file: batter2.htm
No, Do not Always Arrest the Batterer!
Evan Stark
From Evan Stark, "Mandatory Arrest for Batterers: A Reply to Its Critics," in Eve S. Buzawa and Carl G. Buzawa, Do Arrests and Restraining Orders Work? (Sage Publications, 1996). Copyright 1996 by Sage Publications, Inc. Reprinted by permission. Notes and some references omitted.
MANDATORY ARREST FOR BATTERERS:
A REPLY TO ITS CRITICS
THE CONTEXT AND THE PLAYERS
As a long-time activist in the battered women's movement, I welcome the debate about mandatory arrest.
There is a certain irony in defending the police powers of the state against critics who helped lay the basis for the proarrest policy in the first place. Larry Sherman's conversion is the most remarkable. Relying on recent evidence that certain men become more violent after arrest, Sherman has replaced his erstwhile ardor for arrest with an equally passionate belief that mandatory arrest laws should be immediately repealed, especially in cities "with substantial ghetto poverty populations with high unemployment rates." . . . [T]here are profound methodological flaws in both [Larry] Sherman and [Richard] Berk's original Minneapolis Domestic Violence Experiment and the replication studies that are the basis for Sherman's current position....
Sherman's picture is constructed from a bygone devotion to positivist criminology. His is a world defined by the technological optimism of the pre-1960s, yet to be polluted by culture, politics, gender, and history.... In Sherman's [world], causality is singular, universal, and undirectional, and social science research guides professional/government intervention to right injustice. In the post-Law Enforcement Assistance Administration (LEAA) era, this neoKeynesian belief in state intervention appears as naive as an early Doris Day film. Still, there is something seductive about a "preventive conceit" ... that envisions modifying such complex behaviors as violence with minor adjustments in the criminal justice response (arrest vs. no arrest). Would it were so.
. . . Sherman's faith in- the instrumentality of professional intervention-- including the instrumentality of his own research--supports his belief that arrest increases violence. Hence, [the idea that] eliminating mandatory arrest will reduce the violent response [is] almost certainly wrong... because the paradigms from which they arise... discredit the role of social initiative--in this case, the role of the community based women's movement--in shaping the outcome of policy change.
... I suspect that most people view policing as a body of available resources, like schooling or health care, say-- a sort of lottery that circumstances periodically force us to enter. Given the complex political realities that shape crime (as well as, e.g., health, welfare, learning) and mediate how police and the public interact, it is hard to predict whether one's welfare will be helped or hindered in any given encounter or, more globally, what determines how a given set of actors (offenders, patients, students, police) will respond to a specific class of interventions (e.g., mandatory arrest). If people "play" nonetheless (call police, attend school, go to the emergency room), it is not because they believe they will win, but because this is one of the few shows in town at which they can win and because winning--in this case, having police resources at your disposal--is a highly desirable outcome in the long run.
... [O]ne can think of the proarrest strategy as a "basket of goods" that may include everything from a mere warning, handcuffing, or an arrest warrant through a weekend in jail, mandated treatment, a stalker's law, the community intervention programs,... real prison time, the provision of court-based advocates, and so forth.... Sherman... [is] recommending that this keystone to reform be removed from the basket because it fails to fulfill an important policy objective--the reduction of violence. Sherman's implication... [is] that the criminal justice system should no longer be the focal point of society's response to woman battering.
Even before we can ask whether the replication experiments are sufficiently robust to justify abandoning mandatory arrest, we need to question their fundamental premise: that the wisdom of arrest should be assessed solely in terms of its effects on violent behavior, whatever they are. To do this, we need to conceptualize the presumed object of the arrest policy--woman battering--and unpack the "demand" that arrest is designed to satisfy....
BATTERING: THE SOCIAL PHENOMENON
Feminists have hardly been unambivalent about the wisdom of a proarrest strategy. Who is more aware that depending on protective intervention by a male institution is a two-edged sword? As the concern with "wife torture" emerged from the movements in the 1870s to protect animals and children, British feminist Frances Power Cobbe addressed the same questions we are debating today, and she did so, incidentally, with the same sort of evidence.
