Regional Seminar
“Legal Measures to Combat Violence against Women,
including Domestic Violence”
The Hague, 21-22 February 2007

Speech by Carol Hagemann-White
Professor, Faculty of General Pedagogy and Gender Studies,
University of Osnabrück, Germany

Assessment of the implementation of legal measures

The implementation of legal measures – interaction with social interventions and institutional cultures

Legal and social interventions in European-level policy debate
In the Council of Europe Recommendation Rec(2002)5 on the Protection of Violence against Women, a major section concerns legislation. Member states are urged to ensure that every act of violence against women is punishable, to take swift and effective action against perpetrators, to investigate and prosecute acts of violence, initiating prosecution by the public prosecutor.

With respect to civil law, the recommendation speaks only to the possibility of compensation for damages, which may in some cases be provided by the state, but may de fact be a further measure to penalize the perpetrator. Measures of protection, such as banning a perpetrator from the home, are framed as orders or actions taken by the judiciary; there is no mention of the victim’s right to request such measures or to decide if these fit her needs. The discussion of legal frameworks is thus focussed on what the state should do towards perpetrators of violence.

At the same time, the Recommendation articulates as its fundamental principle that all measures should be based on empowerment of victimized women and their free exercise of basic rights, ensuring all necessary measures of protection, support, and services. Such measures also imply legal frameworks, ranging from civil law redress to the regulation of access to benefits, and the legal framing of this dimension is less clearly spelled out. The two sides of legal strategy – deterrence and punishment of actions not to be tolerated, and protection and support of recovery and of self-determination for victims – interact in several ways. Most obviously, sanctions against interpersonal violence are rarely effective unless witnesses are able and willing to co-operate; the victim is not the only, but a prime witness. Precisely because violence against women is a structural problem in society, it is usually difficult for victims to assist the law in sanctions. Recognition of this difficulty has been a moving force behind building inter-agency networks in many countries at this meeting. Beyond that, however, sanctions against the perpetrator can rebound and increase harm to the victim, when she is, for example, financially dependent on the man who attacked her, or when it is easy for him to intimidate her or take revenge.

There is thus a tension between sanctions and empowerment that cannot be easily resolved. As a growing number of national prevalence surveys show, most of the violence that women experience is exercised by men within their circle of personal relationships; and in general, people are not happy to take a family member to court. Overwhelmingly, what women want is to feel safe from further threat of violence – not incidentally, the Swedish comprehensive Act in 1998 was called “Women’s peace”. Strikingly, assessment of the Domestic Violence Act in Luxembourg found that, while the vast majority of the women saw both the police intervention (including eviction of the perpetrator) and proactive advisory services as beneficial, and most of them then took civil and criminal measures against the perpetrator, only half (52%) felt safe in his presence. If this is the case even where women have been supported in a well-co-ordinated network of interventions, how much less safe must women feel before they enter a “chain of intervention” or where this chain is not sufficiently secured. The tension between empowerment of women towards self-determination and the interest of the state in sanctions and deterrence cannot be fully resolved.

Across the Council of Europe I see some problematical developments during the current wave of efforts to implement transnational commitments. Examples can be cited in which legal frameworks define a confrontation between the state and the perpetrators, while the victims have only a secondary role. (This is the tradition in rape law, and is a prime reason why victims have often found court trials humiliating or traumatic.) In the case of domestic violence, the role of children, always a conflict when there is violence against women in the home, has become particularly ambivalent. When it is recognized that children suffer seeing their mothers victimized, in some countries, the woman who fails to leave her husband can be prosecuted for child maltreatment.

These difficulties are compounded by the fact that the actual effects of legal measures depend very much on the differing ways in which the institutions of policing and justice are organized and on what I will call here the institutional cultures in which they are embedded. I will discuss this in three areas:

Special laws

In the countries at the present meeting, there has been a long history of social activism raising awareness of the problem of violence against women since the 1970s and leading to the establishment of a broad network of service organisations, usually initiated by women’s NGO’s, that gradually won at least partial funding out of public money. Legal reform strategies were developed only later, after the specialized services for victimized women had established themselves as the indispensable foundation for addressing violence against women, and in consequence, gained a strong voice in the political sphere. They speak for the principle, affirmed in the CoE Recommendation, that “All measures should be focused on the needs of the victims”.

