domestic violence (see e.g., Akkermans et al., 2020; Peterman & O’Donnell, 2020; Steketee et al., 2020). A group that was in particular affected by lockdown/stay-at-home measures, adolescents, do also show such a decrease in violence during the pandemic. In Florida, statewide trends in juvenile arrests for domestic violence decreased significantly upon school closure and increased again when schools reopened (Baglivio et al., 2022). According to a systematic review including studies on COVID-19 and child abuse in different countries, the rate of child abuse reports declined, but the school closures that isolated children at home with their possible perpetrators might have contributed to this decrease in reports (Kourti et al., 2021).

In sum, findings on the (short-term) effects of the restrictions on domestic violence are mixed. Some studies show an increase (see e.g., Boserup et al., 2020; Kourti et al., 2021; Piquero et al., 2021; UN Women Headquarters, 2020), whereas others do not show a difference between pandemic numbers of domestic violence and numbers from before the pandemic (see e.g., Akkermans et al., 2020; Kruisbergen et al., 2020; Peterman & O’Donnell, 2020; Steketee et al., 2020).

Domestic Violence and Covid-19-Related Restrictions in the Netherlands

As in many if not all other countries in the world, domestic violence is a major social problem in the Netherlands. Around 8% of all Dutch people report having been the victim of domestic violence in the last year. Most often, this violence is structural, occurring daily, weekly, or monthly. Coercive control is the most common form of domestic violence, followed by physical violence (Statistics Netherlands, 2021). Dutch numbers confirm the traditional view that those who commit domestic violence are often male, while women are more often victimized. Around 90% of the offenders is male, and 60% to 80% of the victims is female (Beijersbergen et al., 2018; Van der Veen & Bogaerts, 2010). Since July 2013, it is mandatory for professionals such as doctors and teachers who suspect a case of domestic violence to follow the so-called Domestic violence and child abuse protocol, enabling an effective response to signs of violence. Various health programs are aimed at reducing the extent and severity of domestic violence, mainly by raising awareness.

Since the first half of 2019, all over the world, measures have been taken to limit the spread of Covid-19. To interpret the results of this study, it is important to know the measures and their time path in the Netherlands during the first 15 weeks of the pandemic. The first case of a Covid-19 infection was identified on February 27, 2020. Since March 9, people have been asked to follow hygienic rules: washing hands frequently, sneezing in the elbow, and not shaking hands. On March 12, the government announced the first isolating measures to limit the spread of the virus during a press conference. People were asked to work from home if possible, and events were canceled. Three days later, additional measures were announced: schools and daycare closed (except for children from parents in essential professions), cafés and restaurants closed, sports clubs closed, and people were asked to keep at least 1.5-m distance from each other, except for those living together. A week later, a so-called intelligent lockdown started: people were urgently requested to stay at home as much as possible and only to invite three visitors at a time. Above all, those with health complaints like a cold, fever, or respiratory complaints, as well as those living with them, had to stay at home. From May 11 on, it was possible to slowly let go of some measures, starting with the (partly) reopening of primary schools and daycare. On June 1st, secondary schools followed, and cafés, restaurants, museums, and theaters could open their doors again, although only for small numbers of people. Table 1 gives an overview of the most relevant measures.

Table 1.

Overview of Measures Taken in the Netherlands Between March 16 and June 30, 2020.

March 16	Schools and daycare, cafés and restaurants and sport clubs closed, cultural institutions closed, keep 1.5 m distance
March 23	Intelligent lockdown: stay at home as much as possible, maximum of three visitors at home
May 11	Primary schools and daycare reopen, children allowed to go at least half of the time.
	Children up to 12 years of age allowed to sport outside again
June 1	Secondary schools reopen, for part of the time
	People can meet as long as they keep 1.5 m distance
	Cafés and restaurants open, with a maximum of 30 guests
	Museums and theaters reopen, with a maximum of 30 guests
June 8	Primary schools reopen totally, all children allowed simultaneously

Current Study

It hardly ever happens that the situational context provides the opportunity to study the effects of changed settings on levels and the nature of crime. The corona pandemic and measures taken to combat it, provide such an opportunity. Within days, the lives of many changed as a consequence of severe measures leading to a sudden decrease in social contacts and increased isolation. Coincidentally, just months before the pandemic, we gathered data on reported domestic violence incidents, enabling us to expand this study and gather the same type of data during the first months of the pandemic. The current study aims to describe the levels and nature of domestic violence during the pandemic and compare these to the situation prior to the pandemic. The study is motivated by three research questions:

• RQ1: To what extent has the corona pandemic affected levels of domestic violence?

