The relation between the individual assessment of victims by Dutch police and repeat victimization

ABSTRACT

Dutch police officers perform an individual assessment for all victims they come into contact with. A victim’s vulnerability is assessed, based on six indicators and the employment of protection measures is decided upon. This study explored to what extent the indicators and protection measures are associated with repeat victimization. Analyses were based on Dutch police records from 146,585 victims and contained information from the individual assessment procedure and information repeat victimization within 1 year. Logistic regression analyses revealed that indicators and protection measures were significantly related to repeat victimization. The individual assessment may, therefore, contribute to the identification of victims vulnerable for repeat victimization as indicators referring to experiencing harm or a high-risk crime were related to repeat victimization. Also, the individual assessment might contribute to the protection of high-risk victims, as protection measures were found to mitigate the relation between some indicators and repeat victimization.

KEYWORDS:

• Repeat victimization
• victim directive
• police
• individual assessment

The support and protection of victims of crime has received increasing attention around the world (Thunberg, Willems, et al., 2020). In line with this, the European Parliament and Council accepted Victim Directive 2012/29/EU to establish minimum standards on the rights, support, and protection of victims of crime. The Victim Directive sets a bar for the rights of victims to ensure that ‘victims of crime receive appropriate information, support and protection’ (Art. 1). Article 22 requires Member States to:

Ensure a timely and individual assessment of victims to identify protection needs and to determine whether and to what extent a victim would benefit from special measures due to their particular vulnerability to secondary and repeat victimization, to intimidation and to retaliation.

(Victim Directive 2012/29/EU, p. 72)

The individual assessment as described in Art. 22, is arguably one of the most important aspects of the Victim Directive (Forti et al., 2018). In The Netherlands, a national program has been established for the implementation of the Victim Directive (Lünnemann et al., 2016). As a result, since June 2018, Dutch police officers use a standardized individual assessment procedure for all victims who come into contact with the police. In this individual assessment procedure, police officers assess a victim’s vulnerability and protection needs. In September 2019, the Inspectorate of Justice and Security reported that the use of the individual assessment procedure varies greatly between police officers. Victims’ vulnerability and protection needs are, therefore, not always correctly identified. Despite the partly unsuccessful use of the individual assessment procedure in practice, the Inspectorate also concludes that a considerable share of victims is correctly assessed regarding their vulnerability and protection needs. It is, however, unknown to what extent the individual assessment procedure indeed contributes to the support and protection of victims. The aim of the present study was, therefore, to investigate to what extent the individual assessment by Dutch police officers contributes to the identification and protection of victims at risk of repeat victimization.

The police play an important role in the support and protection of victims (Trujillo & Ross, 2008). In the Netherlands, annually over 800.000 victims report a crime to the police (Statistics Netherlands, 2017). Victims generally encounter police officers as the first representatives or ‘gate keepers’ of the criminal justice system: they are the starting point of further access to interventions, protection and victim support services (Russell & Light, 2006; Stanley et al., 2011). Research has shown that victimization is a highly non-random event: the number of victimization incidents is not equally distributed among victims. A small number of all victims account for a large proportion of all victimization incidents (Gabor & Mata, 2004; SooHyun et al., 2017; van Dijk, 2001). Consequences of repeat victimization are even more adverse than consequences of a single victimization incident (Beckmann, 2019; Méndez-López et al., 2021; Obsuth et al., 2018; Simmons & Swahnberg, 2021; Turner et al., 2006). Following the Victim Directive, police officers should support and protect especially those victims at risk of repeat victimization. To adequately protect victims from repeat victimization, police officers need to be able to correctly identify victims at risk of repeat victimization and systematically apply protection measures (Rasquete et al., 2014).

Repeat victimization

Repeat victimization increasingly gained interest from research and practice in the 1990s (Pease et al., 2018). The notion that repeat victimization accounts for a large share of all victimization experiences was considered as ‘probably the most important criminological insight of the decade’ (Skogan, 1996, p. 3). It was argued that investigating repeat victimization and its causes might benefit crime prevention policy (Hindelang et al., 1978), as it is a way to effectively allocate scarce prevention resources (Farrell, 2005). Indeed, research has shown that repeat victimization prevention strategies may be effective in reducing repeat victimization (Grove et al., 2012). Originally, the prevention of residential burglary was the prime focus (Forrester et al., 1990). However, repeat victimization has been demonstrated for other types of crime as well such as property, personal, or violent crimes (Farrell, 2005).

In the literature on repeat victimization, the term repeat victimization typically refers to an individual experiencing two or more crimes of the same type within a specific period (Finkelhor et al., 2007b; Weisel, 2005). Previous studies, for instance, focused on specific crimes such as sexual abuse or intimate partner violence (Finkelhor et al., 2007a; Pinto-Cortez et al., 2020; Turner et al., 2016). As the Victim Directive requires an individual assessment of all victims of crime, we focused on repeat victimization in general, regardless of crime type. We, thereby, included the possibility of poly-victimization (i.e., experiencing different forms of victimization, such as psychological, physical, peer, or sexual victimization). As repeat victimization often occurs within one year, (van Dijk, 2001), we defined repeat victimization as being victimized within one year after an index victimization incident.

