• 1 Introduction

  In the last decade, victim protection has rightly received political attention. On European Union (EU) level the Council Resolution of 10 June 2011 advocated the active protection of crime victims as ‘a high priority for the European Union and its Member States’.¹x Council Resolution of 28 June 2011, OJ 2011 C 187/01. A previous call to both the Commission and the Member States to improve effective legislation and practical support measures for victim protection,²x Council Document 17024/09, OJ 2019. was followed by a joint directive from the European Parliament and the Council.³x EP and Council Directive of 13 December 2011, OJ 2011 L 338/2. This directive applies to:

  protection measures that aim specifically to protect a person against a criminal act of another person, which may in any way endanger that person’s life or physical, psychological, and sexual integrity, for example by preventing any form of harassment, as well as that person’s dignity or personal liberty, for example by preventing abductions, stalking, and other forms of indirect coercion, and which aim to prevent new criminal acts or to reduce the consequences of previous criminal acts.

  As for the Netherlands, there has indeed been an increased recognition and development of legal protection measures in recent years. This recognition generally stems from a social and political call for strengthening the legal position of victims of crime.⁴x N. Elbers, S. Meijer, I. Becx, A.J.J.G. Schijns & A. Akkermans, ‘The Role of Victims’ Lawyers in Criminal Proceedings in the Netherlands’, OnlineFirst European Journal of Criminology 1 (2020). It was particularly promoted by the normative EU framework of the Directive 2012/29/EU. This directive provided the Member States with minimum standards for the rights, support and protection of victims of crime. In the Netherlands, where the directive was implemented in 2017, a wide variety of victims’ rights has been formalised in legislation and policy.⁵x A.K. Bosma, M.S. Groenhuijsen & M. de Vries, ‘Victims’ Participation Rights in the Post-Sentencing Phase: The Netherlands in Comparative Perspective’, 12 New Journal of European Criminal Law 128 (2021). Besides rights to be informed, to speak at criminal hearings, or to make a victim impact statement, the directive also concerns the right to be protected from secondary and repeat victimisation. In this respect, the legal modalities for the protection of victims have been expanded. Alongside new modalities based on administrative law and civil law, the expansion mostly concerns modalities for imposing penal protection orders (PPOs) on the crime suspect or convict. In general, these orders are aimed at protecting the initial victim, primarily from secondary and repeat victimisation but also in a broader sense from any danger for the victim’s dignity or psychological and sexual integrity.⁶x S. van der Aa, ‘Protection Orders in the European Member States: Where Do We Stand and Where Do We Go from Here?’, 18 European Journal of Criminal Policy and Research 183 (2012). Previous research clearly shows a need for protection among victims (of violent crime), especially when the perpetrator is known.⁷x A. ten Boom and K.F. Kuijpers, ‘Victims’ Needs as Basic Human Needs’, 18 International Review of Victimology 155 (2012). These victims often suffer from repeat victimisation and escalation of violence, as acknowledged by the legislator.⁸x H.C.J. van der Veen and S. Bogaerts, Huiselijk geweld in Nederland. Overkoepelend syntheserapport van de vangst-hervangst-, slachtoffer- en daderonderzoek 2007-2010 (2010).^,9x Ministerie van Volksgezondheid, Welzijn en Sport, Factsheet Huiselijk Geweld (2013). Currently, the Dutch PPOs can be imposed at every stage of the criminal trial process, including the execution of judicial decisions. Either as a stand-alone penal sanction or as part of a conditional sanction modality.¹⁰x T. Fischer, I. Cleven & S. Struijk, Handhaving en veiligheid bij strafrechtelijke contact-, locatie- en gebiedsverboden ter bescherming van slachtoffers (2019).
  In this article, we describe the Dutch legislation and regulations regarding PPOs in four specific legal modalities. These modalities are the criminal behaviour order (Art. 509hh Criminal Procedural Code (CPC), the suspension of remand (Art. 80 CPC), the conditional sentence (Art. 14a Criminal Code (CC)) and, lastly, the freedom-restricting measure (Art. 38v CC). Given the earlier mentioned EU normative framework for victim protection we also examine which aims the legislator had in mind when developing and shaping these modalities and the subsequent possibilities to impose the protection orders. Naturally though, protecting victims of crime does not stop with legislation. The effectiveness of PPOs depends largely on how they are implemented and enforced in practice. Recent studies have shown that the actual protection of victims by means of protection orders still has many limitations. For example, the study by Fischer, Cleven & Struijk shows large differences between the numbers of violations reported by victims and the numbers of violations registered in the files of the public prosecution and probation service.¹¹x Fischer et al., above n. 10, at Table 7.1.^,12x See also: S. van der Aa, K. Lens, F. Klerx, A. Bosma & M. van den Bosch, Aard, omvang en handhaving van beschermingsbevelen in Nederland. Deel 1: Wettelijk kader en handhaving (2012). This discrepancy indicates that with many protection order violations, victims experience a lack of formal criminal justice responses. Moreover, the study shows that important differences between victims and professionals arise regarding the intended aim of protection orders when interpreting the concept of effectiveness in terms of safety perceptions.¹³x Fischer et al., above n. 10. Professionals tend to focus on objective safety (i.e. less repeated victimisation), whereas victims view effectiveness in a broader, more subjective manner (i.e. including their perception of safety).
  Obviously, professionals such as probation or police officers have a certain discretion in making enforcement decisions.¹⁴x Van der Aa et al., above n. 12. Other factors and considerations besides evidence may also play a role in their willingness to report violations of a penal protection order. The question arises whether there is a discrepancy between the legislative aims of the PPOs and the implementation in practice. If so, is this due to limited possibilities and scarce resources to actually meet the legislative expectations, or are there more fundamental causes? This important question, which we will address in this article, corresponds to the results of many studies discussing the dilemmas of the individual frontline professionals working in public services. The influential study by Lipsky was the first to provide an extensive analysis on this issue, thereby introducing the nowadays famous notion of ‘street-level bureaucracy’.¹⁵x M. Lipsky, Street-Level Bureaucracy: Dilemmas of the Individual in Public Services (2010). For Lipsky, this notion was largely determined by coping mechanisms referring to the general responses that professionals develop to deal with the challenges brought about by inadequate resources, few controls, indeterminate objectives and discouraging circumstances.¹⁶x Ibid., at 82. This focus on coping mechanisms retains great significance for socio-legal studies because of what these mechanisms can tell us about the ‘how’ and ‘why’ of the implementation of law and policy.¹⁷x S. Halliday, N. Burns, N. Hutton, F. McNeill & C. Tate, ‘Street-Level Bureaucracy, Interprofessional Relations, and Coping Mechanisms: A Study of Criminal Justice Social Workers in the Sentencing Process’, 31 Law & Policy 405, at 406 (2009). Moreover, the reasoning of frontline work professionals about their coping strategies may give insight into the agency narratives that apply to the mechanisms in the enforcement of protection orders. In this respect we will distinguish between the state-agent and the citizen-agent narrative.¹⁸x S. Maynard-Moody and M. Musheno, ‘State Agent or Citizen Agent: Two Narratives of Discretion’, 10 Journal of Public Administration Research and Theory 329 (2000). Both narratives may help to understand the motivations and considerations of frontline workers judgements and actions.
  Since Lipsky’s seminal work, numerous empirical studies have been published on this topic¹⁹x L. Tummers, V. Bekkers, E. Vink & M. Musheno, ‘Coping During Public Service Delivery: A Conceptualization and Systematic Review of the Literature’, 25 Journal of Public Administration Research and Theory 1099 (2015). examining various public functions, including the police officer, probation officer, public prosecutor and judge. Yet, extant studies are predominantly conducted within their respective professional fields.²⁰x Halliday et al., above n. 17, at 406. In this explorative article though, we go beyond the field-level focus and examine the possible discrepancies and tensions between the aims and outcomes of PPOs using an integrated approach that includes all relevant professionals in the criminal justice chain. Our research question is whether and to what extent conditions are present that complicate the enforcement of PPOs. Subsequently, which coping mechanisms are applied by different professionals regarding this enforcement in order to meet the legislative aims for effective victim protection.
  To answer these questions, we will first analyse and describe the legal framework on Dutch PPOs in the earlier mentioned four legal modalities (Section 2), followed by a discussion of the theoretical framework on ‘street-level bureaucracy’ (Section 3) and our methodology (Section 4). After having presented the results of our study (Section 5), the article concludes with a discussion and some final remarks (Section 6).

• 2 Legal Framework

  The Dutch legal possibilities for imposing a penal protection order on a suspect or convict have substantially been increased throughout every stage in the Dutch criminal process. The corresponding legal modalities vary from the very first stage prior to detention on remand to the actual detention on remand, and from the sentencing procedure to the execution of the sentence and, correspondingly, the detainee’s conditional release. Especially the latter forms of PPOs – imposed as back door sentencing as opposed to front door sentencing²¹x N. Padfield, ‘Legal Constraints on the Indeterminate Control of “Dangerous” Sex Offenders in the Community: The English Perspective’, 9 Erasmus Law Review 55 (2016); N. Padfield, R. Morgan & M. Maguire, ‘Out of Court, Out of Sight?’, in M. Maguire, R. Morgan & R. Reiner (red.), The Oxford Handbook of Criminology (2012) 955. – have recently been increased.²²x S. Struijk, ‘Vrijheidsbeperking na detentie: in hoeverre zet de rechter de achterdeur open?’, 6 Sancties 353 (2018); Bosma et al., above n. 5. To protect victims as effectively as possible, there is also a development concerning the earliest possible stage in the criminal process. Impelled by the aforementioned EU Directive 2012/29/EU, a new ‘individual assessment’ policy was implemented in the Netherlands.²³x Stb. 2016, 310. Since the actual implementation on June 1, 2018 the policy compels the police to make an individual assessment of the victims’ risk, vulnerability and protection needs. At later stages, the policy also enables the Public Prosecution Service, the judiciary, the Probation Services and Victim Support services to systematically assess the vulnerability of each victim and to determine whether protective measures are necessary during the criminal process. In this regard, both the conceptualisations of victims’ risk and protection needs to refer to secondary victimisation, repeat victimisation, intimidation and retaliation.
  The legal framework shows that there is no uniform aim of PPOs. Clearly the focus is on protecting citizens against crime, particularly from second and repeat victimisation. Yet, there may be other aims involved such as the protection from intimidation and retaliation or the reduction of the consequences of previous criminal acts on victims. Given this plural aim of PPOs and the corresponding discretion for the executive professionals, the question arises how these professionals act and decide regarding the orders’ enforcement. To address this issue, we interpret the aforementioned four legal modalities for imposing a penal protection order and reflect on the extent to which they correspond to their plural aim.