Firstly, it is examined whether levels of domestic violence changed during the pandemic by longitudinally studying the number of incidents of domestic violence reported to the police in a major city in the Netherlands.

• RQ2: To what extent has the corona pandemic affected the nature and characteristics of domestic violence?

Secondly, it is examined if and to what extent the nature of domestic violence changed during the pandemic. In studying the nature of the incidents, the type of violence as well as location, reciprocity, and injuries will be taken into account. Furthermore, the characteristics of suspects and victims and their relationships before the pandemic will be compared with those of suspects and victims during the pandemic.

• RQ3: To what extent has the corona pandemic affected the levels and nature of the involvement of minors in domestic violence?

Thirdly, because minors are thought to be more vulnerable during the pandemic, it will also be examined to what extent minors were involved in domestic violence incidents and in which way they were involved.

Methods

Firstly, it is necessary to describe how common crime cases are settled in the Netherlands to explain the data used for this study. Since 2008, the Dutch public prosecution service has the possibility to offer the defendant an out-of-court settlement of the criminal case, called a penalty order, when there is sufficient evidence with regard to the offense at issue. These cases do not go to court unless the defendant does not adhere to the imposed restrictions. The rationale behind these out-of-court settlements is that simple cases do not necessarily have to be dealt with by a criminal court, and the public prosecutor can settle these cases in the shortest timeframe possible (Jacobs & Van Kampen, 2014). In line with this process, a road map is developed to quickly select frequently occurring criminal cases that do not require going to trial and can be settled out of court. Since 2011, this road map has become known as ZSM, a Dutch abbreviation for as soon as possible. More detailed, the goal of ZSM is to handle a case “as rapid, astute, selective, with one another and society-oriented as possible” (Jacobs & Van Kampen, 2014, p. 74). The underlying assumption of the objective of speediness is that a quick settlement is beneficial to all involved: defendants, victims, the police, and society. An out-of-court settlement is not an end in itself; instead, ZSM is meant as a process to decide the most preferable and meaningful way to deal with a criminal case in the shortest time frame possible. The public prosecutor can settle by imposing a penalty order or sending the case to trial. Using data from ZSM for the current study, the study is limited to the bulk of all cases, that are frequently occurring criminal cases. More severe cases of domestic violence are not handled at ZSM and cannot be settled by a public prosecutor.

To indicate a possible effect of the measures taken to combat Covid-19, incidents of domestic violence that were handled by the ZSM location in a relatively large city in the Netherlands were studied. Our data include daily incidents of domestic violence during 3.5 months before the pandemic—August 16 to November 30, 2019—and the first 3.5 months from the start of the nationwide shutdown—March 16 to June 30, 2020. At the time Covid-19 started to spread in the Netherlands, the pre-pandemic data were being gathered in the context of a study into the nature and numbers of incidents of domestic violence and child abuse handled by ZSM. Since data could only be collected at the Public Prosecutor’s Office, the lockdown measures severely limited the possibilities to expand the dataset. In between several Covid-19 waves, we were able to expand the dataset with the first pandemic period, but unfortunately were not able to also include the period in between—from December 2019 to mid-March 2020.