In many countries, for example those that more recently entered European institutions, there is no comparable history of social activism, and legislation is seen as a lever to begin the needed process of social and institutional change. Based on UN recommendations, there has been a strong trend to call for specific laws on domestic violence. These are typically laws that define the offences covered, the relationships to be considered “family”, “household” or “domestic”, the penalties and the procedures; a number of them have been passed or are being prepared in countries that recently joined the EU.

The earliest and in some ways strongest was the Cyprus law enacted 1994, revised and extended in 2002, which considers violence in the family to be an offence against the state. Acts causing physical, sexual or mental harm are punished more severely if they occur within the family than elsewhere. Any act of violence committed in the presence of a child is considered violence against that child, being likely to cause psychological harm. In consequence, the spouse – for example, a wife who has been struck by her husband while a child is at home – is a compellable witness against her husband under threat of being prosecuted herself. Family counsellors are given full police powers of investigation and are directly involved in the prosecution, and public officers of health, education and social welfare are required by law to report “concerns, suspicions or evidence of family violence” to the Attorney General. Thus, social measures are centred on the effort to prosecute perpetrators, based on the interest of the state in eliminating violence from the family sphere.

Laws of this type have been spreading especially into countries with little previous activity to address violence against women. The emphasis may be on protecting the family, rather than protecting women. Typically, the definition of domestic violence makes no reference to gender and includes child maltreatment, elder abuse, attacks on siblings and other relations as well as abuse of a spouse or cohabiting partner. It is not at all clear how some very broad definitions of what constitutes “violence”, including inflicting emotional pain, will be implemented in legal practice. Will the expansion well beyond the traditional categories of punishable offences upgrade or downgrade the “domestic”? Acts that would, in another context, be considered criminal violence, may now be classified as domestic violence and possibly carry a lesser penalty. On the other hand, a number of states have defined violence within marriage or the family as an aggravating circumstance, calling for a higher penalty. Overall, it can be said that a specific law addressing domestic or gender-based violence needs to be carefully crafted to ensure that the interaction with existing legal frameworks has the desired effect.

To judge by the number of cross-references, the Spanish organic law represents such an integrative framework. It is specific, in being directed at adult relationships and gender inequality, and is also innovative in taking a truly comprehensive, multi-pronged approach, addressing gender violence in context including educational and prevention measures, procedural as well as criminal and civil law provisions, social and economic rights, setting up fast-track specialized courts, and more. Perpetrators can only be sentenced to prison or to community service accompanied by re-education; fines, which often burden the wife, are not possible. Any conviction for gender violence suspends the right to own a weapon, and may exclude the perpetrator from exercising any parental authority for up to five years.

In all cases, whether a specific law is passed or existing criminal law applied, or a combination of both, monitoring is needed to assess whether the police are investigating all cases that come to their attention, what proportion of cases are actually prosecuted and for what offences, and how courts dispose of the cases. Based on the information given by member states of the CoE on legislation and its implementation, it seems that very few countries have a monitoring system which would enable them to know where the new legal activities are actually leading in practice. Only Spain has set up a central observatory to collect and analyze data on all cases as they move through the policing and legal system; only the UK and Sweden seem to have an inspection system that, at least at intervals, reviews the actions of statutory agencies towards domestic violence and/or towards sexual assault and rape. The UK has computerized police, prosecution and court records, and is now able to “flag” all cases where an offence of any kind is related to domestic violence; this now allows them to track outcomes over time.

Although many countries have replied that they keep statistics by sex and by relationship between perpetrator and victim, they usually do not disaggregate their criminal statistics by a combination of these facts, so that one cannot even locate data on homicides by intimate partners by sex across Europe. Without data, there can be no monitoring, and legal strategies are largely guesswork.

Protection of victims from further violence

Without effective victim protection, sanctions may exist only on paper. You have already discussed the Austrian model at length. Its innovative provisions for evicting the perpetrator from the home have since met with great interest throughout Europe. It has also met with certain reservations, so that modified measures are emerging.

The Austrian model, since implemented in Germany, Luxembourg, and Switzerland, as well as the Czech Republic and Slovenia, authorizes police officers to expel a person who poses a danger to another from the home and prohibit his return for about 10 days, regardless of ownership or whether the persons are related. The decision on eviction and barring orders lies exclusively with the police officers on site, not with the victim, and compliance with barring orders should be enforced. An advisory centre is usually to be informed about the intervention within 24 hours, and will then actively contact the victim and offer her offer information and support, including how to obtain a court order that extends the police ban.