Repeat victimization is often explained from two different theoretical perspectives: the population heterogeneity perspective and the state dependence perspective (Pease, 1998). According to the population heterogeneity perspective, some individuals are more prone to victimization. Individual and social characteristics such as low self-control, lifestyle patterns, and neighborhood environment might ‘flag’ suitability as a target (Clay-Warner et al., 2016; Tseloni & Pease, 2003). According to the state dependence perspective, initial victimization ‘boosts’ the likelihood of repeat victimization by setting in motion processes that facilitate repeat victimization (Nagin & Paternoster, 2000). An initial victimization incident can alter individuals or their social environments such that repeat victimization is more likely. Examples of these alterations are increased participation in high-risk activities (e.g., substance use) and mental health issues (Blokland & Nieuwbeerta, 2010; Clay-Warner et al., 2016; Turanovic et al., 2018). With regard to the latter, mental health issues such as post-traumatic stress disorder (PTSD) symptoms, anxiety, and depression can occur after victimization, and have been found to be associated with an elevated risk of repeat victimization (Gong et al., 2019; Jaffe et al., 2019; Kunst & Koster, 2017; Rancher et al., 2019; Turner et al., 2017). PTSD symptoms, as a result of victimization, may result in exaggerated ‘freeze, fight or flight’ responses, which reduce the ability to react effectively to future criminal threat (Cougle et al., 2009; Kunst & Winkel, 2013). Symptoms of depression and anxiety may signal vulnerability to potential perpetrators (Clay-Warner et al., 2016).

Individual assessment by the Dutch police

Since June 2018, an individual assessment procedure has been introduced by the Dutch police for victims of crime who come into contact with the police. It is, therefore, not necessary for a victim to make an official report to the police. To perform the individual assessment procedure, all police officers received training. Police officers who perform the individual assessment were most intensively trained: they attended a four half-days training. Aspects discussed in this training were basic knowledge regarding victim rights, how to assess vulnerability and how to protect vulnerable victims. To perform the individual assessment procedure adequately, the police force was extended substantially. The individual assessment procedure is performed at first contact between a victim and a police officer. In the case of multiple victimization incidents (i.e., repeat victimization), the procedure is performed for each registration of a victimization incident. Police officers can view the outcome of previously performed individual assessment procedures in the case of repeat victimization. The individual assessment procedure consists of two parts. First, police officers assess a victim’s vulnerability to repeat victimization. Victim’s vulnerability is assessed using six indicators that follow from the Victim Directive and Dutch legislation. Police officers register (using tick boxes) whether these indicators are present based on their own judgement, interaction with the victim, and knowledge on the victim’s police contact history in the past five years. When at least one out of six indicators is present, victims are considered to be vulnerable. Second, with vulnerable victims, police officers discuss and, when deemed necessary, employ protection measures to reduce the risk of repeat victimization.

Indicators of vulnerability and repeat victimization

The concept of vulnerability has come to play an important role in research and practice (Walklate, 2011). To provide a comprehensive overview of the concept vulnerability, including definitions and dimensions, falls beyond the scope of this study. This study follows the Victim Directive in which vulnerability is derived from 1) personal characteristics of the victim, 2) the type or nature of the crime, and 3) circumstances of the crime. Next, the Victim Directive provides a list of victim or crime characteristics that indicate vulnerability. Dutch police used this list to define six indicators of vulnerability. The assessment of vulnerability is, therefore, based on the Victim Directive and not on an empirically validated instrument. The six indicators of vulnerability are labelled as: ‘minor’, ‘hate crime’, ‘harm’, ‘high-risk crime’, ‘victim-offender dependence’, and ‘police officer risk appraisal’. Theoretically, the indicators ‘minor’, ‘hate-crime’ and ‘harm’ were expected to be related to an elevated risk of repeat victimization as the presence of these indicators may point at the development of psychological distress. Following the state dependence perspective mentioned before, psychological distress can lead to repeat victimization (Gong et al., 2019; Jaffe et al., 2019; Kunst & Koster, 2017; Rancher et al., 2019; Turner et al., 2017). The indicator ‘minor’ reflects whether the victim is aged under 18. Research shows that children and adolescents are more likely to be revictimized (Thunberg, Bruck, 2020) and to suffer more from anxiety, depression, PTSD symptoms, and anger as a result of victimization than adults (Bryson et al., 2020; Obsuth et al., 2018; Smith et al., 2020). The indicator ‘hate crime’ reflects whether the victim experienced a bias or discriminatory motivated crime. Hate crimes are committed, at least in part, out of prejudice or hatred towards the victim as a representative of a specific social group (Van Kesteren, 2016), based on, for instance, ethnic background, religion, gender, sexual orientation, or disability (Ellonen et al., 2020). Research shows that victims of hate crimes report more negative consequences, such as depression, anxiety, anger, and fear, than victims of other types of crimes (Iganski & Lagou, 2015).

