Emory Law Journal

Engendering Fairness in Domestic Violence Arrests: Improving Police Accountability Through the Equal Protection Clause
Niji Jain Notes and Comments Editor, Emory Law Journal; J.D. Candidate, Emory University School of Law (2011); B.A., Stanford University (2008). I would like to thank Professor Victoria Nourse for her guidance in developing the idea for this Comment, Professor Kay Levine for her mentorship and help in clarifying my writing, and Professor Martha Fineman for her advice and support. My sincerest appreciation also goes to the executive editorial board of the Emory Law Journal, particularly to Laura Pisarello and Jena Sold, for their wisdom, editorial expertise, and patience. Finally, I could never have completed this project without the love and confidence of my family and friends.

Abstract

When police decline to respond to reported violations of restraining orders, victims of gender-based violence and their children suffer tragic consequences. Congress enacted 42 U.S.C. § 1983 to remedy problems of this sort by lifting the shield of immunity when a state actor violates an individual’s constitutional rights. A credible threat of liability for police officers is imperative to encourage police to act in a way that protects individuals from harm. However, the Supreme Court has substantially limited the possible § 1983-based causes of action a victim of gender-based violence can bring against a police officer. The only remaining avenue for liability is an equal protection claim. However, equal protection challenges have met such consistent rejection in the lower federal courts that many scholars believe they are not a tenable strategy for police accountability.

This Comment develops an equal protection claim distinct from those that have failed in lower courts. Rather than relying on disparate impact-based arguments, this Comment formulates a new equal protection claim by reframing the challenged state action and the basis of classification. The claim challenges an officer’s use of gender stereotypes in responding to victims. This Comment will show that basing a discretionary decision about how to respond to a victim on a stereotypical assumption about women is an unconstitutional gender classification that fails intermediate scrutiny and therefore violates the victim’s right to equal protection.

Introduction

In 1999, Jessica Gonzales obtained a domestic violence restraining order 1Town of Castle Rock v. Gonzales, 545 U.S. 748, 751 (2005). On May 21, 1999, a Colorado trial court issued a temporary restraining order prohibiting Mr. Gonzales from “molest[ing] or disturb[ing] the peace” of Ms. Gonzales and their children, ages seven, nine, and ten, and from coming within one hundred yards of the family home. Id. On June 4, 1999, with both Mr. and Ms. Gonzales present in court, the judge made the order permanent as part of their divorce proceedings after modifying it to allow Mr. Gonzales “a midweek dinner visit arranged by the parties.” Id. at 752–53 (quoting Gonzales v. City of Castle Rock, 366 F.3d 1093 (10th Cir. 2004) (internal quotation marks omitted)). protecting herself and her three daughters against her abusive ex-husband, Simon Gonzales. 2Brief for Petitioner at 6, Gonzales v. United States, Case 12.626, Inter-Am. Comm’n H.R. (2008) [hereinafter Gonzales Brief]. Jessica and Simon Gonzales had been married for nine years, during which time they had three daughters. Id. Since 1996, Mr. Gonzales’s behavior had become increasingly “erratic.” Id. Ms. Gonzales decided to file for divorce in 1999, after Mr. Gonzales attempted to hang himself in the family garage and Ms. Gonzales determined that she needed to take steps to protect herself and her children from his behavior. Id. In the month after she obtained the restraining order, Ms. Gonzales called the local Castle Rock police on four or five occasions to report serious violations of the order by Mr. Gonzales. 3 Id. at 9. On none of these occasions did the Castle Rock police arrest Mr. Gonzales, despite having probable cause to do so. 4 Id. A month after the order was issued, Mr. Gonzales kidnapped the three girls from the front yard of the family home. 5 Castle Rock, 545 U.S. at 753. When Ms. Gonzales realized the girls were missing, she immediately called the police department because she suspected that Mr. Gonzales had taken them. Despite Ms. Gonzales’s indication to the dispatcher that Mr. Gonzales’s actions were in violation of a restraining order, the Castle Rock police took no action in response to Ms. Gonzales’s request for help. 6Gonzales Brief, supra note 2, at 10, 11. When two police officers were dispatched to her home two hours later in response to a second phone call, they told her they could not do anything, despite the restraining order, because the children were with their father. 7 Castle Rock, 545 U.S. at 753; Gonzales Brief, supra note 2, at 12.

Later that night, Ms. Gonzales reached Mr. Gonzales on his cell phone, and he informed her that he had the girls at an amusement park in Denver. 8 Castle Rock, 545 U.S. at 753. Again, Ms. Gonzales contacted the police. But the officer with whom she spoke refused to send anyone to the amusement park, instead telling her to wait and see if Mr. Gonzales returned the girls. 9 Id. According to Ms. Gonzales, the officer refused to send someone to the amusement park because it was outside of the Castle Rock Police Department’s jurisdiction, and refused to put out an all points bulletin—an electronic dissemination of information about a wanted person—because it would “needlessly go statewide and would cost the state money.” Gonzales Brief, supra note 2, at 14. At midnight, when the children still had not returned, Ms. Gonzales drove to Mr. Gonzales’s apartment to see if he and the children were there; when she found no one at home, she drove to the police station to make an incident report. 10 Castle Rock, 545 U.S. at 753. The officer to whom she submitted the report “made no reasonable effort to enforce the [restraining order] or locate the three children. Instead, he went to dinner.” 11 Id. at 754 (internal quotation marks omitted).

At approximately 3:20 a.m., Mr. Gonzales arrived at the police station, parked his pick-up truck, and opened fire on the station with a semi-automatic handgun that he had purchased earlier that evening. 12 Id. The police fired back, killing Mr. Gonzales. Inside his truck, they found the dead bodies of all three Gonzales daughters. 13 Id.

Although the sad facts of Castle Rock occurred more than ten years ago, such cases remain salient because the Supreme Court has held that the federal government has no substantive role in regulating and preventing violence against women. 14 See United States v. Morrison, 529 U.S. 598, 598–99 (2000). In Morrison, the Court struck down a provision of the Violence Against Women Act, which created a federal cause of action allowing victims of gender-based violence to sue their perpetrators in federal court, as an unconstitutional exercise of congressional power under both the Commerce Clause and section 5 of the Fourteenth Amendment. Id. The Court held that such a remedy encroached too far upon states’ rights despite extensive congressional findings that gender-based violence affects interstate commerce and that pervasive bias in state justice systems against victims of gender-motivated violence was resulting in systematic underprosecution of gender-based violence crimes. Id. The attorneys general of thirty-six states had endorsed this remedy provision as “a particularly appropriate remedy for the harm caused by gender-motivated violence” in light of the “States’ own studies demonstrat[ing] that their efforts to combat gender-motivated violence, while substantial, are not sufficient by themselves.” Brief of the State of Arizona et al. as Amici Curiae Supporting Petitioners, United States v. Morrison, 529 U.S. 598 (2000) (Nos. 99-5, 99-29). This Comment begins with the premise that the police should have done more to enforce Ms. Gonzales’s restraining order. Failure to do so resulted in the death of three of the four individuals protected by a court-issued order of protection and significant emotional trauma to the fourth. One way to encourage the enforcement of restraining orders is to present police with a credible threat of liability if they fail to do so. This Comment develops a new litigation strategy under § 1983 for victims who are harmed by, and wish to hold police accountable for, police failure to enforce their restraining orders. Part I of this Comment describes the present state of police unaccountability, including the Supreme Court’s rejection of both substantive and procedural due process theories for police liability under § 1983. Part II analyzes the failure of equal protection-based § 1983 challenges in lower courts. Part III presents an equal protection argument distinct from those that lower courts have thus far heard and rejected. This claim argues that police may no longer use gender stereotypes when responding to violations of restraining orders because this use violates a victim’s rights to equal protection of the law.

I. Police (Un)Accountability Today: A Sad State of Affairs

Police accountability to survivors of gender-based violence, 15This Comment uses the terms domestic violence and gender-based violence interchangeably. Many feminist scholars prefer the term gender-based violence because it is more inclusive and attempts to remove the idea that gender-based violence is simply a “domestic” problem not worthy of legal or community attention. However, both terms are widely used in legal scholarship and in litigation documents. particularly with regard to the enforcement of restraining orders, 16This Comment will also use the terms restraining order, protective order, and order of protection interchangeably. State statutes generally refer to such court-issued orders using one of these terms, and the differences between them is not relevant to the issues presented here. is at a shockingly low level today. This unaccountability is the result of layers of immunity and judicial reluctance to infringe on that immunity. Section A provides background information regarding restraining orders for gender-based violence situations. Section B further discusses the historical and continuing obstacles to police accountability in this area due to immunity, while section C reviews past attempts to overcome immunity and explains the reasons these attempts have failed. Finally, section D explains why police enforcement of restraining orders is necessary to protect gender-based violence victims, illustrating the need for police liability.

A. Historical Background and Restraining Orders Today

The facts of the Castle Rock case are egregious but sadly common. Nationally, there is a consistent lack of police enforcement of restraining orders for gender-based violence victims. 17Peter Finn & Sarah Colson, Civil Protection Orders: Legislation, Current Court Practice, and Enforcement, in Legal Interventions in Family Violence 43, 43 (1998) (noting that one major limitation of civil protection orders is their “[w]idespread lack of enforcement, even though such orders allow for expansion of police arrest powers and increase the ability of police to monitor repeat offenders”). Sixty percent of restraining orders are violated within one year, and twenty-nine percent of these violations involve severe violence. 18 Nat’l Ctr. for Victims of Crime, Fact Sheet: Orders of Protection 2 (2005), available at http://www.ncvc.org/ncvc/AGP.Net/Components/documentViewer/Download.aspxnz?DocumentID=39669. Any violation of a restraining order is an offense for which the restrained party may be arrested. 19 Id. However, only forty-four percent of violations result in arrest, 20 Id. and some studies have found the arrest rate to be as low as twenty percent. 21Robert J. Kane, Police Responses to Restraining Orders in Domestic Violence Incidents: Identifying the Custody-Threshold Thesis, 27 Crim. Just. & Behav. 561, 563 (2000).