A growing number of Council of Europe states are introducing a measure to evict the perpetrator; indeed, it is rapidly becoming a common element of strategy. However, a majority of countries evidently consider it inappropriate to give this power to the police. Although concerns about police powers are quite different, say, in Norway than in Bosnia, there is an overall tendency to define barring orders, like other safety orders, as a judicial measure. Such court decisions are unlikely to take effect quickly. Many of them are dependent on criminal proceedings already having begun, usually presupposing that the victim will have made a statement and is prepared to testify before any protective measures are taken. (Just to round off the picture: in Poland the court can issue an order to remove the perpetrator from the home only after he has voluntarily moved out!) A few laws, and some legal systems, foresee court decisions as an emergency measure with an ex parte decision.

Whether or not a court-issued barring or restraining order can be effective will depend very much on the way the institutions of policing and justice are understood and how and when they interact. Very little is said about this in country reports on implementing legal measures, because the writers of the reports tend to take the functioning of their own institutions for granted. I am on thin ice here making comparisons, and hope that you will correct me if I am wrong.

Let us take, as a measure for implementation, the “modal case”: A woman is attacked by her partner at 10 p.m. on Friday evening, and although there are no injuries needing medical care, she is obviously hurt and frightened, she may have screamed for help, and someone calls the police.

Before the introduction of the barring order in Germany or Austria, the police were fairly helpless and often very unhappy about being unable to help. In German criminal law, the police are defined as the assistants of the prosecutor, and like the public prosecutor (and unlike the norms of an adversarial legal system), they have to gather both evidence of possible guilt and evidenced of innocence, submit it to the prosecutor for decision on whether there is a crime, and whether it can or should be prosecuted. This decision will not happen rapidly, probably not for at least a week. Only then can it go to court. On the Friday evening, the police have no power to arrest; they could call an emergency judge to issue a warrant, for example, if there is reason to believe the perpetrator might go into hiding, or if a very serious crime is imminent and cannot be prevented any other way. What the police could do was to take a man into custody if he seemed very drunk or out of control, but they could only keep him until he was calm and sober, not more than 8 hours as a rule, so he would be back in the flat, probably angry, by 6 in the morning. The only way to protect a woman was to bring her to a refuge, which the police often did.

This seems to be different in the UK. Awareness that domestic violence is serious has been met by an increase in police arrest. Common assault has been added to the list of offences for which a police officer can make an arrest without a warrant. If the police assess that there has been a crime, or have reason to believe that a crime may be committed, they may even keep the man in jail until taken before a magistrate. Increasingly, the UK is setting up specialized courts and aiming to fast-track domestic violence cases. A man may be brought before the magistrate within 24 hours, and there is a stricter policy on setting bail: When the risk for the woman seem high, he can thus be kept in detention until trial.. To the extent that this system works, or could be made to work by inspection and supervision, it makes the Austrian barring order look superfluous. Why leave the man out there where the woman can never know when he might come back, when you could put him safely in jail?

However, safety is not the prime concern here, but rather branding wrong-doing as such, with a potentially deterrent effect. In actual fact, arrest does not ensure a woman’s safety for very long. The magistrate may only schedule a hearing (or pass a serious case on to High Court), or sentence the man to a fine or to community service. In emphasizing arrest, British policy sees the police as reinforcing the law, and arrest as a relatively normal event that can happen to ordinary people. Possibly, a much larger proportion of the British population (including a large proportion of women victims) than, for example, either in Scandinavia or in Germany, thinks there is no great harm done if someone who has behaved badly has to spend a few days in arrest, or a month in jail: It will teach him to think twice about how he behaves the next time. There may also be a difference in the cultural perception of the role of policing, and how far it is associated with prevention and protection.

There are, of course, weaknesses to prioritizing an arrest response. It assumes that domestic violence situations include identifiable criminal acts and provide enough evidence for a conviction. Most cases are handled by a magistrate, who works as a volunteer, and not by a professional judge. Quite a few men will be let off without a penalty, although the conviction rate is rising. For their own safety, women must actively seek protection orders. This disadvantages the more vulnerable women additionally; to avoid this, significant resources must be invested in supporting women.