  2.1 Criminal Behaviour Order

  Since 2010, the public prosecutor is permitted to impose a so-called behaviour order upon a suspect of a public order offence, especially in case of a high risk of recurrence or incriminating behaviour towards a person (Art. 509hh CPC). Specifically concerning victim protection, this order may consist of a ban to communicate or be in a certain area with a specific person (Art. 509hh CPC). The behaviour order is generally aimed at preventing or stopping people from causing public nuisance and vandalism.²⁴x Kamerstukken II 2007/08, 31467, nr. 3, at 18. Although the legislator primarily focused on hooligans and other troublemakers, the behaviour order was also explicitly designed for victim protection. More specifically, the order may be imposed to protect victims from (further) harassment and other seriously incriminating behaviour by an offender.²⁵x Kamerstukken II 2007/08, 31467, nr. 3, at 1-2 & 29. Although these protective measures occur at an early stage of the criminal proceedings, the order cannot be imposed without meeting the legal requirement of the suspicion of a concrete criminal offence. The legal system also provides for additional victim protection stating immediate execution of the behaviour order contrary to the usual legal procedure. Moreover, violation of the behaviour order legally constitutes a new criminal offence for which the offender, if reported by the police officer, may be held in custody and taken to court (Art. 184 CC). In this respect, violating a behaviour order specifically based on victim protection even constitutes an aggravated offence (Art. 184a CC).

  2.2 Suspension of Remand

  As in most other countries, the Dutch CPC offers the possibility to conditionally suspend pre-trial detention (Art. 80 CPC). Both the suspension and the attached conditions are a judicial decision. The conditions are diverse but always related to one of the legal grounds for the suspension.²⁶x Kamerstukken II 1913/14, 286, at 3 & 84-85. As these grounds include serious flight risk or serious risk for public safety (Art. 67a CPC), it is clear that victim protection as such is not one of them. In fact, both this legal modality and the attached conditions were originally not designed for victim protection, but for adequate behavioural change of the offender with renewed pre-trial detention as a threat of non-compliance. Currently, the conditions attached to the suspension of remand may explicitly be aimed at protecting the victim from repeat victimisation. This aim is not due to changed legislation but based on the current policy of the Public Prosecution Service.²⁷x Aanwijzing voorwaardelijke straffen en schorsing van voorlopige hechtenis onder voorwaarden (Stcrt. 2020, 62553). Protection orders such as a ban to communicate with the victim²⁸x In some countries known as a ‘no contact’ provision. are explicitly part of the attached conditions. In order to enhance the effectiveness of such a ban, the Public Prosecution Service states in its policy that it is important to combine the ban with a provision to stay away from the living environment of the victim and/or other specific areas near the victim’s residence. The offender’s compliance of such a ‘stay away’ provision may be enforced by the possibility of electronic monitoring. Given the strong infringement this monitoring causes on the personal life of the offender, not only a positive advice of the Probation Service is needed but also a careful consideration of proportionality. In this consideration, the interests of the victim need to be accounted for.²⁹x Stcrt. 2020, 62553, above n. 27.
  Although electronic monitoring is not restricted to certain offences, the Public Prosecution Service is explicitly considering electronic supervision in cases of serious domestic violence to prevent repeat victimisation. Hence, the threat that victims (may) experience is accounted for.³⁰x Ibid. The legislator has also acknowledged this fear of victims by granting all decisions regarding the suspension of remand to be immediately effective (Art. 86 CPC). Besides the immediate effect, the effectiveness of a protection order within this legal modality is also enhanced by the regulation that the suspect must declare his compliance with the imposed conditions (Art. 80 par. 1 CPC). This regulation is in contrast to the earlier mentioned legal modality of the criminal behaviour order where such a commitment is not required in advance. For a suspect, whose pre-trial detention was suspended, violating the imposed protection order may result in arrest and subsequently a revoking of the suspension (Art. 82 CPC). An additional disadvantage of revoking the suspension is that in case of a conviction the days spent in pre-trial detention are legally deducted from the imposed penalty thereby shortening the duration of an intended order.

  2.2.1 Conditional Sentence

  The policy document of the Public Prosecution Service is also applicable to the legal modality of the conditional sentence (Art. 14a CC). This modality has already been constituted in the Dutch CC since 1915.³¹x Stb. 1915, 247. Like the modality of the suspension of remand, this modality was adopted by the legislator in an offender-orientated manner. This offender-orientation is still the case, and even strengthened by new legislation in 2012 to increase the possibilities for and the effectiveness of conditional sentences that are tailor-made to the offender’s individual problems and needs. The conditions imposed on the offender are generally aimed at public safety, preventing recidivism and behavioural change of the offender.³²x Kamerstukken II 2009/10, 32319, nr. 3, at 1. Yet, due to the same legislative development the victim’s interests may currently also be taken into account. These interests are primarily seen from the perspective of victim protection in terms of repeat victimisation. Because of this perspective, the same regulation as with the suspension of remand applies that the perpetrator must declare his compliance with the imposed conditions.³³x Kamerstukken II 2010/11, 32319, nr. 7, at 1-2. If not, the judge will not impose a conditional sentence but perhaps another community sanction with the possibility of a penal protection order instead (i.e. the freedom-restricting measure described in the next subsection), or even a custodial sentence.
  In addition, the effectiveness of the victim protection may be enhanced by many specific legislative regulations. First, alongside the standard probation period of maximum three years the legislator introduced the legal possibility of setting the probation period to a maximum of ten years in case of a serious risk that the offender will again commit a crime aimed at or causing danger to the inviolability of one or more persons’ body (Art. 14b par. 2 CC). This option includes cases of domestic violence, sex offences, stalking, and other harassment of the victim. The probation period can be extended up to the maximum duration of the initial probation period (Art. 6:6:19 CPC). A second regulation in the CC to strengthen the effectiveness of a penal protection order through a conditional sentence is the fact that compliance with this order is supported by mandatory probation supervision (Art. 14c CC). A third regulation states the legal possibility of declaring both this mandatory supervision and the imposed conditions immediately executable, for which the same legal requirements apply as for the extended probation period of ten years (Art. 14e CC); a victim’s request for protection is not a sufficient condition. The judge must rule that the offender poses a serious risk for committing another sex offence or violent act, considering the seriousness of the previously committed offence, the circumstances involved and the risk assessment.³⁴x Kamerstukken II 2010/11, 32319, nr. 7, at 29. In such criminal cases, a penal protection order is usually supported by both an extended probation period and immediate enforcement effect. Additional support for offender compliance is given by a fourth regulation concerning the possibility of electronic monitoring for the full or partial duration of the probation period (Art. 14c CC). According to the legislator this type of monitoring is particularly useful for checking the compliance of PPOs.³⁵x Kamerstukken II 2009/10, 32319, nr. 3, at 20. As with the modality of the suspension of remand, the Public Prosecution Service is explicitly considering electronic monitoring in cases of serious (domestic) violence or sex offences, taking into account the threat that victims (may) experience.³⁶x Stcrt. 2020, 62553, above n. 27. A final regulation that may enhance the effectiveness of a penal protection order within this legal modality is the premise to act swiftly in case of non-compliance.³⁷x Kamerstukken II 2009/10, 32319, nr. 3, at 3 & 11. The Probation Service commissioned to supervise the offender compliance is obliged to immediately report non-compliance to the Public Prosecution Service (Art. 6:3:14 CPC). In addition, an immediate and provisional arrest is possible after non-compliance or even if only a fear of non-compliance occurs (Art. 6:3:15 CPC). As a result, the offender is back in custody pending the judicial decision to execute the sentence (Arts. 6:6:20 and 6:6:21 CPC).
  Regarding the victims’ interests, the conditional sentence is not only seen from a protective but also a restorative perspective. Among the possible legislative conditions belongs the duty to pay a financial compensation to the crime victim for the damage suffered or the duty to repair this damage (Art. 14c par. 2 CC). Both conditions are only meaningful and expedient if the victim agrees with financial or material compensation and wants to cooperate in the implementation thereof, because this condition requires contact with the perpetrator. As such, this restorative perspective is accompanied by the dominant objective of victim protection as acknowledged by both the legislator³⁸x Kamerstukken II 2010/11, 32319, nr. 7, at 19. and the Public Prosecution Service.³⁹x Stcrt. 2020, 62553, above n. 27).