Cobbe (1878) believed that battering was rooted in women's status as men's property and that the only effective response was full economic and political independence. Women's subordinate position in the public sphere and their private vulnerability were inextricable, each a precondition for the other. Cobbe supported the arrest and prosecution of batterers as means to reduce women's political isolation.... Noninterference when women were assaulted constituted active support for male dominance....
[A]fter studying arrest and court statistics, she concluded that the courts were focusing only on the most extreme cases of violence. As a result, she believed, criminal justice was effectively establishing a permissible level of harm: By punishing only "severe" injury, the court response actually caused the minimum levels of domestic violence to rise....
I share Cobbe's fundamental conviction that violence against women is a political fact. This means that everything about it--when, how, why, and where it is used, whom it affects, the nature of intervention, and most important, the consequences of intervention for all involved --reflects the relative power of men over women and the struggles by particular men to assert and women to escape this power. Whether this violence is expressed through child sexual abuse, harassment, rape, or battering, the key is the selection of females as victims on the basis of their gender. Because of its roots in sexual inequality, whatever occasions violence in a given encounter, the ultimate cause of "battering" (as well as its consequence) is the denial of women's civil rights....
In trying to conceptualize battering, we need to picture ongoing forms of control that are at once both personal and social, including economic exploitation, isolation from family and friends, intimidation, and a host of rules governing everyday activities. Ann Jones and Susan Schecter use the term coercive control to describe the systemic fusion of social and individual dominance that undermines the physical, psychological, or political autonomy of even the strongest, most aggressive and capable women....
The term entrapment describes the cumulative effects of having one's political, social, and psychological identity subordinated to the will of a more powerful other who controls resources that are vital to one's survival.
BATTERING: THE CRIME
... [A]lthough safety and the reduction of fear remain important goals, their realization depends on making women's "empowerment" the ultimate standard against which the efficacy of various interventions is judged.
... lB]attering, the social phenomenon, occurs at three levels simultaneously: the political level of female subordination, the level of interpersonal assault, and the level of coercive control at which women's social vulnerability is exploited for personal gain. Battering, the experience, arises from the particular ways in which these three levels interrelate in a given relationship over time. The challenge is for criminal justice to recontextualize the sorts of disembodied acts of assault recorded by medicine, such as "punched with fist."... [I]n terms of entrapment and control, Sherman recognizes that this episodic focus also governs police work. The same could be said of the calculus of harms that guides criminal proceedings in the courts. Lacking a conceptual frame to understand the historical nature of battering, police, judges, physicians, and other professionals fall back on kitsch psychology or on cultural stereotypes to explain the patterns they observe among individuals, families, or entire groups who appear "violence prone." Absent a theory of coercive control, the traumatic effects of battering seem to be derived from victim psychology...
Because the element of control is what links the assaultive dimensions of abuse to the political fact of female inequality, there can be no hope of preventing battering simply by regulating the degree of violence. This is why we call for "zero tolerance" of force in interpersonal relationships and oppose basing police intervention on a calculus of physical harm....
Reframing battering, the crime, in terms of inequality, coercive control, and entrapment allows us to think in new ways about mandatory arrest. The critics make three powerful points:
1. Mandatory arrest does not work. A1though arrest might reduce violence in some instances, it may actually increase it in others. Only in allowing police to exercise informed discretion --what Sherman termed structured discretion--can we meet the objectives of deterrence....
2. Mandatory arrest is inhumane. Arrest does little for the victim and less for the offender. Worse, it jeopardizes the fundamental integrity of family life. It is far better to limit arrest to the most dangerous cases, use treatment for batterers, and offer a range of compassionate family supports....
3. The very people we are trying to protect do not want it. It appears contradictory for us to value a woman's claim to liberty as a fully endowed citizen and then to devalue her assessment of a policy carried out in her name, such as mandatory arrest. Allowing women --and police--the discretion to decide whether arrests should occur satisfies the consumer interest of the former and sustains the morale of the latter.