Historically, police indifference to pleas for help from domestic violence victims stemmed from the notion that domestic violence was a private or family matter in which police should not interfere, despite victims’ requests for help. 22G. Kristian Miccio, If Not Now, When? Individual and Collective Responsibility for Male Intimate Violence, 15 Wash. & Lee J. Civil Rts. & Soc. Just. 405, 434 (2009). Indeed, well into the 1960s, many police training manuals instructed officers to help resolve disputes rather than protect victims, and encouraged officers to avoid making arrests in marital and family violence situations by giving parties time to cool down or encouraging batterers to take a walk around the block. 23Deborah L. Rhode, Social Research and Social Change: Meeting the Challenge of Gender Inequality and Sexual Abuse, 30 Harv. J.L. & Gender 11, 17–18 (2007).

Since the battered women’s movement of the 1960s and ’70s, all fifty states 24Am. Bar Ass’n Comm’n on Domestic Violence, Domestic Violence Civil Protection Orders (CPOs) by State, Am. Bar Ass’n (June 2007), http://www.abanet.org/domviol/docs/DVCPOChartJune07.pdf. have enacted legislation criminalizing domestic violence and allowing judges to issue civil restraining orders 25 Nat’l Ctr. for Victims of Crime, Legal Ser. Bulletin No. 4, Enforcement of Protective Orders 1 (2002), available at http://www.ncjrs.gov/ovc_archives/bulletins/legalseries/bulletin4/ncj189190.pdf. The consequences of violating an order of protection vary by jurisdiction; they often include arrest and charges of civil or criminal contempt, misdemeanor, or even felony. Id. at 1–2. to protect gender-based violence survivors and their children when there is a risk that the abuser will inflict physical harm on his or her partner. 26 See Eve S. Buzawa & Carl G. Buzawa, Domestic Violence: The Criminal Justice Response 235 (3d ed. 2003). Unfortunately, these protective measures have done little to help many victims of abuse because of police apathy and resistance in responding to calls for help.

In response to this continuing police indifference, many states have taken steps to eliminate police discretion in situations of gender-based violence. Thirty-two states have enacted statutory “mandatory arrest” provisions when an officer has probable cause to believe that an individual has violated a restraining order. 27 See Miccio, supra note 22, at 420. Colorado, where the Gonzales tragedy occurred, is one of these states; the Colorado statute 28The text of the Colorado statute reads: “A peace officer shall arrest, or, if an arrest would be impractical under the circumstances, seek a warrant for the arrest of a restrained person” when the restrained person has violated a provision of the order. Colo. Rev. Stat. § 18-6-803.5(3)(b) (1999). instructs officers that they “shall use every reasonable means to enforce this restraining order.” 29Gonzales Brief, supra note 2, at 7. This language was included, in all capital letters, on the back of the restraining order issued to Ms. Gonzales by a Colorado court. 30The mandate on the back of the restraining order states, “YOU SHALL USE EVERY REASONABLE MEANS TO ENFORCE THIS RESTRAINING ORDER. YOU SHALL ARREST, OR, IF AN ARREST WOULD BE IMPRACTICAL UNDER THE CIRCUMSTANCES, SEEK A WARRANT FOR THE ARREST OF THE RESTRAINED PERSON WHEN YOU HAVE INFORMATION AMOUNTING TO PROBABLE CAUSE THAT THE RESTRAINED PERSON HAS VIOLATED OR ATTEMPTED TO VIOLATE ANY PROVISION OF THIS ORDER . . . .” Id. (emphasis omitted).

Despite the strong wording of such statutes, police regularly fail to follow these legislative mandates. In these situations, it is difficult for victims to hold police accountable even when their refusal to enforce restraining orders results in egregious or fatal harm to victims of gender-based violence and their children. The next section discusses these barriers to accountability.

B. State Immunity from Claims of Ineffective Enforcement

The Eleventh Amendment confers sovereign immunity upon state governments, protecting them from suit in both federal and state courts. 31 Erwin Chemerinsky, Federal Jurisdiction 405–06 (5th ed. 2007). Local government entities, which include most police departments, are not protected by sovereign immunity. 32Monell v. Dep’t of Soc. Servs., 436 U.S. 658 (1978). However, most jurisdictions have statutory provisions or caselaw stating that neither individual officers nor municipal entities may be held liable for a mere negligent failure to prevent the commission of a crime, unless the police had a special relationship with the victim, which exists only when the situation involves state-created danger or state custody of the victim. 33Miccio, supra note 22, at 429–30. Therefore, unless a state legislature expressly lifts immunity by creating a specific cause of action against the police, 34For further discussion of this possibility, see infra text accompanying notes 61–65. a plaintiff-victim cannot hold an officer legally accountable for any state law tort. 35Miccio, supra note 22, at 428. The Supreme Court’s description of the public duty doctrine has been adopted by most state legislatures. Id. at 429. Because of the vast scope of police immunity, individuals harmed by police failure to respond or protect have little recourse, even when police abuse their discretion and disobey mandatory arrest laws.

One of the only avenues through which individuals can challenge police conduct is 42 U.S.C. § 1983, which creates a federal cause of action when a person acting under color of state law causes the deprivation of an individual’s federal constitutional or statutory rights. 3642 U.S.C. § 1983 (2006) (“Every person who, under color of any statute, ordinance, regulation, custom, or usage, of any State . . . subjects, or causes to be subjected, any . . . person . . . to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress.”) A cause of action against an individual police officer under § 1983 requires a plaintiff to show (1) that the individual was acting under color of state law, 37Monell v. Dep’t of Soc. Servs., 436 U.S. 658 (1978). Section 1983 creates a cause of action against individual officers who violate an individual’s constitutional right while acting under the color of law. Chemerinsky, supra note 31, at 492. Historically, the cause of action was limited to individual officers, but the Court has since broadened its scope to allow a municipality (but not a state government) to be considered a person for § 1983 purposes. Id. at 493, 508–09. However, municipalities may not be held vicariously liable for the actions of individual officers; municipal liability extends only to official policies or customs that cause violations of the Constitution or federal law. Id. at 493. and (2) that the conduct caused the deprivation of a constitutional or federal statutory right 38Although a federal statute could theoretically provide a cause of action under § 1983, the Supreme Court’s decision in Morrison forecloses this avenue for relief. See supra note 14. of the plaintiff-victim. 39However, even if a plaintiff is successful in bringing a § 1983 claim, her options for relief are limited. While the text of § 1983 contemplates monetary damages as well as injunctive and declaratory relief, plaintiffs must clear an additional hurdle if they wish to recover damages from a police officer. Police officers enjoy qualified immunity from civil suit for injuries caused by their discretionary conduct. Pearson v. Callahan, 129 S. Ct. 808, 815 (2009). The purpose of the qualified immunity defense is to protect state actors from the perpetual fear of liability for every exercise of discretion, which would hinder their ability to fulfill their public function. Kit Kinports, Qualified Immunity in Section 1983 Cases: The Unanswered Questions, 23 Ga. L. Rev. 597, 600–01 (1989). The protection provided by qualified immunity is vast: it is an affirmative defense that insulates officers from being sued at all, permitting complete dismissal of a claim against them. Pearson, 129 S. Ct. at 815. The shield of qualified immunity can be lifted—and an officer exposed to liability for damages—only if the constitutional right violated by the officer was “clearly established.” Harlow v. Fitzgerald, 457 U.S. 800, 818 (1982). The field of jurisprudence surrounding § 1983 litigation and particularly the qualified immunity defense is extremely complicated and in constant flux. Sheldon Nahmod et al., Constitutional Torts 461 (3d ed. 2010). A complete discussion of these issues is beyond the scope of this Comment. However, it is important to note that a plaintiff-victim would have to show that an officer violated her right to equal protection, and that this particular equal protection right was clearly established, before she could recover any compensatory or punitive damages from an individual officer. While the potential for relief under § 1983 is theoretically expansive, the Supreme Court’s jurisprudence in this area has made it a complex remedy with limited practical application.

C. Past Attempts to Overcome the Barriers of Immunity

Over the years, the Supreme Court has limited the availability of § 1983-based recourse for domestic violence victims by narrowing the scope of police actions that may qualify as a deprivation of a constitutional right. 40 See Miccio, supra note 22, at 416. The Court has rejected claims made by victims of violence that police failures to protect violate their constitutional rights to either substantive or procedural due process, cutting off two of the three 41The third is an individual’s right to equal protection of the law. This will be discussed in much further detail in Parts II and III infra. constitutional violations that would form the basis of a § 1983 challenge. Subsections 1 and 2 summarize these past attempts under § 1983 to hold police liable for their failure to protect victims of violence and describe the reasoning used by the Court to reject these claims. In the wake of these decisions, many commentators have recommended that state courts and state legislatures act to increase police accountability. The third subsection discusses these proposals in more depth and then explains why they would fail.

1. Substantive Due Process

The Supreme Court has denied a plaintiff-victim’s right to pursue state accountability for a failure to protect under the doctrine of substantive due process. In DeShaney v. Winnebago County Department of Social Services, the Court denied recovery in a § 1983 suit against a county agency for failure to protect under a theory of substantive due process. 42489 U.S. 189 (1989). The Court held that “a State’s failure to protect an individual against private violence simply does not constitute a violation of the Due Process Clause.” 43 Id. at 197. The Court decided that the Due Process Clause does not obligate the state to affirmatively provide protective services to an individual, 44This notion has become known as the public duty doctrine, which asserts that the police owe a duty to protect the public in general, as opposed to individual victims. Miccio, supra note 22, at 428. unless a “special relationship” existed between that individual and the police. 45 DeShaney, 489 U.S. at 197–99. A special relationship does not exist between the State and an individual simply because the “State learns that a third party poses a special danger to an identified victim, and indicates its willingness to protect the victim against that danger.” 46 Id. at 197 n.4. The Court thus limited the scope of the special relationship doctrine to instances in which an individual is in the custody of the state or the state is responsible for creating the danger posed to the victim. 47 Id. at 200; Miccio, supra note 22, at 428. The Court emphasized the distinction between a constitutional duty to protect, which could render the state liable under a § 1983 due process cause of action, and a duty to protect under state tort law, which could give the victim a cause of action in state court for a breach of duty under state law. 48 DeShaney, 489 U.S. at 201. For a more detailed discussion of state tort liability, see also infra text accompanying notes 61–65. The law in this area is unlikely to change because of the vast acceptance of the public duty doctrine. 49Miccio, supra note 22, at 428–29; see also Hudson v. Hudson, 475 F.3d 741, 745 (6th Cir. 2007) (rejecting, on similar facts, the plaintiff’s arguments that the police could be liable to the victim under either the special relationship or the state-created-danger theory).