Whereas the indicators ‘minor’ and ‘hate-crime’ are assumed to be related to psychological distress, the indicator ‘harm’ explicitly reflects whether the victim suffered considerable physical, mental, or emotional distress. Victimization can lead to a variety of negative consequences. However, the magnitude of these consequences may differ (Rühs et al., 2017; Turanovic, 2019). Not all victims experience emotional consequences after victimization (Shapland & Hall, 2007) and victims cope differentially with victimization (Heinze et al., 2016). It is argued that victims who suffered considerable harm are more likely to experience negative consequences of crime and develop psychological distress.

The indicator high-risk crime reflects whether the victim suffered a high-risk crime. Following the Victim Directive, Dutch police classified several crimes as high-risk, such as stalking, sexual assault, and domestic violence. Victims of such crimes have been found to be at increased risk of repeat victimization (Izdebska & Beisert, 2021; Walker et al., 2019; Ørke et al., 2020) because of the repetitive nature of these crimes due to the continued exposure to the offender. The indicator ‘victim-offender dependence’ reflects whether the victim was dependent on, or had a relationship with the offender. Studies suggest that victims have an increased likelihood of repeat victimization when the offender is important to them, when they are dependent on the offender, or when they are financially connected to the offender (Akhter et al., 2020; Ruiz & González-Calderón, 2020; Wenzel et al., 2004). In these cases, the offender’s influence on the victim is not limited to one point in time, resulting in an elevated risk of repeat victimization (Ørke et al., 2020). The sixth indicator, ‘police officer risk appraisal’, reflects to what extent police officers decide a victim to be at high risk of repeat victimization. This final indicator allows police officers to assess a victim as vulnerable without the explicit presence of one of the previously defined vulnerability indicators. Based on their own judgement and information from the victim, police officers can decide that a victim is vulnerable. Robinson et al. (2018) demonstrated that police officers are able to recognize important risk factors for repeat victimization. Regarding information from victims, several studies showed that victims can provide important and relevant information that improve risk assessment: victim risk appraisals appear to be valid predictors of repeat victimization (Bowen, 2011; Kebbell, 2019).

Protection measures

In the individual assessment procedure, Dutch police officers can employ two types of protection measures: measures related to the process of reporting a crime to the police and measures related to the aftermath of reporting a crime to the police. Several protection measures related to the process of reporting a crime (hereafter: process protection measures) can be employed. Most of these measures aim to limit victims’ psychological and emotional harm (Victim Support Europe, 2013). As psychological distress can lead to repeat victimization (Gong et al., 2019; Jaffe et al., 2019; Kunst & Koster, 2017; Rancher et al., 2019; Turner et al., 2017), these measures are believed to indirectly reduce the risk of repeat victimization. First, additional information can be provided to victims as victims often report that they did not receive sufficient information from the police (Holmberg et al., 2020; Wedlock & Tapley, 2016). Insufficient information may lead to psychological distress due to feelings of uncertainty, fear for safety, and isolation (Victim Support, 2012). A second process protection measure that can be employed is ensuring a victim’s privacy (shielding a victim’s name or address). Victims may be reluctant to report a crime to the police due to concerns about retaliation (Berg et al., 2013; Peterson et al., 2018). Ensuring a victim’s privacy may lead to less retaliation, thus less repeat victimization (Boateng, 2018). A third process protection measure is interviewing victims under special conditions such as audio-visually recording victims’ interviews or conducting the interview in rooms designed or adapted for that specific purpose. Police interviews are important for victims’ experiences and life outcomes, such as psychological distress (Tipton, 2021). Majeed-Ariss et al. (2019), for example, found that victims whose interview had been video-recorded experienced less anxiousness, were less intimidated and generally more satisfied than those whose interview had not been video-recorded.

In addition to process protection measures, aftermath protection measures can be employed. Aftermath protection measures may reduce the risk of repeat victimization by promoting emotional recovery and limiting psychological distress, or by providing physical protection. Several aftermath protection measures can be employed. First, adequate transfer of a case between police officers is such an aftermath protection measure. Adequate transfer between police officers is important as the police officer performing the individual assessment may not be the one investigating the crime. It is important to fully inform other police officers involved and ensure that they follow through with the case, resulting in higher victim satisfaction. Victim satisfaction may promote emotional recovery following a victimization incident (Kunst et al., 2015), thereby, indirectly, reducing the risk of repeat victimization. A second aftermath protection measure that can be employed is referral to victim support services. Following the Victim Directive, victims have the right to be referred to victim support services. Victim support services can support victims by providing information on the rights of victims, emotional and psychological support, and advice related to financial and practical aspects. Access to victim support services can mitigate the impact of victimization and reduce repeat victimization (Rivas et al., 2016; Voth Schrag et al., 2020). Third, police officers can apply a wide range of physical protection measures that aim to reduce repeat victimization. Protection orders (POs) are arguably the most used physical protection measure to intervene in situations of (domestic) violence (Cordier et al., 2021). POs intend to diminish contact between offenders and victims, resulting in less repeat victimization (Messing et al., 2021). The empirical evidence for the effectiveness of POs, however, is mixed. Although repeated violence is not completely prevented with POs, research does suggest that violence is reduced for at least some victims (Cordier et al., 2021).