DOES MANDATORY ARREST WORK?
In one of the more dramatic self-critiques in the history of criminal justice research, Larry Sherman, an architect of the Minneapolis Domestic Violence Experiment, now argues that mandatory arrest does not deter domestic violence, except among very select groups (e.g, married, employed men). If the replication experiments failed to show the deterrent value of arrest, they did show, Sherman claims, that arrest actually increases violence in the long run, particularly among unemployed, unmarried, and minority males. Although Sherman bemoans the rush to implement the earlier research findings, on the basis of current findings he argues that as a general practice mandatory arrest should be ended immediately, particularly in cities with large ghetto populations. In place of mandatory arrest, he favors "structured police discretion," presumably so that alternatives can be used in cases in which arrest has been shown to increase risk. Because one logical outcome of this position is clearly absurd--namely, to arrest only white, married, employed batterers--in effect, Sherman's argument would virtually end arrest in domestic violence cases, at least in all major urban centers, except in the most severe cases.
No mea culpas are required from Sherman because the results of his Minneapolis Domestic Violence Experiment were misinterpreted....
[T]he major importance of the Minneapolis Domestic Violence Experiment was to give women's advocates (who already favored arrest) a powerful weapon to use with lawmakers who viewed domestic violence, like child abuse, as a social welfare rather than a criminal justice problem.... [T]he Minneapolis results were accepted so uncritically because, as a management strategy, police discretion in domestic violence cases had already become politically untenable.
But what did the battered women's movement hope to accomplish through arrest? After all, having established an unprecedented woman-run, community based alternative service movement, why risk losing our identity by placing so much emphasis on a male-run system that... we located on the far right of the political spectrum in terms of community participation, access to influence, attitudes toward women, and bureaucratic isolation?
The first reason for mandating arrest was to control police behavior.... The fact is that, in disregarding battering, minimizing its consequences, and blaming the victims of abuse, police were no different from other professional groups.... [T]here can be little question that police behavior has changed significantly with mandatory arrest. Whereas street-level resistance by police remains widespread, arrests for assault had risen 70% between 1984 and 1989, largely because of domestic violence laws.
However individual officers responded to calls for help, the absence of a standard for police practice increased women's sense of powerlessness and thus posed a major obstacle to their empowerment....
The interest in controlling the police response had as much to do with accountability to shelters as to women generally. The legal mandate that expands the package of services that shelters can offer, including services to police, increases the safe mobility of shelter residents and advocates, provides a rationale for a regular shelter presence in the courts, and lays an empirical basis that shelters can exploit for custodial orders, orders of protection, and expanded services. These functions, in turn, provide the substantive basis for ongoing negotiations among shelters, the criminal justice system, and family and children's services, negotiations that have been formalized in many cities through coordinating councils, networks, or other forums that assume quasi-official oversight responsibility for the law. Because shelters often secure temporary restraining orders (TROs), their power has been greatly enhanced in places where police liability for failing to arrest has been extended to failure to enforce orders of restraint and protection. Liability is a particularly important source of redress....
The second reason for mandatory arrest involved immediate protection from current violence. Arrest provides a meaningful opportunity for battered women to consider their options and gives those women ready to end the relationship time to go elsewhere or to obtain a protective order. The amount of time the batterer is physically out of the picture is crucial. In Milwaukee, for example, a woman was three times as likely to be assaulted if police left without an arrest than if the batterer was arrested and then released....
Third in importance was the desire to reduce the overall incidence of domestic violence both directly, because arrest might deter recidivism, and by sending a clear message that battering was unacceptable. Whereas the replication studies clearly do not support the Minneapolis findings, other studies do. In Lincoln, Nebraska, for example, implementing a mandatory arrest law reduced recidivism dramatically--from 83% to 53%....