2. Procedural Due Process

In 2004, the Court likewise rejected procedural due process as a means of holding police liable for failure to protect gender-based violence victims. After her daughters were killed, Jessica Gonzales sued the Castle Rock Police Department, as well as several individual officers, under § 1983 for violations of her procedural due process rights. She argued that “she had a property interest in police enforcement of the restraining order against her husband; and that the town deprived her of this property without due process by having a policy that tolerated nonenforcement of restraining orders.” 50Town of Castle Rock v. Gonzales, 545 U.S. 748, 755 (2005). The Supreme Court ultimately rejected her claim, holding that she did not have a property interest in police enforcement because procedural due process does not protect benefits for which a person has only “an abstract need or desire” or “a unilateral expectation.” 51 Id. at 756 (quoting Bd. of Regents of State Colls. v. Roth, 408 U.S. 564, 577 (1972)) (internal quotation mark omitted). Rather, the procedural due process doctrine protects only those benefits to which an individual has a “legitimate claim of entitlement.” 52 Id. (quoting Roth, 408 U.S. at 577). In other words, procedural due process claims are only available when the individual has been promised a tangible benefit by the state, usually in the form of monetary assistance or a more traditional interest such as real property.

Writing for the Castle Rock majority, Justice Scalia stated that an individual cannot claim that a benefit is a protected entitlement if the benefit is subject to the state’s discretion. 53 Id. He then declared that “[a] well established tradition of police discretion has long coexisted with apparently mandatory arrest statutes” 54 Id. at 760. such that Colorado’s mandatory arrest statute was not truly mandatory. 55 Id. For a state to create an entitlement by statute, the language of the statute would have to indicate more strongly that police were required to act. 56 Id. The Court concluded that the Colorado statute failed to indicate strongly enough that police were required to act to protect victims because the statute gave officers a choice of actions: they could either arrest the individual whom they had probable cause to believe violated the restraining order, or they could seek a warrant if arrest was impracticable. This choice, the majority reasoned, left room for police discretion and fell short of creating an entitlement to protection. Id. Moreover, he opined that

[i]t is hard to imagine that a Colorado peace officer would not have some discretion to determine that—despite probable cause to believe a restraining order has been violated—the circumstances of the violation or the competing duties of that officer or his agency counsel decisively against enforcement in a particular instance. 57 Id.

Castle Rock effectively forecloses any possibility of bringing a procedural due process claim for purposes of police liability in the future. Given Justice Scalia’s broad language, it will be nearly impossible to create a statute worded strongly enough to satisfy the majority’s standard. 58Curt Stokes, Note, Restraining Orders and Law Enforcement Liability After Town of Castle Rock, Colorado v. Gonzales, 26 St. Louis U. Pub. L. Rev. 99, 120–21 (2007). Furthermore, Justice Scalia asserted that even if a statute did create a recognizable entitlement, that still might not be enough to merit protection of a property interest. 59 Castle Rock, 545 U.S. at 766. The majority’s reasoning functionally eliminates the possibility of success for a future procedural due process claim, and indeed several federal courts have rejected other gender-based violence survivors’ claims based on the reasoning in Castle Rock. 60Amber Fink, Every Reasonable Means: Due Process and the (Non)enforcement of a Restraining Order in Gonzales v. Town of Castle Rock, 24 Law & Ineq. 375, 395 (2006). Many critics of the Castle Rock decision have argued that Justice Scalia’s opinion goes against extensive legislative history documenting the intent of the Colorado legislature to remove police discretion in domestic violence situations. See, e.g., Castle Rock, 545 U.S. at 779–83 (Stevens, J., dissenting). Critics and feminist scholars have also argued that Justice Scalia’s conclusion disregards the plain meaning of the word “shall” in the statute and reflects the Court’s continuing indifference to victims of gender-based violence. See, e.g., Miccio, supra note 22, at 420. After the Supreme Court ruled against her, Jessica Gonzales filed suit against the United States in the Inter-American Commission on Human Rights, alleging that the United States had violated her human rights by failing to protect her and her children from Mr. Gonzales’s actions. Gonzales Brief, supra note 2. That decision is pending; if Ms. Gonzales wins, it will be a strong critique of the Court’s historical resistance to require police officers to protect women from gender-based violence. The European Court of Human Rights recently heard a similar case, Opuz v. Turkey, in which a Turkish victim of domestic violence sued the Turkish government for its failure to protect her and her family members from violence by her husband. In a landmark decision, the court ruled for the plaintiff-victim and, for the first time, held a government accountable for failure “to take adequate steps to protect victims of repeated domestic violence, even absent any active malfeasance on the state’s part.” Tarik Abdel-Monem, Opuz v. Turkey: Europe’s Landmark Judgment on Violence Against Women , 17 Hum. Rts. Brief 29, 29 (2009). Especially interesting is that the court agreed with the plaintiff that the Turkish government violated, inter alia, the prohibition on discrimination contained in the European Convention on Human Rights. Id.

3. State Tort Remedy

Since the Supreme Court’s decision in Castle Rock, many commentators have advocated for the creation of a state tort remedy, as suggested by both the Castle Rock and DeShaney majority opinions, 61 Castle Rock, 545 U.S. at 768–69; DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 201–03 (1989). which would allow victims to sue police officers and police departments under state tort law for failure to enforce a restraining order despite a mandatory arrest statute. 62 See, e.g., Joshua J. Bennett, Case Note, Town of Castle Rock v. Gonzales and the Awful Truth: Utah’s Mandatory Enforcement Laws Make Police Enforcement of a Victim’s Protective Order Optional, 8 J.L. & Fam. Stud. 405, 422–23 (2006); Danielle Lynn Lordi, Comment, Police Liability Under State Tort Law for Failure to Enforce Protection Orders: The Last Demand for Accountability, 85 Or. L. Rev. 325, 346–49 (2006); Stefanie T. Scott, Note, Trying to Touch the Untouchables: The Challenges Faced by Texas Plaintiffs Asserting Failure-to-Protect Suits Against Police Departments, 27 Rev. Litig. 539, 557–60 (2008); Stokes, supra note 58, at 125. However, this type of reform likely will be unsuccessful at both the legislative and judicial levels. Police lobbies are influential and police departments have significant political clout, making it unrealistic to think that legislatures would enact such legislation. 63 Kristian Williams, Our Enemies in Blue: Police and Power in America 2, 139 (2004) (“Few changes in public safety or security policies can be made without the tacit approval of the police unions, and the officers’ associations are routinely consulted on changes in the criminal code, or in city policies that might indirectly affect police work.”). For example, one article written after the Castle Rock decision advocates a strict liability standard for police departments when an officer fails to arrest despite having probable cause to believe that a restraining order has been violated. 64Lordi, supra note 62, at 348. Given the high deference shown to law enforcement in Castle Rock, it is improbable that any legislature would ever impose such a strict standard on the police.

Also, the historical deference to police discretion is too entrenched for a state tort cause of action to be a plausible strategy. Even if states do enact legislation exposing police to liability, it is unlikely that the elected judges who sit on many state courts will rule in favor of plaintiffs and slap police officers or departments with large monetary penalties by awarding substantial damages. 65Miccio, supra note 22, at 428. Low damage awards will fail to provide the incentive necessary for police officers and police departments to change their practices with regard to victims of gender-based violence.

D. Making the Case for Police Accountability: Why Police Enforcement Is Necessary

There are several compelling reasons for encouraging police enforcement of protective orders. First, when enforced, restraining orders are effective at curbing violence by both decreasing the likelihood of a future attack 66Victoria L. Holt et al., Do Protection Orders Affect the Likelihood of Future Partner Violence and Injury?, 24 Am. J. Preventative Med. 16, 18–21 (2003). and deterring a rise in the severity of future attacks. 67Adele Harrell & Barbara E. Smith, Effects of Restraining Orders on Domestic Violence Victims, in Do Arrests and Restraining Orders Work? 214, 234 (Eve S. Buzawa & Carl G. Buzawa eds., 1996) (“The odds of severe violence in cases in which an arrest had been made were less than half that of cases in which no arrest had been made [in the year following the arrest].”). Second, failing to enforce a restraining order negatively affects a victim’s ability to escape her situation in a very real way. The risk of homicide by an abusive partner is highest when a victim decides to leave a relationship, 68Jacquelyn C. Campbell et al., Risk Factors for Femicide in Abusive Relationships: Results from a Multisite Case Control Study, 93 Am. J. Pub. Health 1089, 1090 (2003). Moreover, there is evidence that suggests that obtaining a restraining order itself can increase the likelihood of retaliatory violence, making its enforcement even more important to avoid endangering the victim rather than protecting her. Fink, supra note 60, at 379–80 (citing Brief of National Network to End Domestic Violence et al. as Amici Curiae Supporting Respondents at 20, Town of Castle Rock v. Gonzales, 545 U.S. 748 (2005) (No. 04-278)). and obtaining an order of protection may be an indication that a victim is preparing to leave or has left. Thus, an abused woman 69This Comment recognizes that gender-based violence occurs against both men and women, but it refers to victims as women because of the endemic and disproportionate prevalence of gender-based violence by men against women. Similarly, not all police officers are men, and the use of gendered pronouns runs the risk of incurring the same criticism that forms the premise of the argument advanced by this Comment. However, for the sake of differentiation between officers and victims, and having made this concession, this Comment will refer to police officers using the male pronoun where necessary for clarity. and her children may be most vulnerable at the time that she obtains a protective order; 70For example, the Gonzales tragedy occurred one month after Ms. Gonzales obtained a divorce and permanent restraining order against her husband. Castle Rock, 545 U.S. at 753. lack of enforcement may expose her to a brutal attack, or death. 71 See, e.g., id.; Estate of Macias v. Ihde, 219 F.3d 1018, 1026 (9th Cir. 2000) (ex-husband shot and killed victim, as well as victim’s mother and himself, after police officer told victim there was probable cause that ex-husband had violated restraining order and would be arrested); Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026 (3d Cir. 1988) (estranged boyfriend murdered victim at her place of work after stalking her and breaking into her home in the middle of the night in violation of a recently expired emergency restraining order).