  2.2.2 Freedom-Restricting Measure

  The last modality to be discussed for the legal framework is the freedom-restricting measure that came into effect in 2012 (Art. 38v CC).⁴⁰x Stb. 2011, 546, 615. This penal sanction offers the possibility to the judge to impose a protection order to prevent recidivism or incriminating behaviour towards victims or witnesses.⁴¹x Kamerstukken II 2010/11, 32551, nr. 3, at 1 & 11. Besides perpetrators who persistently cause public disorder, the freedom-restricting measure is specifically aimed at perpetrators who display seriously incriminating behaviour towards the victim or a witness of a previous crime even if this behaviour is not punishable such as continuously ringing the bell at the victim’s house or walking by.⁴²x Kamerstukken II 2010/11, 32551, nr. 3, at 6-7. Although this measure may only be imposed when the perpetrator is actually convicted, the legislator deliberately provided it with a preventive instead of punitive character.⁴³x Kamerstukken I 2011/12, 32551, nr. C, at 9. This preventive character is resembled in the statutory regulation in two ways. First, by the possibility that the perpetrator is not punished for the committed crime in addition to the imposed freedom-restricting measure. Second, the measure may only be imposed if there’s fear of recurrence or future incriminating behaviour towards the victim or other persons. Since this is a risk assessment with all the associated uncertainties, the judge must not only consider the proportionality of the modality but also weigh the different interests of the perpetrator and the victim.⁴⁴x The latter interests may be derived from a victim impact statement (Kamerstukken II 2010/11, 32 551, nr. 6, at 12 & 14).
  The legislator has clearly regarded the freedom-restricting measure as an alternative to the conditional sentence in cases where the committed crime is too light to constitute an effective threat of punishment by the conditional sentence, or where the suspect is not willing to comply with the conditions demanded by the public prosecutor.⁴⁵x Kamerstukken II 2010/11, 32319, nr. 7, at 5-6; Kamerstukken I 2011/12, 32 551, nr. C, at 2. In this regard, the legislator will strive for the most effective victim protection possible. For the same reason, the maximum duration of the freedom-restricting measure was increased from two to five years in 2015.⁴⁶x Stb. 2015, 255. As with the three aforementioned modalities, the judge may declare the freedom-restricting measure to be immediately executable (Art. 38v CC). Compared to the other modalities, the legal threshold for this enforcement mode is quite low because the court only has to determine that there is a fear of recurrence or future incriminating behaviour towards the victim or other persons. On the contrary, as this measure generates a relatively minor restriction of freedom of the perpetrator, no mandatory supervision, and electronic monitoring is possible.⁴⁷x Kamerstukken II 2010/11, 32551, nr. 3, at 9; Court of Rotterdam 22 January 2016, ECLI:NL:RBROT:2016:543. The consequence of violating the protection order imposed by the freedom-restricting measure is alternative detention for a specific duration (Art. 38w CC). As with the conditional sentence, an immediate and provisional arrest of the perpetrator is possible after non-compliance is established or if fear of non-compliance occurs (Art. 6:3:15 CPC).

  2.2.3 Concluding Remarks on the Legal Framework

  Overall it can be stated that at the legislative level all four legal modalities for imposing a penal protection order are primarily aimed at protecting the victim from repeat victimisation. Other aims of protection orders, such as preventing secondary victimisation or protecting danger to the dignity or psychological and sexual integrity of victims, are not explicitly intended. Moreover, the legal characteristics clearly demonstrate the different legal frameworks of the PPOs with various powers for professionals in their enforcement.

• 3 Theoretical Framework

  Following up on the legal framework of Dutch PPOs, this section describes the theoretical framework to analyse the actual practice of the protection orders and the accompanying complex enforcement tasks. As stated in the introduction, professionals in the criminal justice chain appear to be typical frontline work professionals. Therefore, this section successively discusses the insights from the literature on typical conditions in the daily work, the discretion and types of agency, and the coping strategies of those frontline work professionals.

  3.1 Conditions in the Working Field

  When Lipsky describes the working field of street-level bureaucrats, five conditions appear to really determine their work.⁴⁸x Lipsky, above n. 15, at 27-28.

  1. Resources are chronically inadequate to fulfil the tasks frontline workers are responsible for.

  2. Service demands tend to grow infinitely in answer to the available supply.

  3. Expectations about goals are ambiguous, vague or conflicting.

  4. The monitoring of the outcomes or impact of their actions is hard or impossible.

  5. Citizens who are subject to the actions of the frontline workers usually do not volunteer for that role.

  Empirical studies on frontline workers in the criminal justice system underline the ubiquitous presence of those conditions. The condition of limited time and resources is put forward by many studies⁴⁹x M-J. Geenen, E. Kolthoff, R.C. van Halderen en J. de Jong, ‘Street-level bureaucrats in de justitiële jeugdinrichting? Hoe groepsleiders hun discretionaire ruimte benutten’, 58 Tijdschrift voor Criminologie 70 (2016).^,50x J.D. Berrick, S. Peckover, T. Pösö & M. Skivenes, ‘The Formalized Framework for Decision-Making in Child Protection Care Orders: A Cross-Country Analysis’, 25 Journal of European Social Policy 366 (2015). and it is hard to refute that this is at least partly caused by growing service demands on frontline workload.⁵¹x J.M. Wilson, ‘Articulating the Dynamic Police Staffing Challenge: An Examination of Supply and Demand’, 35 Policing 327 (2012).^,52x D. van der Wiele, ‘Zet niet nog meer slachtoffertaken weg bij de politie’, Centrum voor Criminaliteitspreventie en Veiligheid 1 (2019), accessible through https://ccv-secondant.nl/platform/article/zet-niet-nog-meer-slachtoffertaken-weg-bij-de-politie. Moreover, the presence of conflicting legal rights, unique needs and interests of the involved individuals is a major challenge in the frontline workers’ daily practice,⁵³x Berrick et al., above n. 50. referring to condition three about the expectations. The fourth condition primarily results from the extensive range of ways in which frontline workers may have impact on individuals and society.⁵⁴x D. DeHart and C. Shapiro, ‘Integrated Administrative Data & Criminal Justice Research’, 42 American Journal of Criminal Justice 255 (2017). But even the monitoring of more objective outcomes, such as an actual flow of cases through the criminal justice system from first police contact to trial and execution, is not as yet realised in most European countries.⁵⁵x J-M. Jehle, ‘Attrition and Conviction Rates of Sexual Offences in Europe: Definitions and Criminal Justice Responses’, 18 The European Journal on Criminal Policy and Research 145 (2012). The fifth condition is rather obvious for both types of ‘clients’ of the system. On the one hand there are the (potential) victims who do not choose to be in a position in which they depend on the criminal justice system to achieve personal protection and justice. On the other hand, the condition is also true for the suspects and perpetrators who generally prefer to stay away from frontline workers to avoid interruptions of their criminal activities and possible penal sanctions.
  A condition that is not specifically included in Lipsky’s summation though mentioned in other empirical studies is the challenging nature of the cooperation of frontline workers within and between different professions in the criminal justice system.⁵⁶x Geenen et al., above n. 49.^,57x Berrick et al., above n. 50. Those studies describe not only a lack of exchange of information and expertise but also poor trust relationships among frontline workers as a source of stress and a barrier in reaching meaningful individual and collective action to handle or serve clients or citizens. The lack of adequate collective action may result in a highly problematic relation with clients or citizens. As described in a Dutch study on group leaders in youth detention, the lack of collective action among group leaders led to opportunities among the clients to manipulate.⁵⁸x Geenen et al., above n. 49. The conclusion states that improving cooperation asks for an open and consequently more vulnerable interaction, which is, according to the authors, hard to reach given the culture in a majority of professions throughout the criminal justice chain.

  3.1.1 Discretion and Agency

  Along with the conditions described earlier, exemplary for the work of frontline professionals in (among others) the criminal justice system is a high level of discretionary space. Despite the presence of laws, protocols and guidelines, the nature of their work gives them great discretionary space to act.⁵⁹x Fischer et al., above n. 10.^,60x Lipsky, above n. 15, at 4. Therefore, Lipsky describes those agents as typical street-level bureaucrats,⁶¹x Lipsky, above n. 15, at 3. defined as frontline workers acting in fields in which expectations about the effectivity and responsivity of their actions are high but resources are always scarce. Because of their position, they play an important role in the relation between citizen and state and may have strong impact on the lives of individuals.⁶²x Lipsky, above n. 15.
  In answering the question on how this discretion develops in practice, studies did not find much overt opposition among frontline workers against the rules or (enforcement) tasks imposed by legislators and policymakers. Instead, the specified goals are most often endorsed by the frontline workers.⁶³x Geenen et al., above n. 49.^,64x E.Z. Brodkin, ‘Reflections on Street-Level Bureaucracy: Past, Present, and Future’, 72 Public Administration Review 940 (2012). However, the practical implementation of the policies that should lead to the goals is problematic under the given conditions, and desired outcomes are hard to reach.⁶⁵x T. Lindhorst and J.D. Padgett, ‘Disjunctures for Women and Frontline Workers: Implementation of the Family Violence Option’, 79 Social Service Review 405 (2005).^,66x M.K. Meyers, B. Glaser & K. MacDonald, ‘On the Frontlines of Welfare Delivery: Are Workers Implementing Policy Reforms?’, 17 Journal of Policy Analysis and Management 1 (1998).^,67x B.D. Smith and S.E.F. Donovan, ‘Child Welfare Practice in Organizational and Institutional Context’, 77 Social Service Review 541 (2003). Maynard and Musheno state:

  ‘Rules and norms are the working definition of the ‘right way’ to do things, including implementing public policy. But the tension between rules and norms and situations that arise at the frontlines creates conditions in which the ‘right way’ must be negotiated on the ground.⁶⁸x S. Maynard-Moody and M. Musheno, ‘Social Equities and Inequities in Practice: Street-Level Workers as Agents and Pragmatists’, 72 Public Administration Review 16, at S18 (2012).’