In addition to the direct effects of protection measures on repeat victimization, there is also mixed evidence that the positive relation between risk factors and repeat victimization is mitigated by protection measures (Svalin & Levander, 2020). Some studies found an interaction effect between risk factors and protection measures in relation to repeat victimization, indicating that the association between risk factors and repeat victimization may be attenuated by protection measures (Belfrage et al., 2012; Storey et al., 2014; Svalin et al., 2017). Others, however, did not find such a moderating effect of protection measures (Storey et al., 2014; Svalin et al., 2017, 2018). It is important to study the moderating effect of protection measures on the association between risk factors and repeat victimization as it may provide information on which vulnerable victims benefit most from protection measures.

Current study

The Victim Directive and Dutch legislation require police officers to identify vulnerability to repeat victimization for all victims who come into contact with the police. As a result, in the Netherlands, police officers perform an individual assessment for all victims. To date, however, it is unknown to what extent risk appraisals based on the individual assessment are related to repeat victimization. Furthermore, previous studies have been limited by the use of dichotomous variables reflecting high or low levels of protection measures, thereby disregarding the type of protection measures employed. As a result, little is known about the relative importance of specific types of protection measures (Svalin & Levander, 2020). The present study extends previous research by distinguishing two types of protection measures (i.e., process protection measures and aftermath protection measures) and investigating their association with repeat victimization. Summarizing, the present study aimed to shed light on the extent to which the individual assessment procedure, which was introduced by the Dutch police as a result of the European Victim Directive, contributes to the identification and protection of victims at risk of repeat victimization. We aimed to answer the following research questions:

1. To what extent are indicators and protection measures from the individual assessment procedure of the Dutch police associated with repeat victimization?

2. To what extent is the relation between indicators and repeat victimization moderated by the application of protection measures?

Method

Sample

In order to answer our research questions, we analyzed official police records from the registration system used by the Dutch Police. All individuals that were registered as a victim between 1 July 2019 and 30 November 2019 in the Netherlands were included in the dataset. The most recent victimization incident in this period was coded as the index incident. Information regarding indicators and protection measures from the individual assessment procedure associated with this index incident was collected. Information about gender of the victim and repeat victimization within 12 months after the index incident was also collected. Individuals that had died within these 12 months, whose date of birth or gender was unknown were removed from the data. The anonymized dataset contained 146,585 victims, aged between 0 and 102 years (M = 42.4, SD = 18.9), 43% of these victims were female and 57% were male. Regarding the analyses on protection measures, we analyzed a subset of vulnerable victims (i.e., victims with at least one indicator checked). Of all victims in the sample, 30.8% were considered to be vulnerable, resulting in a subsample of 45,186 victims, aged between 0 and 100 years (M = 35.9, SD = 20.6), 52% of these vulnerable victims were female and 48% were male.

Measures

Repeat victimization was operationalized as a dichotomous variable, indicating whether a victim was registered as a victim within 12 months at least once after the index incident (1) or not (0). All six indicators of vulnerability (‘minor’, ‘hate crime’, ‘harm’, ‘high-risk crime’, ‘victim-offender dependence’, and ‘police officer risk appraisal’) were included as dichotomous variables, indicating whether the police officer had registered the presence of that particular indicator (1) or not (0). Two dichotomous variables pertaining to protection measures related to the process and aftermath of reporting a crime to the police, respectively, were created. These variables indicated whether (1) or not (0) at least one process or aftermath protection measure had been applied.

Analytical design

Logistic regression analyses were conducted to investigate the associations between indicators and protection measures on the one hand and repeat victimization on the other hand. First, we focused on the relation between the indicators and repeat victimization (Model 1). Protection measures and gender were included as control variables, as these variables might be associated with both the indicators and repeat victimization. Second, we explored the relation between protection measures and repeat victimization (Model 2). As protection measures could only be applied when victims were considered to be vulnerable (i.e., at least one indicator present), we only included vulnerable victims in this analysis. The indicators and gender were included as control variables in this model. Finally, we investigated whether the relation between the indicators and repeat victimization was moderated by the two types of protection measures (Model 3). Again, only vulnerable victims were included in the analysis. Indicators, process protection measures, aftermath protection measures, and the interactions between indicators and protection measures were included in this model. Multicollinearity across indicators and protection measures was not an issue as the lowest tolerance value was 0.72 and the highest variance inflation factor (VIF) was 1.49.