In this study, the definition of domestic violence is adopted from the Dutch public prosecution service: domestic violence is violence committed by someone in the victim’s domestic circle (partners, ex-partners, family, close friends). The following forms of domestic violence can be distinguished: physical, sexual, and psychological. All incidents that (1) fall under the public prosecution service definition of domestic violence, (2) took place between August 16, 2019—November 30, 2019 or March 16, 2020 and June 30, 2020 and (3) were reported to the police and handled by ZSM, are included in this study. In total, 434 incidents of domestic violence were reported and included, 206 in the pre-pandemic period and 228 in the pandemic period. For all these incidents of domestic violence the criminal file was studied. Using a comprehensive topic list, information was systematically gathered on the incident, the suspect and victim and how the case was handled. For each case, the date it took place was collected and whether the violence was physical, psychological or against property. If the violence was directed towards a person, it was also recorded whether it was reciprocal or not and if there were any light or severe injuries involved. Also, the location of the incident was collected and whether the victim, the suspect or someone else reported the incident to the police. For suspects and victims, their gender, age and what their relationship is—(ex-)partners, parent/child, other family, friends, or roommates—are gathered. For each case, it is also known whether minors witnessed the violence. Based on all this information, we were able to deduce whether minors were involved and what their role was—suspect, victim, or witness. Each case was followed-up until mid-February 2021 to see how the case was handled by the police, whether the case was settled by the public prosecutor or the court and what the decision was.

This study was authorized by the Board of Prosecutors General of the Dutch Public Prosecution Service. Additionally, the Ethics Committee of the Department of Law and Criminology, Vrije Universiteit Amsterdam, gave their consent.

Results

Longitudinal Levels of Domestic Violence Before and During the Pandemic

Figure 1 shows the weekly number of incidents of domestic violence reported to the police at the ZSM location in a Dutch city with about 500,000 inhabitants. The number of incidents reported each week fluctuates between 3 and 21 in the pre-pandemic period and between 8 and 25 in the pandemic period. In total, 206 reports of domestic violence were filed in the pre-pandemic period of 15 weeks and 228 reports during the first 15 weeks of the pandemic. No differences were found in the weekly number of incidents between the pre-pandemic and pandemic period (M = 13.7 vs. M = 15.2; t(28) = −0.90, p = .38). Also, no differences were found when the same days of the week before and during the pandemic are compared. On Mondays in the pre-pandemic period, the same number of incidents was reported as on Mondays during the pandemic (t(29) = −0.02, p = .99). Also, no differences were found for pre-pandemic and pandemic Tuesdays (t(29) = 0.30, p = .77), Wednesdays (t(28) = 0.33, p = .75), Thursdays (t(28) = −1.50, p = .15), Fridays (t(29) = −0.86, p = .40), Saturdays (t(29) = −0.23, p = .82), and Sundays (t(28) = −0.98, p = .34). The number of reported incidents seems to peak in weeks 20 and 21 (11–25 May 2020). Although this could be just a coincident and result of weekly fluctuations, this peak coincides with the reopening of the schools on May 11. However, further inspection of the data revealed that the sudden increase does not relate to an increase or decrease of incidents with minors involved.

Figure 1.

Weekly number of domestic violence incidents reported to the police.

Nature and Characteristics of Domestic Violence Before and During the Pandemic

Table 2 shows the nature and other characteristics of domestic violence incidents before and during the pandemic. There was no difference in the nature of the incidents, that is, physical abuse, psychological abuse, or destruction of property, in both periods (χ²(3) = 2.33, p = .51). For almost half of the incidents, it was registered whether or not the violence was reciprocal. Reciprocal incidents of violence were relatively more common in the pre-pandemic period than during the pandemic (42 vs. 18%; χ²(1) = 12.94, p < .01). For most incidents, the injuries of the victim were available. Most often, the violence resulted in light injuries for the victim. Only in 2% of the cases, injuries were severe, such as a broken wrist, losing consciousness, or were followed by resuscitation. No differences were found in the distribution of none, light or severe injuries between the pre-pandemic and pandemic periods (χ²(2) = 0.12, p = .94). For most incidents, information could be gathered on where the violence took place. Since most incidents occurred between (ex-)partners, their common home is the most frequent location. The distribution of locations differ between the pre-pandemic and pandemic incidents (χ²(5) = 27.07, p < .001). During the pandemic, incidents of domestic violence more often took place in the homes of suspects and victims and less often in other locations. For a large part of the incidents, it is known who reported the violence to the police. A difference was found between the pre-pandemic and pandemic period (χ²(3) = 9.01, p < .05). Before the pandemic, victims more often went to the police station to file a report, while during the pandemic, incidents were more often reported by witnesses who saw or heard the violence.