  Regarding this individual decision-making by frontline workers, Molander et al.⁶⁹x A. Molander, H. Grimen & E.O Eriksen, ‘Professional Discretion and Accountability in the Welfare State’, 29 Journal of Applied Philosophy 214 (2012). distinguish discretionary space, that is the handling of tools of the professionals, and discretionary reasoning, that is the justifications for decisions professionals may use. An appropriate process of discretionary reasoning asks for adequate information about the contents of the case or situation, the possible reactions of the professional and their consequences (based on the available discretionary space), and the relation of those expected outcomes to the eventual goals.⁷⁰x Berrick et al., above n. 50. In Molander et al.’s study on child protection care, the concept of discretionary space is illustrated by four different elements. First, the availability of relevant information. Second, time to process the information from different parties. Third, the involvement of all relevant parties (e.g., in child protection care: the children and parents) in the process (i.e., by giving and receiving relevant information). And fourth, the adequate monitoring of the outcomes of decision-making.⁷¹x Ibid.
  In line with this discussion on discretion, scholars have promoted a shift from the discretionary decision-making frame towards an ‘agency’ frame of frontline work.⁷²x W. Sewell, ‘A Theory of Structure: Duality, Agency, and Transformation’, 98 American Journal of Sociology 1 (1992); S. Maynard-Moody and M. Musheno, Cops, Teachers, Counselors: Stories from the Front Lines of Public Service (2003); Maynard-Moody and Musheno, above n. 68. This agency frame stresses that workers’ ability to take decisions and act is an essential quality of being human and not a derivative of laws and procedures.⁷³x Sewell, above n. 72. Therefore, to the traditional ‘state-agent’ narrative a ‘citizen-agent’ narrative was added.⁷⁴x Maynard-Moody and Musheno, above n. 68, at S18. In the state-agent narrative, self-interest (i.e., increasing one’s own comfort, safety and work satisfaction) of the frontline workers is addressed to be the strongest force in determining decisions in reaction to the stress following from the impracticality of rules, procedures and laws.⁷⁵x Maynard-Moody and Musheno, above n. 18. Maynard-Moody and Musheno⁷⁶x Ibid. however, find a different story about the way frontline workers use their discretionary space. With their ‘citizen-agent’ narrative based on extensive fieldwork among frontline workers, they describe professionals who react to individuals and circumstances and not to rules and laws. Their reactions are not prompted by self-interest or the aim to make policies, but result from normative choices while trying to make pragmatic and meaningful judgements and actions in the interest of individuals, society or the system.⁷⁷x Ibid. Recent empirical studies support the ‘citizen-agent’ narrative, although the context and nature of the profession, organisational characteristics, and individual characteristics strongly affect the eventual judgements and behaviour of the frontline workers.⁷⁸x M. Sabbe, S. Moyson & N. Schiffino, ‘Citizen-agency Versus State-Agency at the Frontline in Prisons and Probation Services: A Systematic Literature Review’, 55 Social Policy & Administration 206 (2021).^,79x J. Viglione, ‘Street-Level Decision Making Acceptability, Feasibility, and Use of Evidence-Based Practices in Adult Probation’, 44 Criminal Justice and Behaviour 1356 (2017)._,⁸⁰x S. Portillo and D.S. Rudes, ‘Construction of Justice at the Street Level’, 10 The Annual Review of Law and Social Science 321 (2014).
  In the next section, we will discuss the different coping strategies practiced by frontline workers to deal with the challenging incompatibility between the expectations created by the legislator, policymakers, and citizens on the one hand, and the limited resources available to live up to these expectations in a complex reality on the other hand.

  3.1.2 Coping Strategies

  In describing the dynamics in coping behaviour, it is important to distinguish the four levels of coping strategies as presented by Skinner.⁸¹x E.A. Skinner, K. Edge, J. Altman & H. Sherwood, ‘Searching for the Structure of Coping: A Review and Critique of Category Systems for Classifying Ways of Coping’, 129 Psychological Bulletin 216 (2003). At the first and lowest level we find the actual coping occurrences. Here, specific reactions aimed at lowering or dealing with stress can be described. At the second level, Skinner describes ways of coping, for example types of actions with the same background, such as using personal resources. At the third level, families of coping are described. Within such families coping is organised according to its function. The fourth level refers to more general adaptive processes linking coping behaviour to health and general functioning. In their systematic review on coping behaviour of frontline workers Tummers et al.⁸²x Tummers et al., above n. 19. distinguished three different coping families (level 3 by Skinner):

  1. ‘Moving towards clients’; frontline workers not only try to answer the needs of citizens or clients as much as possible despite the huge gap between expectations shaped by formal policies and possibilities, but also try to meet the expectations given the available resources.

  2. ‘Moving away from clients’; frontline workers try to avoid meaningful interaction with citizens or clients to prevent unrealistic expectations.

  3. ‘Moving against clients’; frontline workers confront citizens or clients with the explicit message that they cannot meet their expectations.

  In the complete sample of reviewed studies, the majority of coping occurrences found were examples of ‘moving towards clients’ (43%).⁸³x Ibid. This finding is in line with other studies showing the strong focus of professionals on helping citizens or clients.⁸⁴x J. Johnson Dias and S. Maynard-Moody, ‘For-profit Welfare: Contracts, Conflicts, and the Performance Paradox’, 17 Journal of Public Administration Research and Theory 189 (2006).,⁸⁵x Maynard-Moody and Musheno, above n. 72. The other coping families were represented in respectively 38% (moving away from clients) and 19% (moving against clients) of the coping occurrences. However, in a subsample of studies on police officers the part of coping occurrences that were examples of ‘moving against clients’ was much higher (41%) relative to the complete sample. Moreover, in this sample of police officers the percentage of occurrences that were an example of ‘moving towards clients’ was only slightly lower relative to the complete sample (37%) while the percentage of occurrences represented ‘moving away from clients’ was clearly lower (22%).⁸⁶x Tummers et al., above n. 19, at 1111.
  Examples of ‘moving towards clients’ given in those studies refer to situations in which frontline workers stretch or even break the formal rules for the benefit of the client or citizen,⁸⁷x See also: Geenen et al., above n. 49. situations in which new solutions, structures, or instruments are created to deal with the strain, or situations in which certain citizens are prioritised. ‘Moving towards clients’ also include coping strategies that help frontline workers to deal with their own feelings of unsafety, like the occurrence where groupworkers in youth detention centres use self-initiated reward systems to try to keep control over the clients.⁸⁸x A. Bosma, M. Kunst, A., Dirkzwager & P. Nieuwbeerta, ‘Street-level bureaucracy en verwijzingen naar gedragsinterventies in Nederlandse Penitentiaire Inrichtingen: Discrepanties tussen beleid en praktijk’, Tijdschrift voor Criminologie 4 (2016)._,⁸⁹x See also: B. Crewe, ‘Soft Power in Prison. Implications for Staff-Prisoner Relationships, Liberty and Legitimacy’, European Journal of Criminology 8 (2011). Tummers, et al. also observe that occurrences of rule breaking to ‘move towards clients’ are less prevalent among police officers (22%) and social workers (13%) than for example among health care professionals (64%). A rule bending strategy that was found in the criminal justice literature was that frontline workers combine or align tasks with coworkers in other ways than dictated by formal policies.⁹⁰x Geenen et al., above n. 49.
  Although fewer in number, many coping occurrences appear to fit in the second family (‘moving away from citizen or clients’). A striking example of such a strategy is given by Lindhorst and Pagett,⁹¹x Lindhorst and Padgett, above n. 65. who describe domestic violence-case workers who avoid finding out what service needs exist among their clients because they anticipate these services might be difficult to address. Other strategies found within this family are: sending away citizens who come to an (police) office on a busy moment,⁹²x Tummers et al., above n. 19. the use of stereotypes to make routine decisions (routinising),⁹³x Bosma et al., above n. 88. and the redefinition of tasks and priorities (e.g. starting with the less complex clients) to make them achievable (prioritising).⁹⁴x Ibid.^,95x Tummers et al., above n. 19. The redefinition of tasks is also expressed by the fact that frontline workers lower their personal expectations and ambitions about the successes that can be achieved with their clients. This strategy helps frontline workers to prevent disappointments and lowers the risk of a burnout.⁹⁶x Geenen et al., above n. 49.
  Within the third family of coping strategies, Tummers et al.⁹⁷x Tummers et al., above n. 19. distinguish two types of occurrences, namely rigid rule following and aggression. Rigid rule following is described as a reaction often used ‘as a way to control clients, especially those who are particularly demanding or manipulative’.⁹⁸x Ibid., at 1110. Occurrences regarding this type were found more frequently among police officers and social workers relative to teachers and health care workers.⁹⁹x Tummers et al., above n. 19, at 1112.
  Overall, Tummers’ finding that many coping occurrences illustrate ‘moving towards citizens’ shows that frontline workers are more prone to the citizen-agent than to the state-agent narrative.¹⁰⁰x Maynard-Moody and Musheno, above n. 68, at S18. This however appears to be less apparent for police officers and social workers, which may be associated with the impact of new management movements, focusing on ‘performance management, digitalisation, enhanced technological surveillance, and transparency pressures’.¹⁰¹x Tummers et al., above n. 19.

  3.1.3 Criminal Justice Chain Perspective

  For processes that are a criminal justice chain responsibility (like the enforcement of PPOs), citizens interact with different frontline workers from different formal agencies. The frontline workers make their complex decisions with different perspectives and (partly) different goals and interests in mind. Moreover, the organisations’ ‘internal laws’ that represent the way representatives react in typical situations differ,¹⁰²x Maynard-Moody and Musheno, above n. 72, at S20. showing that ‘discretion is a relative concept’.¹⁰³x Lipsky, above n. 15, at 15. As an example, Lipsky describes:

  ‘Police behaviour is so highly specified by statutes and regulations that policemen are expected to invoke the law selectively. They could not possibly make arrests for all the infractions they observe during the working day.¹⁰⁴x Ibid., at 14.’

  Although the rules may be clear, prioritising appears to be a necessary coping strategy at the level of the police. Such prioritising at the police level may have direct consequences for the courses of action of frontline workers from other agencies (for example public prosecutor or probation officers) who may have other priorities regarding the specified case.

  3.1.4 Street-Level Bureaucracy in the Enforcement of Penal Protection Orders

  Currently, the literature lacks a specific analysis of the coping strategies of frontline workers in the enforcement practice of PPOs. Therefore, it is imperative to exploratively analyse the judgements and behaviour of the frontline workers from the different criminal justice agencies in this specific enforcement process. This explorative analysis is presented in the results section. In line with the theoretical framework, three dimensions will be explored. First, the presence of Lipsky’s conditions for street-level bureaucracy within the enforcement context. Second, the coping strategies frontline workers use in the enforcement process, including the underlying narratives (state-agent or citizen-agent) in the enforcement practice. Third and last, the interaction between the coping strategies of different agents in the criminal justice chain.
  As already stated in the introduction, when combined those explorations may give a more comprehensive description of the enforcement of PPOs in practice thereby also providing relevant input for future improvements.