According to deterrence theory, the function of mandatory arrest is to convey the criminal nature of battering. Although arrests have risen for domestic violence, because prosecution and sentencing patterns in domestic assault cases have changed very little since battering was criminalized, it is unclear what the NIJ [National Institute of Justice] experiments were actually measuring. The results could just as readily be interpreted to mean that arrest is ineffective in isolation from other sanctions. Indeed, just as violence may increase where orders of protection are unenforced, so too is it likely that arrest without serious followup is interpreted by some as a license to abuse.... The policy of mandatory arrest also has the indirect function of setting a standard of zero tolerance for battering that other institutions can emulate.
Fourth, making battering the only crime in which police discretion is removed acknowledges a special social interest in redressing the legacy of discriminatory treatment of women by law enforcement. It also forces the law to juxtapose the subjective experience of women and its traditional accommodation to the interests of propertied (male) strata. Setting the crime of battering apart in this way also helps distinguish battering from the two sets of crimes with which it is commonly confused: familial abuse, in which the victim is a minor or a frail elderly dependent, and assaults or muggings by strangers. Mandating arrest communicates how seriously we take the crime of battering and may have an effect on subsequent violence that is independent of arrest as such.... Mandating equal protection in this way also helps counter the general reluctance of courts to extend to women the civil rights protection granted to racial minorities.
A final reason for supporting mandatory arrest was what might be termed its "redistributive" function: the perception that police service is a resource that had heretofore been hoarded by others and now should be made available to women on a more egalitarian basis....
The same historical and political context in which the proarrest strategy originated explains its uneven implementation. Differences in community response to mandated arrest reflect differences in this context.... Pennsylvania provides an interesting test case for the political hypothesis: There, despite the absence of a mandatory arrest law and the relative backwardness of the state's judiciary, powerful oversight of police practice by the battered women's coalition has made arrest of batterers routine.
The Minneapolis Domestic Violence Experiment also reflected this;... no area more supportive of battered women could have been chosen to prove the efficacy of arrest. But. . . the fact remains that the researchers, the Police Foundation, the NIJ, and those who supported and attacked the findings from Minneapolis failed to set their evidence in a historical, structural, or political context that could have made the contingent nature of deterrence intelligible to local policymakers. Whatever the replication experiments may or may not say about arrest, they speak eloquently to our failure to learn from our mistakes because they, like the original experiment, do not consider the political chain that might help us understand why arrest works in one setting but not in another. At a time when community support for police efforts is widely recognized as the single most important factor in police effectiveness, it is surprising that the NIJ gave no consideration to the political context that set the tone for enforcement either in selecting the research sites to replicate Minneapolis or in evaluating the findings. Shifting the blame to policymakers for implementing arrest prematurely simply makes it easier to curtail arrest now--a right hard on by the battered women's movement --amid broad cuts in other services to women.
What are we to make of Sherman's conclusion that arrest works mainly with men who are employed and married and that recidivist violence after arrest in certain cities appears highest among minorities and the unemployed? Obvious structural factors are at work here. The correlation between crimes of violence (including assault) and the business cycle has been well known in criminology for some time. Because the replication experiments were conducted when the economy was entering the worst recession in three decades, with the most dramatic effects on minorities and the working poor, the most appropriate baseline is not prearrest levels of violence, but rather rates of violence in similar cities where no change was made in police policy....
[A]lthough the employed, white, middle-class husband may be less violent following arrest, he is also more able than his unemployed, black counterpart to leverage resources other than physical force to continue subjugating his wife. Open warfare may be temporarily over in his household, but the control and coercion go on undisturbed. Has arrest worked in this situation?
This point raises the larger issue of what the replication experiments mean even if we accept the evidence as Sherman presents it. Our assumptions about the three levels of battering give us no hint what a change in postarrest violence signifies in terms of coercive control, entrapment, and/or inequality.... [V]iolence may cease because, following arrest, threats of repeat violence may be sufficient to maintain a power advantage over a spouse, as in the case of the employed middle-class husband described above. Conversely, violence may increase because women are empowered by arrest and/or threaten to leave or to have the man arrested again. On the basis of evidence that abuse escalates when males perceive their control threatened, as during separation and divorce, we would predict that postarrest violence would be greatest--as Sherman's data show it is--among those who are least integrated into the job market or other structures from which men garner their authority--namely, the poor, black, or unemployed. Again, violence may increase in this group because arrest works to empower women and undermine male control....