In Germany or Austria, there is a very strong aversion to increasing the police power to arrest, or to “fast-track” convictions. This is founded in historical experience of how police powers or special, rapid-conviction courts can be abused. As paradoxical as it may seem, the choice of the police as the ones to issue a barring order is based precisely on concern for strict procedural correctness and presumption of innocence in criminal law. The police ban is founded on the police mandate to protect the life and safety of citizens; it is a preventive measure that the police are required to take when there is reasonable ground to suppose a threat of violence in the home. The fact that an assault has just occurred creates a high probability that further violence threatens, but neither the police ban nor the court injunction requires evidence for criminal prosecution. The police are keeping the peace and protecting people from harm, as they would in case of fire or flood. Criminal accusations must involve the prosecutor, and permit preparation of a defence, thus taking time.

This is also a weakness of the Austrian model: now that the police have an immediate means of relief in hand, they may be disinclined to report crimes and to collect evidence that will stand up in court independent of the victims’ inclination to testify. Many domestic violence situations permit collecting evidence, far more than popular opinion (it’s only his word against hers) recognized. Photographs of the red marks in her face and the state of the room, statements by neighbours, spontaneous statements recorded on the scene are all very useful. We have found in German evaluation that the police, even if well-trained in respectful interviewing and using the eviction option, rarely collect such objective evidence.

There may be a third model emerging. The institution of the investigative judge working closely with the police in France or Belgium, or the assistant prosecutor within the police force in the Netherlands, could make it possible for the police to have rapid access to a judicial order of eviction from the home without waiting for an actual court proceeding and disposal. The key questions are: would a judicial protection order be available in above the Friday night case, in time to keep the woman in safety until she can seek advice, consider her options, make a safety plan, change the lock on her door and ask for a temporary court order in matters such as child contact, exclusive use of the residence? In many of the German-speaking countries/ Länder, the police barring order has to be presented to some further authority for approval; if a judge were involved from the outset, this additional bureaucratic step could be eliminated. (Court orders can, of course, always be contested). Little written material is available on this model and its relevance for domestic violence.

The high level of success of the Austrian model – very few bans are contested and the level of violations is low – is probably founded on its clear distinction between the actions of the state and those of the victim, and a balance between the state use of force and respect for the victim’s right to decide on her own personal life. The state acts ex officio in removing the perpetrator and in giving a specialized social support service the opportunity to contact the victim. It is then up to the woman to use this period of safety and the resource of counselling according to her wishes and felt needs. This carefully thought out balance between the state’s clear rejection of violence and the empowerment of victims needs to be explicated for other models as well.

Finally, it should be noted that the preference for rapid punitive measures or for a priority of protection and victim support is not only a matter of institutional cultures: The results differ as well. Protective measures mean that a woman can call the police before she is actually injured, when she recognizes from past experience that violence is brewing. A policy of arrest and punishment is hard to apply until some violence has occurred or been directly threatened. Arrest policies will be more effective in protecting women from highly violent men; risk assessment is vital to this goal. Police barring orders will offer protection that is more uncertain, but meets a wider range of circumstances, and may prevent violence. Where the risk seems high, police in Austria or Germany thus may still bring a woman to a shelter after barring the man from the home, and the advocacy services then plan for safety with her.

Implementing sanctions

One of the most striking aspects in the evolution of measures on violence against women across Europe is the tendency to define sanctions without a great deal of discussion on how they may be applied. Nearly all 46 countries in the CoE now penalize rape within marriage, at least nominally. Several member states (Greece, France) have recently lifted the marital exemption, and the remainder may be expected to follow. Yet attrition rates have also been rising. The Regan/Kelly report1 based on questionnaire responses from 21 member states for the period 1980-2003 found conviction rates for rape to be sinking – in some cases dramatically – across Europe, while women’s reporting of sexual attacks has increased. In all countries (except Germany), when reporting rose, convictions sank. In a number of countries (such as Austria, Greece and Poland), reporting has not even increased significantly over the past 20 years, suggesting that awareness and confidence in the justice system have not improved.

There are still very few states that actually make lack of consent the measure of rape, as in the UK and in Belgium, where it is a sexual offence if the perpetrator either knows that the other person does not consent, or is reckless regarding consent. The UK has been making changes to address the “unacceptably low” conviction rate of 6%, including specialist rape prosecutors, training for the police, and procedural improvements. (Such changes were made in Germany beginning in the early 1990s, and this may be a reason why the attrition rate is lower there).

No systematic data are available from any country, to my knowledge, on the proportion of domestic violence offences that reach the courts at all, or that end in conviction. Research on files from the State Prosecutor’s Offices in Vienna and Salzburg during 2001 revealed that about half of the legal proceedings following bodily injuries were abated. A German study of preliminary proceedings in two different prosecution services also found that the majority of domestic violence cases were dismissed. In 95 % and 81 % of cases no court action was brought.