• 4 Methods

  The empirical data used in this article was collected from a large-scale evaluation study on the practice of PPOs in the Netherlands.¹⁰⁵x Fischer et al., above n. 10. In this study, a variety of both quantitative and qualitative data about and among professionals, victims and offenders was collected. For our analysis, we focus on the interviews with criminal justice professionals. These interviews yielded extensive information on the experiences of the professionals with PPOs in the different legal modalities.¹⁰⁶x Next to the four legal modalities described in section 2, some respondents provided information about penal protection orders during the stage of a conditional release. To mirror the professional perspectives of the criminal justice chain enforcement practices we also include information from the victim and offender interviews. This addition is important because enforcement practices may be experienced differently by citizens (either victims or offenders) then professionals expect.¹⁰⁷x I. Cleven, T. Fischer & S. Struijk, ‘In het belang van het slachtoffer’, 1 Tijdschrift voor Criminologie 11 (2020). Data interviews were conducted between February and September 2017 among a total of 33 criminal justice professionals, nine victims and six suspects or convicts. We interviewed employees of the National Police (n = 8), the Public Prosecution Service (n = 13), Probation Service (n = 6), Victim Support Netherlands (n = 3), and lawyers acting on behalf of victims (n = 3). The themes discussed in these interviews are: protection order enforcement strategies, considerations and types of reactions in case of violations of the protection orders, cooperation with other criminal justice organisations, bottlenecks and dilemmas in the enforcement practice, and professionals’ perceptions of the effectiveness of protection orders. These rich qualitative data, although not specifically collected for this purpose, appear to be very useful in answering our current research question on the coping strategies of frontline workers that result from the perceived discrepancies between the legislative aims of the PPOs and their practical implementation. Naturally though, the analysis is purely explorative since existing data are used and the number of interviews is relatively low.
  The interviews were held face to face or by phone and took between 30 and 90 minutes. All interviews were taped and written reports were made afterwards. In the results section we summarise our findings and illustrate them with quotes. All respondents gave written informed consent for the use and storage of the information they provided. Data were pseudonymised and stored on a specifically protected environment only accessible to the main researchers.

• 5 Results

  In this explorative analysis we examine if and to what extent the five important conditions in the work of frontline workers are apparent in the enforcement process of PPOs. Our findings show that especially the first four conditions are clearly present. We will present the findings in a successive manner, thereby integrating our results on the coping strategies and agency narratives that the professionals appear to apply in order to deal with these conditions.

  5.1 Inadequate Resources to Fulfil the Tasks of Frontline Workers and an Infinite Growth of Service Demands

  The first two conditions of Lipsky are jointly discussed, because in our view the question whether the resources are adequate and sufficient enough is in itself related to the broad and comprehensive range of tasks of the professionals working within the criminal justice chain and the services required of them. This is definitely true for the police, as is expressed in the following quote by a liaison police officer dealing with domestic violence.

  ‘When you investigate you try to get it on top of the pile, but I also foresee that if a lot is urgent, then at some point you have ten urgent cases and number ten is no longer as urgent as number one, right? And capacity is already a real problem within the entire chain. I also see that at Youth Care, and Veilig Thuis.’

  In line with previous research,¹⁰⁸x Van der Aa et al., above n. 12. our interviews show the limited capacity of the police for proactive enforcement. As a result, actual reports of violations by the police are rare. Moreover, enforcement interventions as (random) surveillance and home visits by the community police officer are generally only used in high-risk cases. Electronic monitoring, often mentioned as an accurate measure to proactively control location orders, is only used in a limited number of imposed protection orders (5% of protection orders included in the study of Fischer, Cleven & Struijk, 2019). The sparse use of electronic monitoring may be the result of limited supervision capacity. Yet, according to our respondents (i.e. probation officers) it also results from judgements on proportionality as well as the eligibility of the suspect or convict. The latter represents examples of the third condition of Lipsky referring to vague and conflicting goals.
  The respondents describe a great diversity in both the nature and seriousness of protection order violations, which clearly shows the growth of service demands as a result of the protection orders. This diversity can roughly be divided into three categories. First, incidents that constitute an independent criminal offence. Second, incidents that involve a violation of the protection order but do not constitute an independent criminal offence. Third, incidents that are not prohibited but infringe the victim’s sense of safety and well-being, thereby resulting in the victim experiencing the incident as a violation and reporting it to the judicial authorities. The latter category of violations are identified by both interviewed victims and professionals (police officers, case coordinators, public prosecutors, and a supervisor). However, many professionals view such violations as inevitable and less problematic, as expressed by a public prosecutor.

  ‘You have a certain type of behaviour from suspects that is very annoying and troublesome, but you cannot reasonably say ‘yes, this is subject to a violation of the condition’. For example, an accidental encounter in the supermarket […] And then a victim naturally feels misunderstood because she can no longer go somewhere. Yet, it does stop at a certain point, you can't impose a ban on someone in such a way that they are no longer allowed to go to many places, there are just certain limits. […] and then I think it mainly relates to that subjective safety, because yes, if it is indeed all those supermarket visits, they are very annoying for victims, but there will probably be no abuse.’

  In cases like this, the interviews reveal that victims expect an enforcement response from the judiciary. According to some professionals and victims, the suspect or convict is deliberately pushing the boundaries of the protection order. From a legal point of view there is no violation of the order. This legal view can be very unclear, as was illustrated by an interviewed victim.

  ‘On the same day that he was released from jail in the morning he was behind my yard, not really in my yard, but across from it. Knowing that he was banned from the area […] He has a restraining order, now of course he is not actually in my street, but is he allowed to be there? […] The local police officer that immediately came by my house […] and I say, well, I just find it very unclear. And she acknowledged that indeed it is very unclear.’

  A final issue limiting the efficient usage of resources in the enforcement of protection orders, is the fact that, given the complexity of the cases, the quality of risk assessment in individual cases is limited. Therefore, it is hard to prioritise the right cases for proactive enforcement interventions (e.g. surveillance and home visits). Police officers, victims and victim support officers describe the difficulty of prioritising as an important problem and stress the need for more knowledge, specialised officers and better communication between different chain actors to improve the quality of risk assessment. At some places in the police organisation, specialised teams or specialised officers are actively involved in protection order cases, which is a clear example of the coping family ‘moving towards clients’. However, the proportion of cases in which a specialised officer is involved appears to be quite low (i.e. 10% to 15% of the victims in the survey by Fischer et al. reported the involvement of specialised officers).¹⁰⁹x Fischer et al., above n. 10, at 100.
  As a result of the conditions, the police officers we interviewed described the enforcement of protection orders as mainly reactive. As illustrated by an experienced police officer from the Unit Domestic Violence:

  ‘The police only acts when the Probation Services or other supervisors alert the Public Prosecution Service or judiciary that the suspect or perpetrator does not comply with the conditions. The police are then requested to arrest the suspect or perpetrator.’

  The interviewed police officers, public prosecutors, and probation officers all identified the lack of proactive enforcement as a bottleneck for effective victim protection. Yet, some respondents acknowledged that even with increased forms of proactive enforcement victim protection is not guaranteed, because the police cannot keep watch 24/7.
  Another identified bottleneck is the transfer of knowledge about existing PPOs in individual cases and the accessibility of this information within the police organisation. Obviously, being informed is an important resource for adequate enforcement. Yet, it is also important to have easy access to the information because for police officers the enforcement of the orders is always part of a broader range of tasks. Especially the practice of providing the community police officers with necessary information appeared to be problematic. Several respondents addressed the fact that there are no specified protocols to inform those officers, resulting in a dependency on individual initiatives of the public prosecutor whether the community officer is informed or not. In our interviews various examples were given by police officers of their creative methods to alert their colleagues on protection orders in individual cases, like in the following two quotes from police officers.

  ‘I’m going to pass that [imposition] to the base team. I can do that in two, three ways. First, I can have it put in a briefing item for a set number of days. Second, I can email it to colleagues who are on the base team. Third, nowadays we work with Agora, that is actually sort of the facebook of the police […] well that’s where we put an order on and it stays there, contrary to mentioning it in a briefing where it is gone after a week. So I often upload the verdict in that system.’

  In case of additional rapid response appointments with victims, I always separately ‘copy and paste’ the specific protection orders into the rapid response alerts in the system, to provide the rapid response officer with complete information on how to react.
  Both examples show that because standard working methods may not always be appropriate, professionals are searching for strategies to increase the probability of successful enforcement. This clearly relates to the coping family ‘moving towards clients’. Despite these examples of own initiatives by police officers, many respondents – victims as well as professionals – still feel that police officers are generally insufficiently informed about current protection orders.
  Obviously, the lack of information may hamper the police in properly and timely identifying violations of a current protection order. The same is true for police officers not always having sufficient knowledge of the nature of a current order and also for specific cases in which PPOs may coexist with civil and/or administrative protection orders. Some victims we spoke to gave examples of situations where police officers did not know whether it concerned a penal or a civil protection order and even refused to act on a certain violation of the order on the assumption that it was a civil order. Other victims reported that police officers refused to take victims’ declarations of stalking incidents. According to the victims, policemen refused because of a lack of knowledge about the problematic dynamics of stalking. The aforementioned examples relate to coping strategies representing ‘moving away from’ clients. Although we do not know on what scale this occurs, it may be part of an explanation for the observation we described in the introduction that a huge difference exists between the numbers of violations reported by victims and those registered in the files of the Public Prosecution Service.