Table 2.

Nature and Characteristics of the Incident.

	Pre-pandemic	Pandemic
	N (%)	N (%)
Nature of the violence
Physical	154 (75)	163 (72)
Psychological	21 (10)	19 (8)
Destruction	8 (4)	10 (4)
Combination	23 (11)	36 (16)
Total	206 (100)	228 (100)
Reciprocal violence
No	53 (58)	85 (82)
Yes	38 (42)	19 (18)
Total	91 (100)	104 (100)
Injuries
No	43 (28)	56 (29)
Light	108 (70)	134 (69)
Severe	3 (2)	3 (2)
Total	154 (100)	193 (100)
Location
Home of suspect and victim	85 (59)	118 (65)
Home of victim	28 (19)	37 (20)
Home of suspect	1 (1)	4 (2)
Public area	10 (7)	21 (12)
Other	12 (8)	—
Multiple locations	9 (6)	1 (1)
Total	145 (100)	181 (100)
Violence reported to police by
Victim called emergency number	66 (48)	81 (46)
Witness called emergency number	49 (35)	79 (45)
Victim at police station	23 (17)	12 (7)
Victim via social worker	1 (1)	3 (2)
Total	139 (100)	175 (100)

During the pandemic, suspects were more often male than before (92% vs. 85%; χ²(1) = 4.79, p < .05), while there was no difference in the gender distribution of victims (χ²(1) = 0.10, p = .75) (Table 3). The age of suspects (t(432) = −0.13, p = .90) and victims (t(409) = 0.81, p = .42) was not different during the pandemic than before the pandemic. Suspects were 36 years of age on average during the incident, while victims were 37/38 years of age on average. For most incidents, the relationship between the suspect and the (main) victim is known. Three out of four incidents of domestic violence were between partners or ex-partners. A child was using violence against their parent in around one out of seven to eight incidents. In four percent of the incidents, a parent was suspected of using violence against their child. Other incidents involved violence between other family members, friends, or roommates. During the pandemic, the distribution of relationships between suspects and victims was similar to before the pandemic (χ²(6) = 8.69, p = .19).

Table 3.

Characteristics of Suspects and Victims.

	Pre-pandemic	Pandemic
	M (SD)N (%)	M (SD)/N (%)
Male suspect	175 (85%)	209 (92%)
Age suspect	36.2 (12.4)	36.4 (11.6)
Male victim	35 (18%)	37 (17%)
Age victim	37.7 (13.2)	36.7 (11.7)
Relationship between suspect and victim
(ex-)partners	142 (75%)	167 (77%)
Child/parent	29 (15%)	26 (12%)
Parent/child	7 (4%)	8 (4%)
Friends	8 (4%)	4 (2%)
Family	2 (1%)	10 (5%)
Roommate	1 (1%)	1 (1%)
Total	189 (100%)	216 (100%)

Table 4 shows an overview of the final decisions of the public prosecutor and the court before and during the pandemic. The public prosecutor settled 70% of the cases with a penalty order during the pandemic. The other 30% was sent to court. This distribution equals the distribution in the pre-pandemic period (χ²(1) = 0.07, p = .80). A comparison of the criminal cases settled by the public prosecutor before and during the pandemic shows that the content of the settlement differed (χ²(5) = 15.48, p < .05). During the first months of the pandemic, the public prosecutor more often settled a case with a conditional dismissal (35% vs. 21% before the pandemic) and less often with an unconditional dismissal (47% vs. 67% before the pandemic). Court decisions during the pandemic did not differ from those before the pandemic (χ²(8) = 13.68, p = .09).

Table 4.

Settlement of Cases by the Public Prosecutor and the Court.^a.