Third, nonenforcement communicates to abusers that they are invulnerable, silently condoning their actions and emboldening them. This exposes the victim to further harm by a perpetrator who feels that he is above the law. For example, in a case where a woman was killed after numerous violations of a restraining order and multiple calls to the police for help over several months, 72 Ihde, 219 F.3d at 1021–25. her ex-husband reportedly used to “boast[] to friends and others in the community that the deputy sheriffs were on his side, that the Sheriff protected him and not [the victim]. [He] would torment [her] with the same gibe.” 73Estate of Macias v. Lopez, 42 F. Supp. 2d 957, 960 (N.D. Cal. 1999) (internal quotation mark omitted), rev’d sub nom. Estate of Macias v. Ihde, 219 F.3d 1018, 1026 (9th Cir. 2000). Failure to enforce restraining orders perpetuates in communities the idea that gender-based violence is tolerable and will not result in any legal consequences.

Finally, issuing a restraining order and then failing or refusing to enforce it creates a false sense of security for a victim who mistakenly relies on the promise of police protection. 74Jennifer Rios, Note, What’s the Hold-Up? Making the Case for Lifetime Orders of Protection in New York State, 12 Cardozo J.L. & Gender 709, 728–29 (2006). This may result in worse harm than the victim would have experienced had she taken better steps to protect herself, knowing that no one else would help her. For example, Jessica Gonzales explicitly stated: “Had I known that the police would do nothing to locate Rebecca, Katheryn, and Leslie or enforce my restraining order, I would have taken the situation into my own hands by looking for my children with my family and friends.” 75Jessica Gonzales, Statement Before the Inter-American Commission on Human Rights (Mar. 2, 2007), available at http://www.aclu.org/human-rights-womens-rights/jessica-gonzales-statement-iachr. Police indifference fosters distrust of law enforcement and the criminal justice system among victims. 76Press Release, Nat’l Crime Victim Bar Ass’n, National Crime Victim Bar Association Asks Supreme Court to Rule in Favor of Victims on Protective Orders (Feb. 11, 2005), available at http://www.ncvc.org/vb/AGP.Net/Components/documentViewer/Download.aspxnz?DocumentID=39534.

Thus, it is essential that the law create a credible way of holding police officers and departments accountable for nonenforcement of restraining orders in cases of gender-based violence. Without a credible basis for suit, police accountability is no more than an empty threat, and police officers will continue to disregard the safety of gender-based violence victims. 77Miccio, supra note 22, at 416. In cases such as Castle Rock, where the abuser kills himself or is otherwise killed, neither recovery nor individual accountability is possible, highlighting the need for law enforcement accountability and prevention.

II. Equal Protection: A Beleaguered Avenue for Police Accountability

The only remaining avenue for police liability is a § 1983 claim that police failure to enforce restraining orders denies victims their right to equal protection of the law. 78In DeShaney, the majority recognized that “[t]he State may not, of course, selectively deny its protective services to certain disfavored minorities without violating the Equal Protection Clause.” DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 197 n.3 (1989). The Equal Protection Clause of the Fourteenth Amendment states that “[n]o state shall . . . deny to any person within its jurisdiction the equal protection of the laws.” 79U.S. Const. amend. XIV, § 1. The purpose of this protection is to ensure that all persons who are similarly situated are afforded the same treatment. 80 Erwin Chemerinsky, Constitutional Law: Principles and Policies 676 (3d ed. 2006) (noting that equal protection principles protect individuals, not groups, and that an individual can bring an equal protection claim for discrimination in the form of arbitrary government treatment even against a “class of one” (quoting Village of Willowbrook v. Olech, 528 U.S. 562, 564 (2000)) (internal quotation marks omitted)).

This Part introduces the possibility of bringing equal protection claims against police officers who fail to enforce restraining orders. Section A reviews the history of the equal protection doctrine and current equal protection jurisprudence under the Constitution. Section B discusses past equal protection claims that have failed in the lower federal courts. In particular, this section notes the possible levels of scrutiny with which an equal protection claim in this context could be reviewed, as well as the shortcomings of the arguments advanced by past plaintiff-victims.

A. Background and Summary of Equal Protection Under the Constitution

The Supreme Court’s equal protection jurisprudence is comprised of three tiers of judicial review. Any classification used by the state to distinguish between persons, whether written into a statute or used in a more subtle way, must pass some level of review by a court because the law presumes that persons similarly situated should be treated alike. 81 Id. at 669–70. The level of review that a classification will receive depends on the basis of the classification itself. 82 Id. at 670. The Court has deemed that certain classifications, such as race and national origin, are almost never an appropriate basis on which to discriminate between persons; accordingly, these classifications are reviewed under strict scrutiny. 83 Id. at 671. For the state to classify or discriminate on the basis of race or another suspect class, the state must advance a “compelling” interest, and the state action or policy must be narrowly tailored to achieve the stated government interest. 84 Id. If the government interest is not sufficiently compelling, or if a less over- or under-inclusive policy could achieve the government interest, then the statute or classification will be struck down as an unconstitutional violation of the equal protection clause. 85 Id.

The next level of review is intermediate scrutiny. As its name suggests, this level of review is less rigorous than strict scrutiny. 86 Id. at 671–72. Gender is one classification that receives intermediate scrutiny. 87 Id. at 671. For a state actor to justify a classification or discrimination based on gender, it must advance an “important” government interest and demonstrate that the classification is “substantially related” to achieving this interest. 88Miss. Univ. for Women v. Hogan, 458 U.S. 718, 724 (1982). The rationale advanced by the Court for applying a lesser level of scrutiny in these cases is that, unlike individuals of different races or national origins, the sexes do manifest some “inherent differences.” 89United States v. Virginia (VMI), 518 U.S. 515, 533 (1996) (internal quotation marks omitted). On the other hand, these classifications must undergo some scrutiny “to assure that the validity of a classification is determined through reasoned analysis rather than through the mechanical application of traditional, often inaccurate, assumptions about the proper roles of men and women.” 90 Hogan, 458 U.S. at 725–26. Due to these competing principles, cases reviewed under the intermediate scrutiny standard have had mixed success in the federal courts. 91The Supreme Court has invalidated many gender-based classifications challenged under the Equal Protection Clause, but has upheld some. Chemerinsky, supra note 80, at 672; see also infra Part III.B.1.b.

While classifications that are overtly or explicitly based on gender will automatically receive intermediate scrutiny, intermediate scrutiny is harder to invoke when the classification is neutral on its face but has an adverse impact on one gender. In these instances, a court will only apply intermediate scrutiny if the plaintiff can demonstrate that the intent behind the policy was to discriminate against the gender that is disproportionately affected. 92 Chemerinsky, supra note 80, at 672. Otherwise, even in the face of an extremely disparate impact on one gender, a facially neutral classification will receive rational basis review. 93 Id.

All other classifications, such as age, sexual orientation, and disability, also receive the most lenient level of judicial scrutiny: rational basis review. 94 Id. For a classification to pass muster under rational basis review, the government can simply advance a “legitimate” interest and show that the classification is reasonably related to achieving that interest. 95 Id. The purpose of this basis of review is to prevent the state from making arbitrary distinctions between people merely on the basis of hostility toward a particular group. 96 Id. at 677–78. However, rational basis review provides little protection as long as the government can advance some non-arbitrary reason to justify the classification. 97 Id.

The distinction between rational basis review and intermediate scrutiny is important because the level of review largely determines the outcome of the case. The claim of a plaintiff alleging that police treated her differently because of her status as a victim of domestic violence—that the discrimination was based on the type of crime, or the type of relationship between the victim and abuser—would receive only rational basis review. However, a plaintiff-victim has a much greater chance of success if her claim is reviewed with intermediate scrutiny.

B. Past Attempts to Obtain Intermediate Scrutiny: Discrimination Against Women or Against Victims of Domestic Violence?

In the past several decades, many lower federal courts have heard cases with facts similar to Castle Rock, in which either a victim-spouse or her children were severely injured or killed by a partner. 98 See, e.g., Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026 (3d Cir. 1988); Watson v. City of Kansas City, 857 F.2d 690, 693 (10th Cir. 1988) (husband, a police officer himself, raped, beat, and stabbed victim-spouse in the presence of her children, in violation of a restraining order, before committing suicide); Thurman v. City of Torrington, 595 F. Supp. 1521, 1527–29 (D. Conn. 1984) (ex-husband repeatedly stabbed victim in the chest, neck, and throat, in violation of a restraining order, and continued assault after police arrived, kicking victim in the head multiple times, grabbing her child from inside the residence, and dropping child on wounded victim before being arrested). In Hynson v. City of Chester, Legal Department, the Third Circuit noted a “growing trend of plaintiffs relying upon the due process and equal protection clauses, enforceable through § 1983, to force police departments to provide women with the protection from domestic violence that police agencies are allegedly reluctant to give.” 99 864 F.2d at 1030. The Hynson court noted the distinction between “discrimination on the basis of . . . membership in the class of victims of domestic abuse,” which would receive rational basis review, and discrimination based on sex or gender, which would require intermediate scrutiny. 100 Id. at 1031.