  5.2 Ambiguous, Vague, or Conflicting Expectations Regarding Aims

  The third condition concerning the existence of ambiguous, vague or conflicting expectations regarding aims is also clearly present in the enforcement process of PPOs. Victims and professionals seem to interpret the concept of effectiveness of protection orders differently in terms of safety perceptions. Professionals tend to focus on objective safety (less repeated victimisation), whereas victims view effectiveness in a broader, more subjective manner (including their safety perception). This finding is consistent with the discussion in the legal framework. When designing the four legal modalities for imposing PPOs, the legislator aimed primarily at preventing recidivism in general and, more specifically, repeat victimisation. The legislator did not specifically focus on victim safety perceptions. Such differences concerning the intended aims of the protection orders appear to influence the effectiveness of the enforcement and the perception thereof by both victims and professionals. Victims tend to expect that, if the protection order is violated, adequate action will be taken by the professionals through swift arrest and imprisonment of the suspect or convict. As described by our respondents, that thought alone gives the victims more freedom of movement, a greater peace of mind, and a feeling of safety. However, for some of the victims interviewed, these positive feelings waned as no visible or adequate enforcement response had come after the order was violated. This perceived lack of enforcement is identified by almost half of the victims that were interviewed as an important hindering factor for the protection orders to contribute to feelings of safety.
  Several explanations for this perceived lack of enforcement have emerged from the research findings, such as choices and prioritisation, insufficient knowledge of protection orders, and the legal necessity to take proportionality into account when choosing the response. Again, this has a lot to do with different expectations and interpretations. In the following quote, both a victim and her current partner refer to the public prosecutor’s reaction on violations in a threat and domestic violence case:

  ‘Victim: The problem was that at one point the detectives said ‘there is no longer a threat’. Current partner: While at the same day he said, ‘I’m going to kill her.’ Victim: But it will probably be the interpretation of this detective who says, ‘Yes he says a lot, but he has done so for two years already and did nothing.’ And that is the difficult part of this story, the sense of safety that we have and what is actually happening. There is still so much space in between. And that is very difficult for us.’

  By enforcement, the interviewed victims and some police officers generally mean that the suspect or perpetrator is detained, while various public prosecutors also see a warning as a potentially effective enforcement response. The possibility to suffice with a warning was also acknowledged by the legislator, as can be deduced from the following remark in parliamentary history in regard to the conditional sentence: ‘Sometimes a warning will suffice, but a substantial and/or systematic violation of the conditions must lead to the execution of the initial conditional sentence.’¹¹⁰x Kamerstukken II 2009/10, 32319, nr. 3, at 11.

  Proportional and customised responses are also important because protection orders may function as a framework creating a relatively safe situation in which treatment and structural resources for the perpetrator can be arranged to improve his resocialisation. This resocialisation may, when successful, also be beneficial for victims in the long run. Yet, to achieve that goal, proportionality and flexibility in the enforcement may be important, though that often conflicts with short-term needs of victims. In any case, decisions about the best response ask for the use of the right information but interpretations between professionals at different positions in the criminal justice chain can differ strongly. Interviewed community police officers described several examples where the trial judge or the investigating judge has a different interpretation of the likelihood that a suspect or convict will change his behaviour after having violated the protection order for the second or third time. In the following quote by a community police officer such a situation is described:

  ‘I have a case where a young man [with a restraining order] harasses his ex-girlfriend a lot, mistreats her, is arrested and is imprisoned for fourteen days because it concerned quite serious abuse. Then he was being brought before the investigating judge, where suspension of remand is allowed because, according to his lawyer, he has just started a new job and wants to start a study. For us [the police] that was strange, because we knew that he was not going to work or start a study, but that was what the lawyer said and he was allowed to go home of course with suspensive conditions. He goes straight from the jail to that girl’s house, smashes the windows, is arrested on the spot by us and is again allowed to go home after two days because according to the investigating judge it is really important for this boy to begin his study. […] Such a boy, whenever being imposed with a contact ban in future, is convinced that his lawyer will get him released again, so why shouldn’t he do what he wants to do?’

  This quote illustrates that the community police officer tries to ‘move towards the victim’ by prioritising the case and reactions quickly to violations in order to protect the victim, while the investigating judge ‘moves towards the suspect’ in the belief that this may support his resocialisation. According to the police officer, the decision of the judge not only increases the risk for this specific victim but also for future victims.

  Besides customised criminal justice responses on behalf of the suspect or convict, the description of the protection order is sometimes customised too. For example, when a location order is adapted it makes it possible for the suspect or convict to reach the workplace or to visit family. This method of customising the application of protection orders is another example of the coping strategy ‘moving towards’ citizens, with the suspect or convict in the role of citizen. At the same time, it may mean a denial of victims needs or interests for example if the customising decreases the area in which a victim may feel relatively safe. This situation was frequently mentioned by several interviewed victims. Therefore, the coping strategy may be defined as ‘moving away from’ citizens with respect to the victims in the role of citizen.
  The police officers we spoke to do not seem to disagree much from public prosecutors about the extent to which a warning is an appropriate enforcement response, but they do disagree about the circumstances under which this response is appropriate. A few police officers – as well as victims – indicate that they regularly find the response from the Public Prosecution Service too mild. This can be detrimental to the effectiveness of the enforcement. For example, police officers may view that a lack of decisiveness from the Public Prosecution Service affects the credibility of a police warning. An interviewed community police officer indicated that he would rather not issue a warning. The following quote of a public prosecutor illustrates these different narratives of the professionals:

  ‘Sometimes, the opinions [differ]. Look, of course we all have the same goal. But sometimes, look, the public prosecutor can of course assess how a judge would view this. And you are sitting there with […] of course the police are on top of it, they have direct conversations. Experience shows that they often want to act quickly and then it is up to the public prosecutor to find certain middle ground. To communicate to the victim ‘we are there for you’, but also to make a realistic assessment of whether it will actually be enforced by the investigating judge or the trial judge, or not.’

  Some of the public prosecutors we interviewed try to solve this gap between the police and public prosecutors’ coping strategies by imposing the legal modality of the criminal behaviour order (Art. 509hh CPC), thereby increasing the judicial possibilities for prompt enforcement reactions. One of the public prosecutors explained that as follows:

  ‘There will come a point where you say, ‘Yes, he has been incarcerated for so long for this offence, it’s just not possible anymore.’ Anyway, then you can, I mean, then you have to be creative, that you think, well, if all that doesn’t work anymore, then I’m just going to impose a behaviour order, and if he will violate it, then you have a new offence, and I will act on it. So as a prosecutor, you can do it one way or the other.’

  Naturally, the decision on the right and appropriate enforcement response in a specific case is highly dependent on the provability of the reported violation of a protection order. The provability is an important factor given the possibility of a false report by the victim, as acknowledged by a police officer interviewed. At the same time all respondents view provability as an impeding factor for the effectiveness of protection orders. Some respondents stated that this is less problematic when the order is enforced by electronic monitoring. Yet, another important factor is proportionality. According to an interviewed public prosecutor, due to this factor a warning may still suffice in case of minor violations where electronic monitoring was applied.
  Besides the aforementioned discrepancy between the perception of victims and professionals about the effectiveness of a warning as response to a violation, it is important to note that from a criminal proceedings point of view professionals perceive every report by victims as useful. Even if it is not clear that the report indeed concerns a violation of the order, or a new criminal offence, a note will be made in the criminal file of the perpetrator. Both an interviewed public prosecutor and community police officer emphasise this extension of the criminal file as a useful mechanism when the suspect appears to be immune to the threat of punishment. Such a note in the criminal file may for instance increase the penalty imposed on the perpetrator after a subsequent violation according to an interviewed public prosecutor. The note may also lead to a sentence that is better customised to the situation, as the information on violations increased the public prosecutors’ insights about the dynamics of the situation.
  Accordingly, probation officers describe that they use information about violations in the supervision of their clients, as it gives input for periodical talks about their behaviour during the probationary trajectory and the applicable goals. However, as none of the interviewed victims acknowledged them, it appears that these mechanisms in the victims’ perception do not contribute to ensure safety or at least a feeling of safety, perhaps because victims are not aware of this function.
  The examples of a lack of enforcement and a lack of a meaningful response to a reported violation provided by interviewed victims mainly concerned cases of stalking. This is unsurprising as these cases often involve non-criminal acts such as so-called accidental encounters in the neighbourhood. Respondents, such as a public prosecutor, acknowledge the difficulty in these cases to ensure (perceived) victim safety. This is related to the finding that victims and professionals have different and conflicting expectations on the aims of protection orders, as is also described below.
  Several respondents – victims and professionals – acknowledged that the aim of the enforcement is predominantly the absence of violence though to victims this has a broader meaning of creating peace and the absence of any confrontation with the suspect or perpetrator. Some victims and professionals – including a lawyer acting on behalf of victims – expressed their scepticism about the possible deterring effect of protection orders, indicating that if the suspect or convict truly wants to harm the victim, he will not be stopped by an order. Therefore, as long as victims still experience confrontations, they will feel restless and unsafe.