Results

Descriptive results

Repeat victimization within 12 months after the index incident was experienced by 11.5% of all victims and 13.7% of the vulnerable victims. Descriptive results regarding indicators and protection measures are presented in Table 1. The most common indicator was ‘harm’, 17.3% of all victims suffered considerable harm. The indicator ‘hate crime’ was the least common (0.2%). The most commonly registered protection measure among vulnerable victims was providing additional information (34.1%), followed by referral to victim support services (12.9%). In general, process protection measures were more common (37.8%) than aftermath protection measures (14.8%) among vulnerable victims.

Table 1. Descriptive Results.

		Total sample N = 146,585	Vulnerable victims n = 45,186
	Yes	%	%
Minor	11191	7.6	24.8
Hate crime	294	0.2	0.7
Harm	25342	17.3	56.1
High-risk crime	15915	10.9	35.2
Victim-offender dependence	1231	0.8	2.7
Police officer risk appraisal	6420	4.4	14.2
Protection measures	23002	15.7	50.9
Process protection measures	17070	11.6	37.8
Additional information	15389	10.5	34.1
Privacy protection	1456	1.0	3.2
Special conditions interview	500	0.3	1.1
Aftermath protection measures	6691	4.6	14.8
Adequate transfer to another police officer	594	0.4	1.3
Referral to victim support services	5827	4.0	12.9
Physical protection measures	456	0.3	1.0

Indicators, protection measures, and repeat Victimization

The results of Model 1 in Table 2 indicate that four indicators were significantly (p < .01) related to repeat victimization for all victims. The indicators ‘harm’ (OR = 0.93, p = <0.01) and ‘minor’ (OR = 0.64, p = <0.001) were both negatively related to repeat victimization, indicating that victims who suffered considerable harm and children under the age of 18 and were less likely to experience repeat victimization than victims that did not suffer considerable harm and victims aged over 18 years. The indicators ‘high-risk crime’ (OR = 1.62, p = <0.001) and ‘police officer risk appraisal’ (OR = 1.45, p = <0.001) were positively related to repeat victimization, indicating that victims of high-risk crimes were more likely to experience repeat victimization compared to victims of other crimes. Repeat victimization was also more likely when the indicator ‘police officer risk appraisal’ was ticked.

Table 2. Logistic Regression Models on Repeat Victimization for all victims and vulnerable victims.

	Model 1 (N = 146,585)			Model 2 (n = 45,186)
	B (SE)	OR	CI	B (SE)	OR	CI
Indicators
Minor	−0.45 (0.04)	0.64***	[0.59, 0.68]	−0.63 (0.04)	0.53***	[0.49, 0.58]
Hate crime	0.24 (0.16)	1.27	[0.93, 1.73]	0.13 (0.16)	1.14	[0.84, 1.56]
Harm	−0.07 (0.03)	0.93**	[0.89, 0.98]	−0.30 (0.03)	0.74***	[0.70, 0.79]
High-risk crime	0.49 (0.03)	1.62***	[1.54, 1.71]	0.28 (0.03)	1.32***	[1.24, 1.41]
Victim-offender dependence	0.13 (0.08)	1.14	[0.98, 1.33]	0.10 (0.08)	1.10	[0.95, 1.28]
Police officer risk appraisal	0.37 (0.04)	1.45***	[1.35, 1.56]	0.22 (0.04)	1.25***	[1.15, 1.35]
Protection measures
PM process	0.19 (0.03)	1.21***	[1.14, 1.29]	0.11 (0.03)	1.12***	[1.06, 1.19]
PM aftermath	0.31 (0.04)	1.36***	[1.25, 1.47]	0.28 (0.04)	1.32***	[1.22, 1.43]
Gender (ref. = male)	−0.18 (0.02)	0.83***	[0.81, 0.86]	−0.08 (0.03)	0.92**	[0.87, 0.98]
Constant	−2.05 (0.01)	0.13***		−1.75 (0.04)
−2Loglikelihood	103355.34			35301.99
Nagelkerke R	0.01			0.03

Note: CI = confidence interval; OR = odds ratio; PM = protection measures.

*p < .05, ** p < 0.1, *** p < 0.001.

The results of Model 2 in Table 2 show that for vulnerable victims both protection measures categories were positively related to repeat victimization. The likelihood of repeat victimization was higher when process (OR = 1.12, p = <0.001) or aftermath (OR = 1.32, p = <0.001) protection measures were applied. The effects of the indicators were similar to those in Model 1.