While the Supreme Court has never reviewed this type of claim, 101All of these plaintiff-victims had obtained a civil protection order before their attacks. Most sued police departments on both due process and equal protection grounds. This Comment will only discuss the courts’ equal protection analyses since the Supreme Court has rejected both procedural and substantive due process theories as a basis for bringing a § 1983 action on these facts. the rationales of lower courts, as well as related Supreme Court jurisprudence, strongly suggest that the Court would deny an equal protection claim that relied on arguments similar to the ones lower courts have heard. The following subsection reviews the arguments that victims have advanced, as well as the reasoning of the various courts that have adjudicated and denied these claims.

1. Failed Attempts at Intermediate Scrutiny

Given the difficulty of successfully challenging state action under rational basis review, a key task is to convince a court to invoke intermediate scrutiny. Most plaintiff-victims have attempted to do so using some species of a disparate impact argument. They have alleged that failure to enforce the restraining order of a domestic violence victim is gender-based discrimination because women are primarily (and disproportionately) affected by domestic violence. In other words, police policies and practices 102Many of the equal protection claims discussed in this section consisted of both an action against an individual officer and an action against the police department itself for a municipal liability claim. See supra note 32 and accompanying text. A municipal liability claim differs from a claim against an individual officer because a victim can sue the police department only if the entity itself caused the constitutional violation. Monell v. Dep’t of Soc. Servs., 436 U.S. 658 (1978). Municipal liability requires the plaintiff to show that the municipality had a policy or custom that caused the violation of the plaintiff’s constitutional right. Id. at 694. However, claims of this type are difficult to prove because there must be “a direct causal link between a municipal policy or custom and the alleged constitutional deprivation.” City of Canton v. Harris, 489 U.S. 378, 391 (1989). However, municipal liability claims provide deeper pockets for damage awards, as the city itself must pay the damages, whereas a claim against an individual officer will expose only that officer to personal liability. Id. For the argument advanced by this Comment, however, additional distinctions between municipal liability and individual officer liability are ancillary and will not be discussed in detail. that provide lesser protection to domestic violence victims discriminate against women, violating their right to equal protection of the law. These past attempts to invoke intermediate scrutiny have failed for two reasons: (1) lack of data showing that police policies affecting domestic violence victims disproportionately affect women, and (2) failure to meet the strict standard for invoking intermediate scrutiny in the context of a facially neutral classification. This subsection describes in detail the reasoning with which past litigants’ claims have been rejected. 103Prior to DeShaney, one district court allowed a plaintiff to bring a § 1983 equal protection claim against the police on the basis of gender because it found that the police department’s practice of affording inadequate protection to domestic-violence-related calls operated as an administrative classification that implemented the law in a discriminatory fashion on the basis of sex. Thurman, 595 F. Supp. at 1527–29 (denying city’s motion to dismiss plaintiff’s equal protection claim). Tracey Thurman was awarded $2.3 million by a federal jury. Anne Sparks, Feminists Negotiate the Executive Branch: The Policing of Male Violence, in Feminists Negotiate the State: The Politics of Domestic Violence 35, 42 (Cynthia R. Daniels et al. eds., 1997). However, the Thurman court’s rationale is no longer good law. The district court based its decision on a finding that the police had an affirmative duty to protect the plaintiff-victim because they were aware of her situation. Thurman, 595 F. Supp. at 1527. Four years later, in DeShaney, the Supreme Court rejected this notion, stating that the police never have an affirmative duty to protect an individual unless the individual is in state custody. It explicitly rejected the argument that state knowledge of the risk or danger faced by the victim is sufficient to create a special relationship that would trigger an affirmative duty to protect. DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 197–200 (1989).

a. Lack of Data to Demonstrate Discriminatory Policies

The first roadblock faced by plaintiff-victims is a lack of data to prove the existence of a discriminatory policy. Because their claims are premised on the existence of a policy that discriminates against all domestic violence survivors, rather than just the treatment afforded to a particular plaintiff, this data is a necessary element.

In Hynson, the Third Circuit created a three-factor test to determine whether a plaintiff could make a prima facie case for an equal protection violation on the basis of gender. 104The court stated that its rationale for this three-part standard was “to eliminate the all too common shotgun pleading approach to these equal protection claims.” Hynson, 864 F.2d at 1031 n.13. After laying out this standard, the appeals court remanded the case to the district court to apply the test. On remand, the district court in Hynson found sufficient evidence for a jury to infer an intent to discriminate against women, but still granted summary judgment for the police officers on the basis of qualified immunity. Hynson v. City of Chester, 731 F. Supp. 1236, 1240–41 (E.D. Pa. 1990). In its decision, the district court did not give any indication as to how it would have decided had this immunity not been available. The first factor of this test requires a plaintiff to provide sufficient evidence from which a reasonable jury could infer that the police had a policy or custom to provide less protection to domestic violence victims than to victims of other types of violence. 105 Hynson, 864 F.2d at 1031. The second and third factors of this test require that discrimination against women was a motivating factor in this policy and that the plaintiff was injured by this policy or custom. Id. See infra Part II.B.1.b. for discussion of the second factor. Many circuit courts have adopted the Hynson test or one that is analogous. See, e.g., Soto v. Flores, 103 F.3d 1056 (1st Cir. 1997); Navarro v. Block, 72 F.3d 712 (9th Cir. 1995); Eagleston v. Guido, 41 F.3d 865 (2d Cir. 1994); Ricketts v. City of Columbia, 36 F.3d 775 (8th Cir. 1994); McKee v. City of Rockwall, 877 F.2d 409 (5th Cir. 1989). This factor presents a significant hurdle for individual litigants. Because most police departments do not collect data about their response rates to reported problems, and individual victims do not have the resources or ability to conduct significant research in preparation for their litigation, many plaintiff-victims’ arguments will be foreclosed at this step. 106 See, e.g., Scott, supra note 62, at 559–62.

For instance, in 2007, the Third Circuit rejected Jill Burella’s equal protection claim in which she alleged discrimination on the basis of gender. 107Burella v. City of Phila., 501 F.3d 134 (3d Cir. 2007). In addition to her equal protection claim, Jill Burella also brought § 1983 challenges on the basis of substantive and procedural due process. The Third Circuit denied both due process claims based on the precedents in DeShaney and Castle Rock. Burella alleged that the Philadelphia Police Department had a custom or policy of providing domestic violence victims with less protection than other victims of violence. 108 Id. at 148. She argued the court should infer discriminatory intent on the basis of gender. 109 Id. As evidence to support this inference, she provided a police sergeant’s statement that victims of domestic violence were predominantly women and an expert report that stated that the Philadelphia Police Department discriminated against female victims of domestic violence. 110 Id. at 148–49.

The Third Circuit declined to infer a custom of inferior protection to domestic violence victims and stated that,

[w]hile statistical evidence and individual arrest records are not per se requirements in this context, such evidence may often be crucial. Indeed, in this case there is a marked absence of any comparable evidence (or even factual allegations) from which a reasonable jury could find an unlawful custom or infer a discriminatory motive. 111 Id. at 149.

A concurring judge observed that “[s]tate protection-from-abuse statutes seem to reside in a rock-solid castle of narrow construction barring any federal constitutional relief for the very victims that the statutes are designed to protect.” 112 Id. at 150 (Ambro, J., concurring in part).

Based on the court’s reasoning and its application of existing precedent, statistical data regarding the differential response to women affected by domestic violence is a necessary prerequisite to succeeding on an equal protection claim that alleges discrimination on the basis of gender. Absent this type of data, there is no way to allege a sound gender-based equal protection argument.

b. Feeney’s Overly Stringent Standard: Showing a Discriminatory Purpose

Further complicating the plaintiff’s burden is that even if she does have data demonstrating that women are disproportionately affected by a police practice relegating domestic violence calls to low-priority status, disparate impact alone is not enough to constitute an equal protection violation when the basis of the classification is not actually gender. In Personnel Administrator v. Feeney, the Supreme Court held that a plaintiff challenging a facially gender-neutral law but asserting a gender-based discrimination claim had to show that the classification was adopted to purposefully discriminate against women. 113442 U.S. 256, 279 (1979). This standard demands that the plaintiff show more than just that the policy had been adopted “in spite of” its adverse effects on women. 114The Court effectively stated that a showing of disparate impact alone, no matter how extreme, was not an actionable violation of equal protection. Id. Rather, the plaintiff must show that the policy was adopted “because of” its adverse effects on women. 115 Id.

The second factor from the Hynson court’s three-factor test, that the victim show “a discriminatory purpose [against women] behind the policy,” 116Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026, 1031 (3d Cir. 1988). was adopted from Feeney. Based on current Supreme Court jurisprudence, a gender-based violence victim would have to demonstrate that police officers or a police department did not respond to her call because they had a policy to provide women—not just domestic violence victims—with an inferior standard of protection. This standard is nearly impossible to meet because the vast majority of formal inequality has been erased from the books, and discrimination now exists in much more subtle forms and at the level of individual officer discretion rather than department-wide policy.

In Watson v. City of Kansas City, a victim-plaintiff sued the local police department claiming discrimination against abused spouses as a matter of training or unwritten policy or custom. 117857 F.2d 690, 694 (10th Cir. 1988). The victim was the spouse of a police officer in the Kansas City police department, and she had statistics demonstrating the dissimilar arrest rates for domestic and nondomestic disputes. 118 Id. at 695. But the Tenth Circuit rejected her argument that this disparity was sex discrimination, citing Feeney for the proposition that a disparate impact on women is not enough. 119 Id. at 696–97 (citations omitted). More recently, in Okin v. Village of Cornwall-on-Hudson Police Department, the Second Circuit similarly held that “[p]roof that discriminatory intent [on the basis of sex] was a motivating factor is required to show a violation of the Equal Protection Clause.” 120577 F.3d 415, 438 (2d Cir. 2009).