  5.3 Hard or Even Impossible Monitoring of Outcomes and Impact

  The results on ambiguous and conflicting expectations as a hindering factor for the effectiveness of protection orders are clearly connected to the fourth condition of Lipsky. First of all, this connection results from the fact that the monitoring of outcomes and impact is problematic because of the ambiguous aims connected to the protection orders. The difference between objective and subjective safety is included, but also the other aims, namely the protection order as framework to facilitate treatment in a community setting, and the instrumental aim concerning the criminal proceedings of supervision points of view (extension of the criminal file and probationary supervision).
  A second reason why the monitoring of outcomes and impact is problematic is the difficult interpretation of the figures on violations of protection orders. Low violation frequencies may be a result of the absence of violations but may also be related to the absence of victim reports of violations, or the active monitoring and/or administration of violations. Consequently, positive developments such as an increasing effort in proactive monitoring or an increase in the willingness of victims to report, all increase the frequencies of violations in the files of the criminal justice system. Note that the victim’s willingness to report strongly deteriorates from limited expectations of or disappointment in enforcement.¹¹¹x Fischer et al., above n. 10, at 95. The other two factors that emerged from the interviews are the extent to which the victim feels dependent on the suspect or perpetrator, and the knowledge about the imposition of the order and the method of reporting a violation.
  A final reason why it is hard to monitor outcomes and impact is the fact that PPOs are rarely used as isolated interventions. Often, the protection order is just one measure among a collection of possible measures (judicial, social and clinical) to regulate a complex victim-offender relationship. Therefore, quasi-experimental designs with a sound treatment and control group are hard to realise.
  Many frontline workers interviewed reflected on this lack of monitoring as very problematic. However, none mentioned clear coping strategies to deal with this situation. At some places specialised officers were put in place to coordinate the enforcement of protection orders (for example police officers specialised in domestic violence). Those specialised officers clearly had more ideas and vision about effective strategies in the enforcement of protection orders than less specialised officers. Victims reported generally positive experiences with specialised teams, indicating that they are relatively successful in increasing knowledge about effective enforcement. However, this increase of knowledge appears to be the result of accumulation of experience instead of systematic monitoring of the outcomes. Therefore, regarding the organisation level, we can conclude that the attribution of specialised officers to the coordination of protection order enforcement is a successful strategy (i.e. ‘moving towards’ clients).

  5.4 Citizens Who Are Subjects of the Frontline Workers Actions Usually Do Not Volunteer in that Role

  The fifth condition is the least apparent condition regarding the enforcement process of protection orders. The involuntariness of the suspect or convict is given in the penal setting in which the order is imposed. The order is part of a legal modality chosen by the public prosecutor or judge as the appropriate response to certain criminal behaviour of the offender. For some legal modalities the offender must comply with the conditions. Yet, this compliance is still not true voluntariness as the decision may be mainly driven by the possible repercussions of non-compliance. For instance, if the offender does not declare his willingness to meet the conditions for the suspension of remand, the detention on remand will continue.

• 6 Conclusion and Discussion

  The penal protection order is one of the instruments deployed to realise a more active protection of crime victims. In many countries, including the Netherlands, this aim has been given a strong and increasing priority in the past decades. Yet, knowledge on the actual practice of the protection orders and the successes, dilemmas and challenges involved is scarce. Our explorative study tried to expand this knowledge by first analysing the legal framework of the Dutch PPOs and the associated specific aims of the legislator. Next, we examine the actual enforcement practice of the PPOs using data from interviews with a great variety of actors involved (professionals, victims and perpetrators). Finally, the practice of penal protection order enforcement is analysed using the theoretical framework of Lipsky’s street-level bureaucracy and the succeeding work of among others Maynard & Musheno and Tummers on coping strategies and agency narratives of frontline workers.
  The analysis of the legal framework shows that in both the legislative and policy choices the main focus is on effectively decreasing the risk of another violent act. Illustrative is the regulation that probation supervision is not only mandatory regarding conditional sentences with a protection order imposed, but the judge may also declare both this supervision and the imposed order immediately executable. The latter decision is restricted to the situation that the offender possesses a serious risk for committing another sex offence or violent act. Simultaneously, priorities to prevent psychological damage for victims caused by for example, stalking, are inevitably low in cases where judges estimate the risk of actual violence as low or mediate. This conclusion is notable, as the international legal framework concerning protection orders does not only focus on repeat victimisation but also on the protection of victims in a broader sense from any danger of bereaved dignity or psychological and sexual integrity.
  Another outcome from analysing the legal framework is that the discussed legal modalities to some extent differ in creating possibilities for the protection orders to actually serve the victims’ interests. For instance, whereas all legal modalities and subsequent protection orders may be declared immediately executable – thereby providing (feelings of) safety for the victim involved – the freedom-restricting measure has the lowest legal threshold for this enforcement mode. The court only has to determine that there is fear of recurrence or future incriminating behaviour towards the victim or other persons. As perpetrators of stalking and domestic violence often show recurring behaviour, the freedom-restricting measure may serve the victim’s interests better than a legal modality with a higher threshold for making the protection order immediately executable. Besides, due to considerations of proportionality the legal framework of the freedom-restricting measure constitutes no possibility for mandatory probation supervision and electronic monitoring, thereby minimising the possibilities for proactive enforcement of the order. Future qualitative research with interviews and case law analysis may be relevant to provide insights into the manner in which judges are aware of these differences and if so, how they weigh these diverging interests of victims and perpetrators when choosing the specific legal modality to impose a protection order.

  Our explorative findings on the enforcement practice of PPOs clearly show the presence of the five conditions Lipsky described as being important in the work of street-level bureaucrats. Particularly the first four conditions are true: inadequate resources to fulfil the tasks of frontline workers, an infinite growth of service demands, the presence of ambiguous, vague or conflicting expectations of aims and lastly a difficult monitoring of the outcomes or impact of the actions of frontline workers. These conditions are related to the strong discrepancy between the aims of the Dutch legislator – namely a significant decrease in the risk of new violent acts from the offender towards the victim and, subsequently, the prevention of repeat victimisation – and the actual enforcement practices. In almost all situations of imposed protection orders the process of monitoring violations is reactive, because only a small proportion of the offenders with a protection order is supervised via electronic monitoring. Naturally, this makes it hard to actually prevent (new) violent incidents. As knowledge on risk indicators for violent escalation is still limited, it is difficult for frontline workers to decide on how many and what type of resources should be invested in which cases inevitably resulting in a ‘moving away from’ client’s strategy.
  However, in case reactive enforcement is the default, we also found several examples of coping strategies by organisations or individual frontline workers that represent strong examples of ‘moving towards clients’. These strategies varied from creative improvements concerning the communication and availability of information about protection orders, the use of specialised officers and teams (both ‘moving towards’ the victim), to the practice of customising the order or enforcement reactions to facilitate treatment or resocialising activities of the suspect or convict (‘moving towards’ the suspect or convict). Despite the problematic conditions in which the frontline workers have to enforce the protection orders, these strategies show that they try to do their best for both victims and suspects or convicts. These findings match with the citizen-agent narrative in which frontline workers try to use their discretionary space to make pragmatic and meaningful judgements and actions in the interest of individuals instead of in self-interest.
  Another outcome is the existence of conflicting coping strategies, decisions and behaviour of professionals at different positions in the criminal justice chain. Professionals often appear to deviate in their assessments on the proportionality, legal possibilities and the expected short-term and long-term effects of enforcement reactions (for example the execution of the initial conditional sentence). An outcome in line with the conclusion that all frontline workers show that they are strongly hindered by limited possibilities for discretionary reasoning resulting from inadequate information about the case and expected consequences of their enforcement reactions. This lack of information is at least partly resulting from the lack of monitoring and corresponding lack of feedback loops in the criminal justice chain. Therefore, our study leads to the recommendation that all criminal justice partners should invest in the accumulation and communication of knowledge on the occurrence of violations of PPOs and on the consequences of enforcement reactions on repeat victimisation and victim’s safety perceptions (both in the short term and in the long term). Monitoring the consequences of strategies in the enforcement of protection orders and relating them to the characteristics of the cases offers a solution. In the end, such monitoring may improve the quality of risk assessment and leads to more customised enforcement strategies and reactions to violations. Moreover, it may help to decrease the observed differences in considerations about proportionality of enforcement responses across different criminal justice agencies.

Noten

• 1 Council Resolution of 28 June 2011, OJ 2011 C 187/01.

• 2 Council Document 17024/09, OJ 2019.

• 3 EP and Council Directive of 13 December 2011, OJ 2011 L 338/2.

• 4 N. Elbers, S. Meijer, I. Becx, A.J.J.G. Schijns & A. Akkermans, ‘The Role of Victims’ Lawyers in Criminal Proceedings in the Netherlands’, OnlineFirst European Journal of Criminology 1 (2020).

• 5 A.K. Bosma, M.S. Groenhuijsen & M. de Vries, ‘Victims’ Participation Rights in the Post-Sentencing Phase: The Netherlands in Comparative Perspective’, 12 New Journal of European Criminal Law 128 (2021).

• 6 S. van der Aa, ‘Protection Orders in the European Member States: Where Do We Stand and Where Do We Go from Here?’, 18 European Journal of Criminal Policy and Research 183 (2012).

• 7 A. ten Boom and K.F. Kuijpers, ‘Victims’ Needs as Basic Human Needs’, 18 International Review of Victimology 155 (2012).

• 8 H.C.J. van der Veen and S. Bogaerts, Huiselijk geweld in Nederland. Overkoepelend syntheserapport van de vangst-hervangst-, slachtoffer- en daderonderzoek 2007-2010 (2010).

• 9 Ministerie van Volksgezondheid, Welzijn en Sport, Factsheet Huiselijk Geweld (2013).

• 10 T. Fischer, I. Cleven & S. Struijk, Handhaving en veiligheid bij strafrechtelijke contact-, locatie- en gebiedsverboden ter bescherming van slachtoffers (2019).

• 11 Fischer et al., above n. 10, at Table 7.1.

• 12 See also: S. van der Aa, K. Lens, F. Klerx, A. Bosma & M. van den Bosch, Aard, omvang en handhaving van beschermingsbevelen in Nederland. Deel 1: Wettelijk kader en handhaving (2012).

• 13 Fischer et al., above n. 10.

• 14 Van der Aa et al., above n. 12.

• 15 M. Lipsky, Street-Level Bureaucracy: Dilemmas of the Individual in Public Services (2010).

• 16 Ibid., at 82.

• 17 S. Halliday, N. Burns, N. Hutton, F. McNeill & C. Tate, ‘Street-Level Bureaucracy, Interprofessional Relations, and Coping Mechanisms: A Study of Criminal Justice Social Workers in the Sentencing Process’, 31 Law & Policy 405, at 406 (2009).