	Pre-pandemic (N = 206)		Pandemic (N = 228)
	Public prosecutor (n = 146)	Court (n = 60)	Public prosecutor (n = 159)	Court (n = 69)
Unconditional prison sentence	—	9 (15%)	—	14 (20%)
Suspended prison sentence	—	15 (25%)	—	31 (45%)
Unconditional community service	10 (7%)	24 (40%)	15 (9%)	32 (46%)
Suspended community service	—	16 (27%)	—	16 (23%)
Unconditional fine	3 (2%)	3 (5%)	5 (3%)	1 (1%)
Suspended fine	—	2 (3%)	—	1 (1%)
Unconditional dismissal	98 (67%)	—	74 (47%)	—
Conditional dismissal	30 (21%)	—	55 (35%)	—
Acquittal	—	10 (17%)	—	7 (10%)
Not yet decided	2 (1%)	5 (8%)	8 (5%)	5 (7%)
Other	3 (2%)	1 (2%)	2 (1%)	12 (17%)

^aMultiple sentences can be imposed in a single criminal case (for example prison sentence and community service). As a consequence, the sum in a column can be larger than the total.

Involvement of Minors in Domestic Violence Before and During the Pandemic

For each incident reported to the police, information from the police file was gathered on whether minors (<18 years of age) were involved. In the pre-pandemic period, minors were involved in 71 of the 206 incidents of domestic violence (34%) (Table 5). In 98 out of 228 incidents during the pandemic, minors were involved (43%). Minors were not more often involved in domestic violence incidents during the pandemic than before, although the difference almost reached significance (χ²(1) = 3.30, p = .07). In about half of these cases, only one minor was involved. In about four out of ten incidents, two minors were involved. In the other incidents, three or more minors were involved. Minors can be involved as suspects of the violence, as a victim, or as a witness. In most cases, minors had witnessed violence between their parents. The most frequent role among minors is as a witness (83% before and 84% during the pandemic). Before the pandemic, 8% was the main suspect in the criminal case, and 8% was the main victim. Four percent was the main suspect during the pandemic, and 12% was the main victim. The distribution between the different roles of minors was not different during the pandemic than before (χ²(2) = 1.89, p = .39). Before as well as during the lockdown, the ages of minors involved in the violence with different roles differ (before: F(2, 60) = 16.67, p < .001; during: F(2, 92) = 11.78, p < .001). Witnesses are younger than suspects (p < .05) and victims (p < .05). The age of minor suspects of domestic violence before and during the pandemic did not differ (t(8) = −0.07, p = .94), nor did the ages of minor victims (t(16) = −0.91, p = .38) or minor witnesses (t(128) = 0.44, p = .66).

Table 5.

Number and Characteristics of Minors (<18 years) Involved in Domestic Violence.

	Pre-pandemic	Pandemic
	N (%)/M (SD)	N (%)/M (SD)
Number of incidents with minors involved	71	98
Number of minors involved
1	36 (51%)	46 (47%)
2	28 (39%)	36 (37%)
3	5 (7%)	9 (9%)
>3	2 (3%)	7 (7%)
Role of minora
Suspect	6 (8%)	4 (4%)
Victim	6 (8%)	12 (12%)
Witness	59 (83%)	82 (84%)
Age of minors per role
Suspect	15.3 (1.5)	15.3 (2.1)
Victim	14.5 (2.1)	12.5 (5.1)
Witness	6.3 (5.0)	6.7 (5.0)

^aNumbers based on minor with the most prominent role in the case.

Conclusions and Discussion

Since the beginning of the Covid-19 pandemic, worries have been expressed on a possible effect on the nature and prevalence of domestic violence. Previous studies show an increased risk of domestic violence in the aftermath of a disaster or social crisis. In this study, we compared the levels and nature of domestic violence incidents reported to the police in a relatively sizeable Dutch city during the first 15 weeks of the pandemic to the levels and nature of these incidents in an equally long period before the pandemic. In contrast with findings in other countries, it appeared that the weekly number of reported incidents during the pandemic in a large city in the Netherlands was not different from the weekly number of reported incidents before the pandemic, leading to the conclusion that levels of domestic violence did not increase during the first 15 weeks of the Covid-19-related restrictions. This finding is in line with other studies on the effects of the restrictions on domestic violence in the Netherlands (see e.g. Kruisbergen et al., 2020; Van Gelder et al., 2021).