Table 3 presents the interaction effects between all six indicators on the one hand and the two types of protection measures on the other hand (Model 3) for vulnerable victims (as protection measures could only be applied for vulnerable victims). Process protection measures did not moderate the relation between the indicators and repeat victimization. In contrast, aftermath protection measures did moderate the relation between four indicators and repeat victimization. The positive associations between repeat victimization and the indicators ‘police officer risk appraisal’ and ‘high-risk crime’ were smaller when aftermath protection measures were employed. The significant interaction effect between indicator ‘minor’ and aftermath protection measures indicates that the negative association between repeat victimization and being a minor was stronger when this type of protection measures was applied. The presence of indicator ‘harm’ was associated with a lower risk of repeat victimization and this negative effect was smaller (but still negative) when aftermath protection measures were applied. In other words: when considerable harm was experienced and protection measures were applied, victims had a higher likelihood of repeat victimization compared to when considerable harm was experienced but no protection measures related to the aftermath of reporting a crime were applied. Gender was included as control variable in all models (Model 1, Model 2, and Model 3). In all models, it was shown that women have a lower risk of repeat victimization than men.

Table 3. Logistic regression model for repeat victimization: interaction effects for vulnerable victims(n = 45,186).

	Model 3
	B(SE)	OR	CI
Indicators of Vulnerability
Minor	−0.56 (0.06)	0.57***	[1.23, 1.50]
Hate crime	0.16 (0.21)	1.18	[0.97, 1.70]
Harm	−0.35 (0.05)	0.70***	[0.77, 1.79]
High-risk crime	0.31 (0.05)	1.36***	[0.64, 0.78]
Victim-offender dependence	0.25 (0.14)	1.29	[1.19, 1.52]
Police officer risk appraisal	0.30 (0.06)	1.34***	[0.90, 1.22]
Process protection measures	0.05 (0.08)	1.05	[0.86, 1.22]
x Minor	0.02 (0.09)	1.02	[0.93, 1.21]
x Hate crime	0.14 (0.32)	1.15	[0.64, 1.23]
x Harm	0.10 (0.07)	1.10	[0.62, 2.15]
x High-risk crime	0.06 (0.07)	1.06	[0.96, 1.27]
x Victim-offender dependence	−0.12 (0.17)	0.89	[0.81, 1.12]
x Police officer risk appraisal	−0.05 (0.08)	0.95	[1.36, 1.99]
Aftermath protection measures	0.50 (0.10)	1.64***	[0.51, 0.78]
x Minor	−0.40 (0.11)	0.63***	[0.65, 0.91]
x Hate crime	−1.34 (0.76)	0.26	[0.60, 1.15]
x Harm	0.24 (0.08)	1.27**	[0.06, 1.16]
x High-risk crime	−0.26 (0.09)	0.77**	[1.08, 1.49]
x Victim-offender dependence	−0.19 (0.17)	0.83	[0.68, 0.97]
x Police officer risk appraisal	−0.21 (0.09	0.81**	[0.87, 0.97]
Gender (ref. = male)	−0.08 (0.03)	0.92**	[0.51, 0.64]
Constant	−1.76 (0.06)	0.17***	[0.87, 0.97]
−2Loglikelihood	35240.66
Nagelkerke R	0.03

Note: OR = odds ratio; CI = confidence interval.

*p < .05, ** p < 0.1, *** p < 0.001.

Discussion

Dutch police officers perform a standardized individual assessment procedure for all victims who come into contact with the police. Police officers register whether a victim is vulnerable to repeat victimization based on six indicators and register whether protection measures need to be employed. To determine whether the individual assessment procedure is a valuable effort to support and protect victims, it is important to study to what extent registered vulnerability to repeat victimization is indeed associated with repeat victimization and to what extent protection measures are associated with (a decreased risk of) repeat victimization. In this study, we set out to explore the relation between the Dutch police’s individual assessment procedure and repeat victimization.

The first research question addressed the associations between the indicators, protection measures and repeat victimization. Our analyses revealed that four out of six indicators were indeed related to repeat victimization. Victims aged over 18 years were more likely to experience repeat victimization than child victims, arguably, because some studies also suggest that victimization risk peaks just after the age of 18 and in the early 20s (DeCamp & Zaykowski, 2015). Victims who suffered considerable harm were less likely to experience repeat victimization compared to victims who did not suffered considerable harm. Although receiving mixed empirical support (for an overview see Averdijk, 2011; Ousey et al., 2008), the ‘once bitten, twice shy’ hypothesis may explain this finding. This hypothesis suggests that victims may modify their behavior (e.g., less involvement in risky activities such as drinking alcohol or taking self-protection actions) after victimization to reduce their risk of repeat victimization (Hindelang et al., 1978). It could be argued that victims who experienced considerable harm are more inclined to modify their behavior to avoid future harm compared to victims that did not experience considerable harm. The indicator ‘high-risk crime’ was positively related to repeat victimization, arguably due to the repetitive nature of these crimes (Izdebska & Beisert, 2021; Stubbs & Szoeke, 2021; Walker et al., 2019; Ørke et al., 2020). Moreover, our findings suggest that when police officers register an elevated risk of repeat victimization, victims are indeed more likely to experience repeat victimization. This study confirms that police officer risk appraisals are valid predictors of repeat victimization (Bowen, 2011; Kebbell, 2019; Robinson et al., 2018).