• 18 S. Maynard-Moody and M. Musheno, ‘State Agent or Citizen Agent: Two Narratives of Discretion’, 10 Journal of Public Administration Research and Theory 329 (2000).

• 19 L. Tummers, V. Bekkers, E. Vink & M. Musheno, ‘Coping During Public Service Delivery: A Conceptualization and Systematic Review of the Literature’, 25 Journal of Public Administration Research and Theory 1099 (2015).

• 20 Halliday et al., above n. 17, at 406.

• 21 N. Padfield, ‘Legal Constraints on the Indeterminate Control of “Dangerous” Sex Offenders in the Community: The English Perspective’, 9 Erasmus Law Review 55 (2016); N. Padfield, R. Morgan & M. Maguire, ‘Out of Court, Out of Sight?’, in M. Maguire, R. Morgan & R. Reiner (red.), The Oxford Handbook of Criminology (2012) 955.

• 22 S. Struijk, ‘Vrijheidsbeperking na detentie: in hoeverre zet de rechter de achterdeur open?’, 6 Sancties 353 (2018); Bosma et al., above n. 5.

• 23 Stb. 2016, 310.

• 24 Kamerstukken II 2007/08, 31467, nr. 3, at 18.

• 25 Kamerstukken II 2007/08, 31467, nr. 3, at 1-2 & 29.

• 26 Kamerstukken II 1913/14, 286, at 3 & 84-85.

• 27 Aanwijzing voorwaardelijke straffen en schorsing van voorlopige hechtenis onder voorwaarden (Stcrt. 2020, 62553).

• 28 In some countries known as a ‘no contact’ provision.

• 29 Stcrt. 2020, 62553, above n. 27.

• 30 Ibid.

• 31 Stb. 1915, 247.

• 32 Kamerstukken II 2009/10, 32319, nr. 3, at 1.

• 33 Kamerstukken II 2010/11, 32319, nr. 7, at 1-2.

• 34 Kamerstukken II 2010/11, 32319, nr. 7, at 29.

• 35 Kamerstukken II 2009/10, 32319, nr. 3, at 20.

• 36 Stcrt. 2020, 62553, above n. 27.

• 37 Kamerstukken II 2009/10, 32319, nr. 3, at 3 & 11.

• 38 Kamerstukken II 2010/11, 32319, nr. 7, at 19.

• 39 Stcrt. 2020, 62553, above n. 27).

• 40 Stb. 2011, 546, 615.

• 41 Kamerstukken II 2010/11, 32551, nr. 3, at 1 & 11.

• 42 Kamerstukken II 2010/11, 32551, nr. 3, at 6-7.

• 43 Kamerstukken I 2011/12, 32551, nr. C, at 9.

• 44 The latter interests may be derived from a victim impact statement (Kamerstukken II 2010/11, 32 551, nr. 6, at 12 & 14).

• 45 Kamerstukken II 2010/11, 32319, nr. 7, at 5-6; Kamerstukken I 2011/12, 32 551, nr. C, at 2.

• 46 Stb. 2015, 255.

• 47 Kamerstukken II 2010/11, 32551, nr. 3, at 9; Court of Rotterdam 22 January 2016, ECLI:NL:RBROT:2016:543.

• 48 Lipsky, above n. 15, at 27-28.

• 49 M-J. Geenen, E. Kolthoff, R.C. van Halderen en J. de Jong, ‘Street-level bureaucrats in de justitiële jeugdinrichting? Hoe groepsleiders hun discretionaire ruimte benutten’, 58 Tijdschrift voor Criminologie 70 (2016).

• 50 J.D. Berrick, S. Peckover, T. Pösö & M. Skivenes, ‘The Formalized Framework for Decision-Making in Child Protection Care Orders: A Cross-Country Analysis’, 25 Journal of European Social Policy 366 (2015).

• 51 J.M. Wilson, ‘Articulating the Dynamic Police Staffing Challenge: An Examination of Supply and Demand’, 35 Policing 327 (2012).

• 52 D. van der Wiele, ‘Zet niet nog meer slachtoffertaken weg bij de politie’, Centrum voor Criminaliteitspreventie en Veiligheid 1 (2019), accessible through https://ccv-secondant.nl/platform/article/zet-niet-nog-meer-slachtoffertaken-weg-bij-de-politie.

• 53 Berrick et al., above n. 50.

• 54 D. DeHart and C. Shapiro, ‘Integrated Administrative Data & Criminal Justice Research’, 42 American Journal of Criminal Justice 255 (2017).

• 55 J-M. Jehle, ‘Attrition and Conviction Rates of Sexual Offences in Europe: Definitions and Criminal Justice Responses’, 18 The European Journal on Criminal Policy and Research 145 (2012).

• 56 Geenen et al., above n. 49.

• 57 Berrick et al., above n. 50.

• 58 Geenen et al., above n. 49.

• 59 Fischer et al., above n. 10.

• 60 Lipsky, above n. 15, at 4.

• 61 Lipsky, above n. 15, at 3.

• 62 Lipsky, above n. 15.

• 63 Geenen et al., above n. 49.

• 64 E.Z. Brodkin, ‘Reflections on Street-Level Bureaucracy: Past, Present, and Future’, 72 Public Administration Review 940 (2012).

• 65 T. Lindhorst and J.D. Padgett, ‘Disjunctures for Women and Frontline Workers: Implementation of the Family Violence Option’, 79 Social Service Review 405 (2005).

• 66 M.K. Meyers, B. Glaser & K. MacDonald, ‘On the Frontlines of Welfare Delivery: Are Workers Implementing Policy Reforms?’, 17 Journal of Policy Analysis and Management 1 (1998).

• 67 B.D. Smith and S.E.F. Donovan, ‘Child Welfare Practice in Organizational and Institutional Context’, 77 Social Service Review 541 (2003).

• 68 S. Maynard-Moody and M. Musheno, ‘Social Equities and Inequities in Practice: Street-Level Workers as Agents and Pragmatists’, 72 Public Administration Review 16, at S18 (2012).

• 69 A. Molander, H. Grimen & E.O Eriksen, ‘Professional Discretion and Accountability in the Welfare State’, 29 Journal of Applied Philosophy 214 (2012).

• 70 Berrick et al., above n. 50.

• 71 Ibid.

• 72 W. Sewell, ‘A Theory of Structure: Duality, Agency, and Transformation’, 98 American Journal of Sociology 1 (1992); S. Maynard-Moody and M. Musheno, Cops, Teachers, Counselors: Stories from the Front Lines of Public Service (2003); Maynard-Moody and Musheno, above n. 68.

• 73 Sewell, above n. 72.

• 74 Maynard-Moody and Musheno, above n. 68, at S18.

• 75 Maynard-Moody and Musheno, above n. 18.

• 76 Ibid.

• 77 Ibid.

• 78 M. Sabbe, S. Moyson & N. Schiffino, ‘Citizen-agency Versus State-Agency at the Frontline in Prisons and Probation Services: A Systematic Literature Review’, 55 Social Policy & Administration 206 (2021).

• 79 J. Viglione, ‘Street-Level Decision Making Acceptability, Feasibility, and Use of Evidence-Based Practices in Adult Probation’, 44 Criminal Justice and Behaviour 1356 (2017).

• 80 S. Portillo and D.S. Rudes, ‘Construction of Justice at the Street Level’, 10 The Annual Review of Law and Social Science 321 (2014).

• 81 E.A. Skinner, K. Edge, J. Altman & H. Sherwood, ‘Searching for the Structure of Coping: A Review and Critique of Category Systems for Classifying Ways of Coping’, 129 Psychological Bulletin 216 (2003).

• 82 Tummers et al., above n. 19.

• 83 Ibid.

• 84 J. Johnson Dias and S. Maynard-Moody, ‘For-profit Welfare: Contracts, Conflicts, and the Performance Paradox’, 17 Journal of Public Administration Research and Theory 189 (2006).

• 85 Maynard-Moody and Musheno, above n. 72.

• 86 Tummers et al., above n. 19, at 1111.

• 87 See also: Geenen et al., above n. 49.

• 88 A. Bosma, M. Kunst, A., Dirkzwager & P. Nieuwbeerta, ‘Street-level bureaucracy en verwijzingen naar gedragsinterventies in Nederlandse Penitentiaire Inrichtingen: Discrepanties tussen beleid en praktijk’, Tijdschrift voor Criminologie 4 (2016).

• 89 See also: B. Crewe, ‘Soft Power in Prison. Implications for Staff-Prisoner Relationships, Liberty and Legitimacy’, European Journal of Criminology 8 (2011).

• 90 Geenen et al., above n. 49.

• 91 Lindhorst and Padgett, above n. 65.

• 92 Tummers et al., above n. 19.

• 93 Bosma et al., above n. 88.

• 94 Ibid.

• 95 Tummers et al., above n. 19.

• 96 Geenen et al., above n. 49.

• 97 Tummers et al., above n. 19.

• 98 Ibid., at 1110.

• 99 Tummers et al., above n. 19, at 1112.

• 100 Maynard-Moody and Musheno, above n. 68, at S18.

• 101 Tummers et al., above n. 19.

• 102 Maynard-Moody and Musheno, above n. 72, at S20.

• 103 Lipsky, above n. 15, at 15.

• 104 Ibid., at 14.

• 105 Fischer et al., above n. 10.

• 106 Next to the four legal modalities described in section 2, some respondents provided information about penal protection orders during the stage of a conditional release.

• 107 I. Cleven, T. Fischer & S. Struijk, ‘In het belang van het slachtoffer’, 1 Tijdschrift voor Criminologie 11 (2020).

• 108 Van der Aa et al., above n. 12.

• 109 Fischer et al., above n. 10, at 100.

• 110 Kamerstukken II 2009/10, 32319, nr. 3, at 11.

• 111 Fischer et al., above n. 10, at 95. The other two factors that emerged from the interviews are the extent to which the victim feels dependent on the suspect or perpetrator, and the knowledge about the imposition of the order and the method of reporting a violation.