Unable to meet the exceptionally high standard required to make a cognizable gender-based equal protection claim, 121Prior to Castle Rock but after DeShaney, the Ninth Circuit heard a case in which the district court granted summary judgment to police officers against a plaintiff who alleged violation of her right to equal protection on the basis of both domestic-violence-victim status and gender. Estate of Macias v. Ihde, 219 F.3d 1018, 1028–29 (9th Cir. 2000). On appeal, the Ninth Circuit reversed the grant of summary judgment and remanded to the district court. Id. The police department then settled with the victim’s family for one million dollars. Lordi, supra note 62, at 336 n.71 (citing Sonoma County, California, Sheriff’s Department Settles Domestic Violence Murder Case for One Million Dollars, 8 Domestic Violence Rep. 1, 15 (2002)). This case, while encouraging, still does not provide a model for a successful equal protection challenge because it did not reach trial. victims are thus left with the difficult task of establishing that a classification on the basis of domestic, as opposed to stranger, violence fails rational basis review.

2. Rational Basis Review: An Impossible Standard to Meet

Several courts have heard cases with facts similar to Castle Rock, in which the plaintiff-victim alleged discrimination on the basis of her status as a victim of violence by a domestic partner as opposed to violence by a stranger. 122 See, e.g., Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026, 1026 (3d Cir. 1988); Watson, 857 F.2d at 690. In each of these cases, the plaintiff had to establish statistical or empirical evidence indicating that police attitude and response to domestic violence victims was dissimilar to police behavior when faced with a request for help from a victim of stranger assault. 123 See, e.g., Watson, 857 F.2d at 696. The cases suggest that while many of these plaintiffs were able to survive summary judgment, this argument is unlikely to succeed on the merits because of the lack of data available to support empirical claims. 124 See, e.g., Scott, supra note 62, at 559–62.

Furthermore, the rational basis test itself makes it extremely difficult for a plaintiff to prevail. Under rational basis review, the government simply has to show that the classification reasonably relates to a legitimate government interest. 125 Chemerinsky, supra note 80, at 672. This standard is “enormously deferential” to the government, and plaintiffs rarely prevail under this level of review. 126 Id. Moreover, unlike intermediate scrutiny, under which the government has the burden to prove that its interest is important and the classification is substantially related, under rational basis review, the burden is on the plaintiff to show the classification fails the rational basis test. 127 Id. Finally, under the rational basis test, a court will accept any legitimate purpose the government can come up with, even if it is simply pretext for another discriminatory purpose; in intermediate scrutiny, a court will evaluate whether the proffered justification is the actual reason the government adopted the classification. If it is not, the court will reject the classification as unconstitutional. 128 Id. at 683. In most cases, police will be able to come up with some rational reason to justify discrimination between victims of domestic violence and those of other forms of violence. The hurdles inherent in rational basis review make it highly unlikely that a plaintiff-victim would prevail in proving the unconstitutionality of actions that are alleged to differentiate between domestic and nondomestic violence victims.

In the discussion above, the ultimate problem for each plaintiff-victim was that her claim required the court to infer that a particular victim’s experience was representative of other women’s experiences. Each individual claim was unable to stand alone on its own facts. 129In Okin, the Second Circuit noted the district court’s finding that the plaintiff “relie[d] exclusively on her own story—which, while heartrending, does not get her where [the test] says she must go.” Okin v. Cornwall-on-Hudson Police Dep’t, 577 F.3d 415, 438 (2d Cir. 2009) (quoting Okin v. Cornwall-on-Hudson Police Dep’t, No. 04 Civ. 3679 (CM), 2006 U.S. Dist. LEXIS 75881, at *58 (S.D.N.Y. Oct. 13, 2006)) (internal quotation mark omitted). This Comment addresses this problem by developing an equal protection claim that challenges an individual police officer’s actions as they relate to a particular victim. The argument presented here will allow a court to evaluate a plaintiff’s claim based on its own individual facts and still invoke intermediate scrutiny.

III. Reformulating the Claim: Reviving Intermediate Scrutiny for Victims of Gender-Based Violence

The dearth of viable litigation strategies for victims of gender-based violence makes clear the need for a new framework for arguing equal protection claims. This Comment proposes a claim that reshapes the way in which litigants conceive of and present the elements of an equal protection claim to create a successful federal constitutional claim for police liability.

To accomplish this, the first task is to identify the government action; the next step is to frame the classification it makes as one based on gender. The final step is to apply the intermediate scrutiny test and evaluate whether the government action is substantially related to an important government interest. The following sections take up each of these steps in turn.

A. The Government Action and Triggering Intermediate Scrutiny: Gender-Based Stereotypes Instead of Disparate Impact

The task in this section is to develop an equal protection argument that invokes gender, and thus intermediate scrutiny, without relying on a disparate impact argument. Subsection 1 pinpoints the government action in question here: the exercise of discretion by a police officer at the moment when he must decide whether to respond to a victim of violence. Subsection 2 then details how to frame that action as one based on gender: by identifying an officer’s inappropriate use of gendered stereotypes in deciding not to respond to a victim.

1. The Government Action

To begin formulating a new equal protection claim, it is necessary first to identify the government action that allegedly violates the Equal Protection Clause. One way to conceptualize the government action is the exercise of discretion by a police officer or department at the moment a victim calls claiming a restraining order has been violated. But this conceptualization puts us right back at Castle Rock. 130 See supra text accompanying notes 50–60. Thus, it is necessary to concede that a police officer’s ability to use discretion is almost certain to pass constitutional muster, even under the intermediate scrutiny standard.

But the question of how the police officer or department uses that discretion is a different matter. In a case like that of Jessica Gonzales, several state actors exercise discretion in a way that ultimately results in the harm suffered. A victim’s contact with the police begins with a 911 call, answered by a dispatcher. As soon as the dispatcher hears the situation, she must decide how to respond to the phone call. She could dispatch an officer to the scene immediately, ignore the call entirely, or do something in between, such as make a note in a file and ask the victim to wait and call back if the children do not return within a few hours. When the dispatcher decides which action to take in response to the allegation that the abuser has violated his restraining order, the dispatcher is exercising discretion. Next, if the dispatcher does send an officer to assist the victim, the action that officer takes once he arrives and sees that the children are missing also involves an exercise of discretion. He could immediately begin to search for them, determine that no such action would be worthwhile and return to the police station without assisting the victim, or do something in between, such as check the registry to see if the restraining order was valid or make a note in a file.

The key inquiry for the equal protection claim developed by this Comment is: What motivated each officer’s decision? On what basis did he exercise his discretion? While an officer is permitted to exercise his discretion when undertaking a police duty, he must do so in a constitutional way—one that does not violate the victim’s right to equal protection. That is, the basis on which the police officer decides how to respond to a victim may not take into account or treat her differently based on a protected status. Thus, if the victim’s sex influences the decision the officer makes, in his discretion, about what action to take, then there is an equal protection claim. 131There is a key distinction between this discussion of discretion and the role of discretion in the procedural due process line of cases, as discussed by Justice Scalia in the Castle Rock opinion. In that case, in order to show that Ms. Gonzales’s procedural due process rights were violated, it was necessary to show that she was entitled to enforcement of her restraining order, and the prerequisite to entitlement was a lack of police discretion. Therefore, she could not make a successful claim unless the mandatory arrest statute eliminated any and all police discretion, an argument that Justice Scalia rejected. For this type of equal protection challenge, in which only an officer’s use of gender stereotyping is at issue, there is no need to eliminate discretion entirely; the only requirement is that the discretion be exercised for an appropriate reason, and not on the basis of an impermissible, unconstitutional stereotype. This claim does not invalidate discretion; it simply challenges how the police use their discretion. Therefore, this strategy is more moderate than both a procedural due process-based claim and a state tort remedy, each of which is premised on violation of a mandatory arrest statute. See supra text accompanying notes 50–59 and 61–62. In fact, this equal protection claim does not require states to have enacted mandatory arrest statutes at all, because it does not question officers’ actual ability to exercise their discretion—it simply requires them to do so on an appropriate basis. Whether the officer’s action actually violates the Equal Protection Clause, however, will depend on the basis of the classification, the government interest the officer advances by exercising this discretion, and the relationship between the action he takes and the government interest he advances. The following subsection describes the classification in this instance: gender, as understood in the context of gender-based stereotypes. It begins with a discussion of relevant Supreme Court jurisprudence in the area of gender-based stereotypes, followed by a development of the argument that it is unconstitutional for police officers to use gendered stereotypes in responding to gender-based violence survivors.

2. Gender-Based Stereotypes in Discretionary Police Decisions

This Comment’s equal protection claim takes issue with a specific action taken by a particular police officer in response to a call from an individual woman. Having identified that specific action, it challenges the basis on which that officer made that decision. The key question is whether gender, in the form of a gender-based stereotype, influenced the decision that officer made in response to that victim’s request for help.

The invocation of gender in this way avoids the disparate impact problem faced by previous gender-based equal protection claims because it narrows the scope of scrutinized conduct to individual instances of mistreatment of a victim by a police officer. This conception of the claim relies on framing the relevant inquiry as whether an officer’s conduct was motivated by the gender of this individual victim—and thus motivated by gender per se—rather than attempting to impute a classification based on gender via a disparate impact argument. Thus, the plaintiff can directly assert that the classification is gender and therefore have her claim reviewed with intermediate scrutiny. This strategy relies on challenging the constitutionality of state use of gender-based stereotypes and invokes the Court’s jurisprudence rejecting outdated but invidious stereotypes about women.

a. Gender-Based Stereotypes and the Law

Gender stereotypes are “archaic and overbroad generalizations about gender.” 132J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 135 (1994) (quoting Schlesinger v. Ballard, 419 U.S. 498, 508 (1975)). It is firmly established within equal protection jurisprudence that state actors may not employ gender-based stereotypes in making decisions about how to treat men and women. In J. E. B. v. Alabama ex rel. T. B., the Court held as “axiomatic” that “discrimination on the basis of gender by state actors violates the Equal Protection Clause, particularly where . . . the discrimination serves to ratify and perpetuate invidious, ‘archaic, and overbroad’ stereotypes about . . . men and women.” 133 Id. at 130–31. In Mississippi University for Women v. Hogan, the Court stated that “[a]lthough the test for determining the validity of a gender-based classification is straightforward, it must be applied free of fixed notions concerning the roles and abilities of males and females.” 458 U.S. 718, 724–25 (1982). Through its decisions in cases such as J. E. B., 134511 U.S. at 139–40. Mississippi University for Women v. Hogan, 135458 U.S. at 725, 730. and United States v. Virginia (VMI), 136518 U.S. 515, 541–42 (1996). the Court has made clear that gender-based stereotypes, which make assumptions about members of either gender, will violate the Equal Protection Clause. The majority in Hogan emphasized that states may not “rel[y] upon the simplistic, outdated assumption that gender [can] be used as a ‘proxy for other, more germane bases of classification.’” 137458 U.S. at 726 (quoting Craig v. Boren, 429 U.S. 190, 198 (1976)).