Evidence suggests that even men who become angry following arrest may be no more violent as a result and that men who are violent after arrest are not necessarily angrier than others about arrest itself. Even if we assume that arrest provokes greater violence among a substantial number of men, Sherman's proposal that we respond by eliminating mandatory arrest appears perverse (to say the least). A far more humane and rational response would be to expand protection for the most vulnerable female populations-- namely, single, minority, and low-income women--as well as to better integrate batterers into opportunity structures without diminishing women's access to resources. The bottom line, however, is that assault is merely one among many means available to men in battering relationships and that its absence, even for some extended period, may signify greater equality or greater dominance.
Let's assume that violence increases following arrest among certain groups and that little can be done to inhibit it. Even this worst case scenario barely affects the most important rationales for mandatory arrest--namely, controlling police behavior, setting a public standard for police response, offering immediate protection, embodying women's civil rights claims, and affording women access to a new "package of resources."...
[O]ther data sets... indicate that mandatory arrest may meet a number of our more important objectives. With respect to protection and control, for instance, recent Supreme Court and appellate court decisions suggest that battered women in states with mandatory arrest laws have far stronger claims to police protection than do battered women in states without such statutes. Furthermore, states with broad mandatory arrest laws are far more likely to allow liability of police officers than might states without such laws, such as New York.
Minority women deserve particular attention because they are most likely to use the service system (including police) and to suffer powerlessness as a result of institutional discrimination. .. . [T]he replication experiments indicate that minority women bear the major brunt of postarrest violence. As a result, argues Sherman, minority women would be the major beneficiaries if mandatory arrest is eliminated.
That arrest poses a series of unique dilemmas to minority women goes without question. But if police were reluctant to intervene generally in domestic violence cases prior to legal reform, they provided virtually no protection to blacks. ... [M]any police conceive of blacks as what sociologist Darnel Hawkins terms "normal primitives" and only intervene in "domestics" when violence overflows into the public arena. This is one reason why violence--almost always by a spouse or a lover--is the major cause of death among black women under 44 years of age. That such attitudes are selffulfilling is suggested by Sherman's observation that police can "predict" which couples (and which apartment build-ings!) are chronically violent. Eliminating mandatory arrest, then, would do little more than reestablish a brutal status quo....
CRIMINALIZATION AS SOCIAL CONTROL
[Here is] a final argument against mandatory arrest: that it reflects a trend to criminalize behaviors and, as such, is a conservative effort to extend "social control." . . . Echoes of the social control argument can... be found in Sherman's... warnings against interference in "family privacy."
... That such intervention may... become a vehicle for imposing other systemic biases (e.g., race or class prejudice) highlights the problems with enforcement but is not an inevitable consequence of society's legitimate interest in limiting unacceptable behavior. Nor is the control function of the law a reason to abandon women's larger justice claims for the equitable distribution of criminal sanctions.
Like legal control, criminalization may or may not advance personal liberty and social justice....
With respect to both control and criminalization,... there is simply no way to avoid the difficult process of making political decisions about where scarce resources will be distributed and on whose behalf. If we think of justice as a good in itself, not merely as a means to an end, we can ask whether the mandatory arrest of batterers represents a more or less equitable distribution of this good.
There is little question in my view that the mandatory arrest of batterers represents a progressive redistribution of justice on behalf of women. One measure of this is a growing suspicion among social scientists and policymakers that, in intervening to counteract the illegitimate coercive power of particular men, the law is being used "politically" to further the larger goals of sexual equality. Would it were so.
REFERENCES
Cobbe, E R (1878). Wife torture in England. Contemporary Review, 32, 55-87.