Regarding protection measures, our findings show that both types of protection measures were positively related to repeat victimization: when protection measures are applied, victims are at increased risk of repeat victimization. This may indicate that protection measures reflect the presence of aspects related to vulnerability for repeat victimization that are not included in the six indicators in the individual assessment procedure. According to the Victim Directive, the individual assessment should take into account: 1) personal characteristics of the victim 2) type or nature of the crime and 3) circumstances of the crime. Apart from the indicator ‘minor’, the individual assessment only consists of indicators that take into account the type or nature of the crime and circumstances of the crime. Personal characteristics of the victim are less explicitly present in the individual assessment. Thus, the individual assessment mostly includes exogenic indicators of vulnerability (van der Aa, 2016). Personal characteristics of the victim, however, may be important indicators of elevated risk of repeat victimization. Individuals with intellectual disabilities, for example, are reported to be at increased risk of poly-victimization (Codina et al., 2020). Arguably, protection measures are applied when victims judged to be at high risk based on personal characteristics, and the indicators do not accurately reflect this high risk. The positive association between protection measures and repeat victimization in our study might result from the absence of indicators reflecting vulnerability due to personal characteristics in the individual assessment. Future research could explore whether this explanation holds.

With regard to the second research question, we explored to what extent the association between the indicators and repeat victimization was moderated by the two types of protection measures. In case of intimate partner violence (IPV), Belfrage et al. (2012) and Storey et al. (2014) demonstrated that repeat IPV was lower when high levels of interventions were registered in high-risk cases. We, therefore, expected that protection measures would mitigate the effect of the indicators on repeat victimization as protection measures aim to limit victims’ psychological and emotional harm which may result in a decreased risk of repeat victimization. In line with previous studies (for an overview see Svalin & Levander, 2020), our findings were mixed. Regarding process protection measures, our findings indicated that process protection measures did not moderate the relation between indicators and repeat victimization. Process protection measures also aim to protect victims from secondary victimization, intimidation and retaliation (Victim Support Europe, 2013). It could be argued that process protection measures are mostly related to preventing secondary victimization rather than repeat victimization and, therefore, were not found to moderate the relation between the indicators and repeat victimization. Aftermath protection measures seemed to serve as a protective factor in the case of a high-risk crime or a high-risk appraisal by the police officer. Regarding the risk factor ‘minor’ we found that aftermath protection measures functioned as a stronger protective factor for child victims than for adult victims. Finally, aftermath protection measures appeared to increase the risk of repeat victimization for victims who suffered considerable harm. A possible explanation for this finding may lie in the specific nature of some of the aftermath protection measures. Physical protection measures (a specific type of aftermath protection measures) entail a wide range of protection measures, including linking a special code to the victim’s location in the police control room to ensure that the control room initiates an immediate response in case of emergency and providing personal alarms and tracking devices for victims. Especially these two physical protection measures may result in a higher likelihood of a subsequent registration as a victim in the police system, as victims more easily contact the police when being revictimized.

The present study has several limitations that provide suggestions for future research. First, the measurement of repeat victimization using official police records is strongly influenced by under-reporting, as crimes are often not reported to the police (Torrente et al., 2017). Victims are even less willing to report repeat victimization when previous police experiences were evaluated negatively (Bolger & Walters, 2019; Koster, 2017). Therefore, not all victimization incidents, and especially repeat incidents, may have been reported to the police and included in the data. Farrell (2005), therefore, argue that victim surveys provide the most accurate measurements of repeat victimization and should be preferred when feasible. Note, however, that large sample sizes may be difficult to obtain when using victim surveys, as victims may be difficult to contact and reluctant to cooperate. Using official police records enabled us to include a large national sample and to gather information on the individual assessment procedure. Given the important issue of under-reporting of crimes in official police records, future studies could combine official police records with victim surveys. Official police records could provide information on the indicators and protection measures from the individual assessment procedure. A combination of police records and victim surveys could be used to assess more accurately which victims experience repeat victimization.

Second, another limitation is the time-window problem (Farrell et al., 2002). This study included victims that reported a crime to the police between July and November 2019 and measured repeat victimization incidents up to one year after the index incident. Although revictimization often occurs within one year (van Dijk, 2001), some repeat victims may experience the subsequent victimization incident beyond that time period. Our one-year follow-up window was motivated by practical considerations. Our time period began in July 2019 as we aimed to investigate to what extent the individual assessment procedure is related to repeat victimization. The individual assessment procedure was implemented in July 2018. Newly implemented procedures often encounter implementation problems (Grove et al., 2012). In the case of the Dutch individual assessment procedure, possibly not all police officers were trained to perform the individual assessment procedure before its implementation. This training issue subsided in the first year, hence the start of our selection period in July 2019. Future studies could explore different time windows (e.g., two- or three-year windows) to replicate our findings.