In VMI, the Virginia Military Institute (VMI) defended its policy of denying admission to all women by arguing that its educational mission depended on the use of the adversative method of educating students, and that this method could not accommodate women. 138518 U.S. at 540. VMI defended its position through expert testimony stating that men and women have different “tendencies” and “gender-based developmental differences.” 139 Id. at 541 (quoting United States v. Virginia, 766 F. Supp. 1407, 1434 (W.D. Va. 1991)). One expert for VMI stated that “‘[m]ales tend to need an atmosphere of adversativeness,’ while ‘[f]emales tend to thrive in a cooperative atmosphere.’” 140 Id. (alterations in original) (quoting Virginia, 766 F. Supp. at 1434) (internal quotation marks omitted). However, the expert—and VMI—conceded that some women were capable of meeting the requirements of a typical male student at VMI and indeed would do well under the adversative model. 141 Id.

The Supreme Court rejected VMI’s arguments and held that this use of gender stereotypes—to make assumptions about the particular characteristics of an individual female candidate—was a classification based on gender that lacked the requisite exceedingly persuasive justification and was therefore an unconstitutional violation of equal protection. 142 Id. at 540. The majority stated that “equal protection principles, as applied to gender classifications, mean state actors may not rely on ‘overbroad’ generalizations to make ‘judgments about people that are likely to . . . perpetuate historical patterns of discrimination.’” 143 Id. at 541–42 (alteration in original) (quoting and citing J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 139 n.11 (1994)). Moreover, the Court emphasized that “generalizations about ‘the way women are,’ estimates of what is appropriate for most women, no longer justify” dissimilar treatment of men and women. 144 Id. at 550. Thus, in VMI, the Court rejected VMI’s use of gender stereotypes to generalize about the nature and tendencies of all women and disapproved the use of those stereotypes to impute characteristics to individual women. 145Two other important cases in the Court’s gender-based stereotyping jurisprudence, mentioned above, are Mississippi University for Women v. Hogan and J. E. B. ex rel. T. B. v. Alabama. In Hogan, a state-run university restricted admission to its nursing program to women. 458 U.S. 718, 720 (1982). A male applicant challenged the policy under equal protection law. Id. at 721. Mississippi argued that the admission policy was an instance of educational affirmative action. Id. at 727. The Court rejected this argument, holding, inter alia, that the State “failed to establish that the alleged objective is the actual purpose underlying the discriminatory classification.” Id. at 730. Instead, the majority stated that the policy of excluding males from the nursing program “tends to perpetuate the stereotyped view of nursing as an exclusively woman’s job.” Id. at 729. In J. E. B., the defendant in a paternity suit challenged on equal protection grounds the government’s use of its peremptory strikes to empanel an all-female jury. 511 U.S. 127, 129 (1994). He argued that the exclusion of men from the jury on the basis of their gender violated equal protection principles. Id. The government defended its use of gender-based peremptory strikes by arguing that men and women have different attitudes and that women would be more sympathetic to a paternity suit and more likely to find in favor of paternity. Id. at 137–38. The government contended that its interest was establishing the paternity of a child born out of wedlock, which justified the use of the gender-based discrimination. Id. at 137 n.8. The Court rejected both the interest advanced by the government and the proffered justification. Id. at 137–38. Discussing the actual interest of the state, the Court stated, “What respondent fails to recognize is that the only legitimate interest it could possibly have in the exercise of its peremptory challenges is securing a fair and impartial jury.” Id. at 137 n.8. Holding that the classification did not substantially relate to advancing the state’s interest under intermediate scrutiny, the Court stated it would “not accept as a defense to gender-based peremptory challenges ‘the very stereotype the law condemns.’” Id. at 138 (quoting Powers v. Ohio, 499 U.S. 400, 410 (1991)). The Court further noted:Respondent offers virtually no support for the conclusion that gender alone is an accurate predictor of juror’s attitudes; yet it urges this Court to condone the same stereotypes that justified the wholesale exclusion of women from juries and the ballot box. . . . We have made abundantly clear in past cases that gender classifications that rest on impermissible stereotypes violate the Equal Protection Clause.Id. at 138–39, 139 n.11. These cases, as well as VMI, demonstrate the Court’s commitment to eradicating the use of gendered stereotypes under equal protection jurisprudence.

With this jurisprudence in mind, the following part develops a claim based on the VMI Court’s principles and reasoning in the context of a police officer’s failure to respond to a gender-based violence victim’s request for the enforcement of her restraining order. The argument uncovers the use of gendered stereotypes by police officers in their responses to past plaintiff-victims. While many stereotypes may arise in the litigation of a particular case, the discussion here is limited to two examples that serve to illustrate the argument: women should be submissive to their husbands, and women are hysterical.

b. Gender-Based Stereotypes and Police Discretion in Domestic Violence Arrests

A compelling gender-stereotyping claim can be made in the context of a gender-based violence victim whose pleas for help are ignored by the police. The focus of this claim is on the discretion exercised by police officers when they respond to an individual victim’s call. In exercising their discretion about whether and how to respond, officers must base their decisions on the facts of the particular case, or another legitimate and appropriate basis. The argument, stated simply, is that police officers may not make inferences about the specifics of an individual woman’s case based on gender stereotypes.

For example, when Jena Balistreri called the police because her estranged husband, against whom she had a restraining order, was attacking her, the officer who arrived at the scene declined to arrest Ms. Balistreri’s husband because the officer “did not blame plaintiff’s husband for hitting her, because of the way she was ‘carrying on.’” 146Balistreri v. Pacifica Police Dep’t, 855 F.2d 1421, 1427 (9th Cir. 1988), superseded by 901 F.2d 969 (9th Cir. 1990). The decision not to arrest Mr. Balistreri was a discretional one because the officer had probable cause to arrest but chose not to do so. 147The officer’s own statement shows that he believed Ms. Balistreri when she told him that she had been attacked by her husband. Furthermore, the facts of the case provide sufficient evidence that Mr. Balistreri was in violation of his restraining order. Id. at 1423, 1426. The reason he did not arrest Mr. Balistreri was that he felt Ms. Balistreri deserved the violence she suffered because of her own actions.

This line of reasoning relies on the first gender-based stereotype identified by this Comment: she deserved it because she should have been more submissive. The officer depends on stereotypes about women for his notions of the characteristics that women should have—they should be silent, they should not complain, or they should be submissive to their husbands. In addition, the officer invokes a gendered stereotype about what Ms. Balistreri actually did to “deserve” the beating—the officer assumes, based on a stereotype, that women who are victims of partner abuse do something to provoke or incite the abuse. This officer ignored the reality of Ms. Balistreri’s situation—that she was injured by her ex-husband on several occasions, in violation of a court-issued restraining order. 148 Id. at 1423. His decision not to arrest Mr. Balistreri was a discretionary one motivated by his assumptions about women, understood through his gendered stereotypes. His conduct violated Ms. Balistreri’s right to equal protection of the law, free from discriminatory governmental action in enforcement of the law. 149 Id. at 1427.

A second gender stereotype that would be improper, if employed by a police officer in his discretionary decisions, is the stereotype of the hysterical woman. For example, after she realized her children were missing, Jessica Gonzales called the police several times. 150Gonzales Brief, supra note 2, at 10–15. The second time she called, two officers were sent to her house, but they said they could not help her because the children were with their father, despite the fact that she showed them her restraining order, which protected Ms. Gonzales as well as her three daughters. 151 Id. at 12. When she called again after speaking with Mr. Gonzales on the phone and discovering his whereabouts, she again requested that the police take steps to recover her children. 152Town of Castle Rock v. Gonzales, 545 U.S. 748, 753 (2005). Upon hearing that the children were in fact with their father, at an amusement park two counties away,

Dispatcher Cindy Dieck entered into the computer that Jessica Gonzales’ children “had been found,” even though Jessica Gonzales had specifically informed Ms. Dieck about her restraining order against Mr. Gonzales. . . . In fact, when Jessica Gonzales called again at around 10 p.m. to state that Mr. Gonzales had still not returned with her children . . . . Ms. Dieck . . . went on to scold Jessica Gonzales, stating that it was “a little ridiculous making us freak out and thinking the kids are gone.” 153Gonzales Brief, supra note 2, at 24 (footnotes omitted).

Here, something other than a thorough understanding of the situation prompted the dispatcher’s belief that Ms. Gonzales’s concern was an overreaction and a “ridiculous” reason to report the children as missing. Given Mr. Gonzales’s history of erratic behavior and multiple run-ins with the police in the recent past, the dispatcher could not have exercised her discretion based on the facts at hand; rather, her reaction must have been based on some other assumption about Ms. Gonzales or her situation. 154 Id. at 9.

Ms. Gonzales’s persistence in trying to persuade the police to enforce her restraining order resulted in a response from the dispatcher that reflected a gendered stereotype: women are hysterical. Despite information that Mr. Gonzales had removed the children from their home without their mother’s knowledge or permission, had taken them to an amusement park two counties away in violation of a restraining order, and had not returned them by 10:00 p.m., the dispatcher still determined that Ms. Gonzales’s situation did not qualify as an emergency. The only way the dispatcher could have reached that conclusion is if she thought that Ms. Gonzales was exaggerating, overreacting, or otherwise misrepresenting the situation. Because the dispatcher either did not have full knowledge about the Gonzales’ situation, or chose to ignore it, and because the dispatcher did not determine that Ms. Gonzales’s request was a non-emergency in light of another, more urgent situation to which her attention was necessary, 155See supra note 11 for an instance in which the officer had neither a factual basis for ignoring Ms. Gonzales’s request, nor another task to which he needed to attend; he simply declined to assist her and went to dinner instead. she must have made this decision based on some assumption she made about Ms. Gonzales. Here, that assumption was that Ms. Gonzales was overreacting: despite Ms. Gonzales’s frightened and urgent requests, the dispatcher decided that her pleas were not so serious as to merit emergency status. This decision indicates that the dispatcher felt Ms. Gonzales was not accurate in her own determination of the danger at hand. This assumption—that Ms. Gonzales was irrational and hysterical—resulted in the dispatcher’s discretionary decision not to respond to Ms. Gonzales’s request with any action.