Most incidents reported to the police were related to physical violence, and fewer incidents of psychological violence or destruction of property. The nature of the incidents was not different during the pandemic than before. The seriousness of the violence, as deduced from the victims’ injuries, also did not change during the pandemic. During the pandemic, however, the violence was less often reciprocal as compared to the pre-pandemic period. As could be expected from the lockdown measures, violence more often took place in the homes of suspects and victims and less frequently in other locations. A difference between the pandemic period and before was also found in those who reported the violence to the police: incidents were more often reported by witnesses during the pandemic. The characteristics of suspects and victims and the relationship between them were not different during the pandemic than before, except for the fact that suspects were more often male during the pandemic than before. Before the pandemic, minors were involved in 34% of all incidents of domestic violence reported to the police. During the pandemic, this percentage rose to 43% of all incidents (which is not significant, p = .07). Most often, minors witnessed the violence between their parents. Less often, a minor was the primary suspect or primary victim of an incident of domestic violence. The role of minors, as well as their ages, were similar before and during the pandemic.

A major advantage of the current study over many previous studies is that we have baseline data available, enabling us to compare the levels and nature of domestic violence during the pandemic with an equally long period before the pandemic. Based on the available data, it is not possible to draw any conclusions on a causal relationship between the measures taken to combat Covid-19 and the levels or nature of domestic violence. Although we have a measurement from a couple of months before corona entered the world and were able to compare these data to data that is gathered in exactly the same way during the first corona lockdown, we do not have a control group of people who did not suffer from the lockdown measures. Also, there is a gap of 3.5 month between the pre-pandemic and pandemic data we have available. As a consequence, we were not able to conduct more sophisticated analyses estimating the effect of the lockdown measures on domestic violence incidents and taking into account seasonal effects on domestic violence, which means that the differences we see (e.g., location of the incident, reciprocal, gender of the suspect) might be a result of the pre- and pandemic data coming from two different periods of the year. But the similarities we see might as well be a result of a seasonal effect, meaning that there might have been differences if we had data from the same time of the year before the pandemic. A previous study found that reported domestic violence incidents tend to increase in spring (Leslie & Wilson, 2020). However, this would be under normal circumstances where people could go out, meet others, and have drinks.

A limitation of the current study is its dependence on official data. Although we included all incidents reported to the police within a specific city and a specific time frame, we do not have information on the diversity of our sample. It is plausible that some groups suffered more from the isolation measures than others due to the fact that they have, for example, smaller homes, small children, or greater financial loss. Previous studies, for example, found that belonging to a disadvantaged group predicts higher odds of intimate partner violence in the year after a disaster and that higher socioeconomic status is associated with lower odds of intimate partner violence (Rao, 2020). Furthermore, we only used information on incidents reported to the police and handled by ZSM. Since we compared reports from before the pandemic to reports during the pandemic, one might assume that a possible selection effect can be ruled out. However, it is possible that the willingness to report an incident to the police is different in corona times than before. On the one hand, it is possible that report rates are lower because victims feel more scared about calling for help because they are at home with their aggressors all the time during a lockdown. On the other hand, it is possible that report rates are higher because there are more witnesses around; neighbors are also more likely to be at home most of the time. Some self-report studies indicate that domestic/partner violence did not change or even improve for part of the study group (Jetelina et al., 2021).

The results of this study may ease the concerns about the potential threat of domestic violence levels going through the roof. Despite the limitations, the current study shows that domestic violence levels and nature did not drastically change during the first months of the corona pandemic in the Netherlands. An advantage of the current study is that data could be gathered right from the start of the pandemic, giving a close look at the effects of the measures on the short term. The downside is that the current study relies on a short window and does not determine whether these results sustain in the long term. Future studies should, if possible, also take into account the effects of the measures in the long run. Covid-19-related measures are still going on and are not likely to be over soon. Therefore, it would be recommended to closely monitor levels and nature of domestic violence and child abuse for as long as measures still apply and longer. This way, governments can take into account threats of possibly increased levels of violence in the measures they impose.