Third, repeat victimization was dichotomized in this study as the individual assessment procedure aims to identify victims vulnerable to repeat victimization in general, thereby disregarding the number of repeat victimization incidents. To further understand repeat victimization and possibly protect victims even better, future studies could explore the associations between indicators and protection measures from the individual assessment procedure and the number of repeat victimization incidents. Fourth, we could not fully assess to what extent registered protection measures were actually implemented. To adequately explore the effects of protection measures, it is important to have information on which protection measures have actually been implemented (Svalin et al., 2017). The individual assessment procedure is relatively new, and police officers may not always employ all protection measures they register in the police system. This might especially be the case for protection measures that depend on other organizations such as Victim Support Netherlands, public prosecution service and municipalities. However, as all police officers that perform the individual assessment procedure received an elaborate training and the majority of protection measures are not related to other organization than the police, we can assume that the majority of the registered protection measures are actually implemented. Future studies, however, could improve our knowledge on the individual assessment procedure and protection measures by exploring to what extent a difference between registered and implemented protection measures exist. We extended previous research by distinguishing two categories of protection measures (i.e., process and aftermath protection measures) as previous studies often investigated protection measures in general (Svalin & Levander, 2020). The categories process and aftermath protection measures consisted of multiple specific protection measures. Due to the low prevalence of the specific protection measures and repeat victimization, we were unable to investigate the associations between specific protection measures and repeat victimization. A final limitation is that the Dutch police registration system does not systematically contain information on individual characteristics such as biological aspects (e.g., physical vulnerability), risky lifestyles (e.g., staying out late drinking), or occupation (e.g., working nightshifts), which may be related to repeat victimization following the population heterogeneity perspective (Blokland & Nieuwbeerta, 2010; Clay-Warner et al., 2016; Gabor & Mata, 2004; Tseloni & Pease, 2003; Turanovic et al., 2018). As our models only explained a relatively small amount of variance in repeat victimization, it is likely that other characteristics are also related to repeat victimization. These characteristics could be investigated using victim surveys.

Notwithstanding these limitations, the present study extends our knowledge on the individual assessment performed by Dutch police officers and its contribution to the prevention of repeat victimization. The individual assessment procedure was introduced in June 2018, and all police officers have been trained since to perform the individual assessment (Inspectorate of Justice and Security, 2021). Our sample consisted of victims that reported a crime to the police between July 2019 and November 2019, thus one year after the individual assessment procedure was implemented. Our findings suggest that several indicators used in the individual assessment indeed point at an elevated risk of repeat victimization. Moreover, aftermath protection measures from the individual assessment procedure appear to be effective in reducing repeat victimization for some vulnerable victims. Although the individual assessment procedure is relatively new and the Inspectorate of Justice and Security concluded that the protection and support of victims is yet insufficient (Inspectorate of Justice and Security, 2021), our study suggests that information from the individual assessment procedure can be used to identify victims at risk of repeat victimization and that some vulnerable victims benefit from protection measures. Our study suggests that improving and further developing the individual assessment might be a viable method to improve the protection and support of victims as, already in its current form, the individual assessment procedure offers information on which victims are at risk of repeat victimization and which victims may benefit from protection measures. Using the information from the individual assessment in combination with, for instance, an actuarial risk assessment instrument for repeat victimization may further improve the support and protection of victims, as it may guide police officers in prioritizing at risk victims and allocate limited prevention resources to these victims. Future research should explore to what extent such efforts indeed prevent repeat victimization. Altogether, we conclude that the individual assessment is a valuable procedure for the identification of victims at risk of repeat victimization and that some vulnerable victims benefit from protection measures.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study is part of a larger collaboration between Tilburg University, Radboud University, Praktikon, and the Dutch National Police. The study was funded by the European Union (Grant number 2015ISFP7). The authors have no conflict of interests.

Notes on contributors

Marc J. M. H. Delsing

Geurts Roos is an external PhD-candidate at the Behavioural Science Institute (Radboud University Nijmegen). Her research focuses on risk assessment of repeat victimization in a policing context

Dr. Marc Delsing is a senior researcher at Praktikon in Nijmegen, The Netherlands. His research interests include child and adolescent development, the evaluation of child and youth care interventions, risk assessment, and (longitudinal) research methodology

Jacqueline A. M. Wientjes

Drs. Jacqueline Wientjes is a behavioral scientist at the Dutch Police. She works on prediction models for the prevention of victimization and juvenile delinquency.

Ron H.J. Scholte

Ron Scholte is full professor at the Department of Developmental Psychopathology at Radboud University. His research is largely practice based, and focuses on social interactions between adolescents and their social environment such as family, peers, school, and youth and mental health care.