If I may be excused for simplifying generations of legal debate: Sanctions deter crime generally by conveying the notion that acts are wrong and socially not tolerated, and specifically by creating a reasonable expectation of punishment. Thus, most people do not rob a bank even if tempted, because (a) they think it is wrong, and (b) they expect they would probably be caught and the consequences be very unpleasant.

Violence against women is not like robbing a bank. It rests on a foundation of centuries in which men exercised violence towards women – towards “their” women, towards “loose” or “available” women – with impunity. Men who batter or rape do not go to any great effort to conceal what they do, nor do they expect, if apprehended, that anything very unpleasant will follow. As long as they do not actually draw blood, they tend not to consider what they have done as being “violence” (cf. the research of Hearn 1998), and are thus unlikely to assume that a law against domestic violence will even apply to them. Women mostly recognize acts that hurt and intimidate as being violence.

If we draft laws, policies and action plans declaring that there will be sanctions for domestic or for sexual violence, the law in itself raises expectations among women, while having (at first) no significant deterrent effect on men, as long as they believe that their actions are not punishable. Raising the penalty level is probably not very useful, since it does not increase, and may even decrease the probability that the penalty will actually be applied in any given case. The focus must be on ensuring (a) that sanctions will be applied, and (b) that men understand that their behaviour will give rise to such sanctions.

Here, the West Yorkshire model of police response to repeat victimization is very promising, and has been assessed as effective. In essence, it consist in identifying repeat offenders – i.e. men whose use of violence has led to police attendance previously – and establishing an automatic increase in the level of intervention with each repeat attendance. This presupposes a database that will allow police to identify, when a call comes in, if the man has used violence against any woman before. Each level of intervention – beginning with an official warning to the perpetrator and rising to court actions – is accompanied by a (standardized) letter explaining the policy and stating what actions could or would be taken. Many cases can be handled on a routine basis, with only a relatively small number requiring special consideration and discretionary assessment. The model is based on early intervention, and – like the Austrian model, but in a different way – avoids the effort of preparing a case for prosecution when the level of demonstrable harm is still low, yet communicates very clearly to the abusive man that his behaviour is unacceptable and both can and will result in sanctions if continued. As a result, the small proportion of chronic offenders, who are the most dangerous, can be filtered out of the numerous calls and be targeted for high attention and special protective measures.

Some countries have been introducing electronic foot bracelets to keep track of perpetrators, and cell phones or other alarm systems for victims. These measures will be useful for women threatened by those dangerous men who do not respond to any other deterrents, but effective use requires that these cases be identified. We know from prevalence studies that up to a quarter of all women experience some physical violence by an intimate partner at some time in their lives. A clear message from the police that this is wrong may help them find their own solution; and many will never call the police at all. Knowledge and skills in risk assessment are essential for the police to identify those cases that might call for special protection measures.

From the global to the local and back again

There is a pressing need for all European countries to empower women, assist them to escape from violence within the family, and provide resources for advice and practical help. This requires establishing and maintaining woman-centred NGO’s to provide assistance and services to victims. Without these, legal action against perpetrators is likely to fail, since women who are intimidated and humiliated will not be able to act as witnesses or to pursue a complaint. Training the police and enforcing high standards of intervention with respect to violence against women also need to be an integral part of the effort to develop good overall services. These are all measures that profit from the engagement of actors on site and “close to the ground”, bringing all stakeholders to the table to identify the most pressing needs for action, as well as developing potential for co-operation and designing programs to fit the local circumstances.

In the longer view, the experience in many countries has been that no single measure is, by itself, effective. In recent years, the network approach has moved to the foreground, as can be seen not only in the UK and Germany, where there has also been significant evaluation research, but also, for example, in France and the Netherlands. There is a shift in focus from „measures and programs“ to developing community intervention strategies through inter-agency cooperation, beginning on a local level. They build on existing services, and on a growing awareness in statutory agencies of the cross-cutting nature of the problem. Building cooperative programs is a process which grows from strong roots in local conditions and human resources. They are also probably the best site for working out how the differing and inter-related needs and rights of the direct and indirect victims, the adults and the children in families, can be balanced. However, it is essential that the knowledge and skills developed in such local cooperation networks be brought back into the strategies and policy-making on the national and transnational level.

Note 1 See footnote 6