The preceding examples demonstrate instances in which police officers relied on gendered stereotypes to make discretionary decisions. These are simply two illustrations; there are many other gendered stereotypes that are invoked frequently in this context. 156 See, e.g., Emily C. Rutledge, Comment, Healing Jurisprudential “Bruises”: A Critique of the Failure of Due Process to Account for the Intersection of Race, Class, and Gender in the Post-Castle Rock Era, 53 How. L.J. 421, 434–35 (2010) (discussing a variety of stereotypes that police hold about battered women generally, and poor, minority victims in particular). Each time a woman calls about her husband or boyfriend violating his restraining order and a police officer declines to respond to her requests because he thinks the victim is lying, calling repeatedly to get attention, or complaining to the police because she is vindictive, the officer has engaged in gender stereotyping. Each of these thought processes, when it results in a decision not to respond or not to arrest, is an inappropriate exercise of discretion. Unless the officer had another reason for deciding against assisting the victim, his inappropriate basis for exercising discretion is actionable under the Equal Protection Clause. 157It is important to note here that gendered stereotypes are not appropriate even if they are true of most women. See supra note 141; J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 139 n.11 (1994) (“We have made abundantly clear in past cases that gender classifications that rest on impermissible stereotypes violate the Equal Protection Clause, even when some statistical support can be conjured up for the generalization.”). Even if many women are hysterical, or do exaggerate the dangerousness of their situations, and even if the officer has personally experienced this, it would still be unconstitutional for that stereotype to influence the treatment given to an individual woman. Obviously, there are examples of women who do not fit the stereotype, such as Jessica Gonzales, whose fear and panic were well-founded. Treating her based on the stereotype both was unconstitutional and resulted in horrific but avoidable consequences. Thus, if police decisions and actions are motivated by assumptions that rely on gender stereotypes, then they will have violated the constitutional rights of the victim.

Despite the many examples of inappropriate exercises of police discretion discussed above, it is important to reiterate that this argument does not advocate an elimination of police discretion entirely. Unlike mandatory arrest statutes, which attempt to force a reluctant police officer to arrest—in every instance—an abuser in violation of a restraining order, this equal protection argument leaves room for officers to exercise their discretion—as long as they do so in an appropriate manner. Indeed, the litigation theory presented by this Comment does not rely on mandatory arrest provisions at all; even if an officer is not required to arrest someone who violated a restraining order, he must not decide whether to do so based on gendered stereotypes. 158By promoting police responsiveness to gender-based violence victims without advocating for, or relying on, mandatory arrest provisions, this argument avoids the many feminist criticisms of mandatory arrest laws. See, e.g., Leigh Goodmark, Autonomy Feminism: An Anti-Essentialist Critique of Mandatory Interventions in Domestic Violence Cases, 37 Fla. St. U. L. Rev. 1 (2009); G. Kristian Miccio, A House Divided: Mandatory Arrest, Domestic Violence, and the Conservatization of the Battered Women’s Movement, 42 Hous. L. Rev. 237 (2005).

This argument recognizes that there are plenty of well-trained, careful police officers who strive to do the right thing and whose decisions are not motivated by outdated, gendered stereotypes. These officers need not be afraid of frivolous claims against them. Discretion is fine, as long as it is not motivated by gendered stereotypes that reflect an unwillingness to take women seriously and a lack of concern for their safety, which ultimately lead to brutal and tragic consequences.

B. The Government Interest: Applying the Intermediate Scrutiny Test

As discussed above, classifications made based on gender are subject to intermediate scrutiny and are constitutional only if use of the classification substantially relates to achieving an important government interest. For the argument advanced by this Comment, a key determinant of the plaintiff’s likelihood of success will be the way in which a court characterizes the government interest at stake. 159 See, e.g., Laurence H. Tribe, Lawrence v. Texas: The “Fundamental Right” That Dare Not Speak Its Name, 117 Harv. L. Rev. 1893, 1900 (2004) (noting that in Bowers v. Hardwick, which involved a statute banning any act of sodomy, “the Supreme Court went out of its way to recast the plaintiff’s claim to substantive protection under the Due Process Clause for [the plaintiff’s] private sexual relationships as an asserted ‘fundamental right to engage in homosexual sodomy’” (quoting Bowers, 478 U.S. 186, 191 (1986))); supra note 141 (describing the J. E. B. Court’s rejection of the government’s argument for the use of peremptory strikes based on gender and the majority’s subsequent reinterpretation of the actual interest at stake). Moreover, the breadth of the characterization of the government interest will affect both whether it is considered “important” and how related the classification is to achieving the interest. If a court construes the government interest as the narrow ability of police officers to exercise their discretion on whatever basis they wish, then a plaintiff will likely be able to win her claim. If, on the other hand, the court characterizes the government interest so broadly that it considers at stake the ability of the police to do their job without a lawsuit at every step, then the victim will likely lose.

For these reasons, it is important to understand the various ways in which a defendant, and a court, could characterize the relevant government interest, as well as the effect each characterization would have on the subsequent equal protection analysis. The next part of this Comment considers several ways the government interest could be framed and asks two questions: is the government interest important, and is the gender-based classification substantially related to that interest? 160In addition to demonstrating that the classification substantially relates to achieving the government interest, the government must also show that the interest advanced during litigation is the actual purpose behind the classification, not a post hoc justification. United States v. Virginia (VMI), 518 U.S. 515, 535–36 (1996) (“[A] tenable justification must describe actual state purposes, not rationalizations for actions in fact differently grounded.”).

1. Interpretation Number One: The Interest in Discrimination

At its narrowest, the government interest could be described as the ability of the police to discriminate against individuals based on gender. In other words, the only government interest at stake would be the police’s ability to provide inferior protection to gender-based violence victims with impunity. Viewed in this way, the state interest is certainly not important—and likely not even legitimate—so it would fail intermediate scrutiny and even rational basis review. This is arguably the correct way to view the government interest, because the only action proscribed by the equal protection claim is gender stereotyping—all the officers must do to avoid liability is premise their discretion on a constitutional reason rather than an unconstitutional stereotype about women. In other words, the plaintiff’s argument here is that gender stereotyping can serve no other purpose than discriminating against women. However, it is unlikely that the Court would accept such a narrow characterization of the interest at stake, and defendant police officers and police departments will advance arguments that the interest at stake is broader and more significant.

2. Interpretation Number Two: The Interest in Allocation of Time and Resources

A second, more moderate way of construing the government interest at stake here is as the ability of the police to exercise their discretion in terms of how to allocate limited time, staff, and resources. As emphasized by Justice Scalia in the majority opinion in Castle Rock, the Court recognizes police discretion as a historically grounded—and likely important—interest. 161 See supra text accompanying notes 53–57; see also supra note 131. Thus, if the reviewing court frames the government interest in this way, the key question will become whether the use of gender stereotypes substantially relates to this interest.

Even if the government interest is construed in this way, a police officer can still make discretionary decisions without violating the Equal Protection Clause. For example, if a police officer chose not to respond to a violence victim’s call because another, more urgent matter prevented the officer from responding at that time, then the plaintiff will not have an actionable claim. The police officer’s actions did not violate equal protection because the basis on which he acted was not a gendered stereotype, and therefore not problematic. If he chose not to respond to a victim’s call because he needed to respond to another situation, he acted on the basis of other facts that created a more pressing need. In this situation, he likely did not invoke any gendered stereotype at all, and at the very least, the stereotypes did not result in treatment that violated the principles of equal protection because they were not the basis on which the officer exercised his discretion. 162A simple failure to respond or protect, without any allegation of dissimilar treatment of similarly situated persons, does not violate equal protection. This is because individual citizens hold no absolute right to police protection. See supra text accompanying notes 33 and 43–44. Equal protection simply ensures a citizen’s right to be protected equally to others that are similarly situated; that is, police may decline to protect everyone, or they may provide poor responses across the board. In this situation, there is no allegation of inappropriate classifications on the basis of a suspect or quasi-suspect class and, thus, no equal protection claim. Alternatively, if he chose not to respond because he had specific facts or knowledge that the victim was not in danger, or because he thought another officer was already responding to the situation, here again the officer would not be exposed to liability because his exercise of discretion was based on something other than a stereotype.

In contrast, police discretion exercised on the basis of gendered stereotypes will fail the balancing test because the use of stereotypes is not substantially related to preserving police resources. In other words, gender stereotyping will never substantially relate to the interest of effectively distributing limited resources because effective distribution of resources requires knowledge of facts that are specific to the situation at hand, rather than reliance on stereotypes about women generally. Since the Supreme Court has rejected the use of gendered stereotypes as an unacceptable justification for institutional decisions, it is likely that a court would find that the use of stereotypes in this context is not substantially related to preserving police resources.

3. Interpretation Number Three: The Broadest Government Interest

However, if sued, police officers will likely argue that the government interest at stake is much broader and more serious than a simple question of discretion and resources. They will claim it goes to the ability of the police to do their jobs and therefore define the government interest at stake as the freedom to make necessary discretionary decisions without constant exposure to liability. 163 Brief for United States as Amicus Curiae Supporting Petitioner at 11–12, Town of Castle Rock v. Gonzales, 545 U.S. 748, 751 (2005) (No. 04-278).

To counter this policy argument, the plaintiff should contend that while discretion is an important government interest, the equal protection claim developed here would affect police discretion and exposure to liability in a very narrow and minor way. First, discretion would be limited only to the extent that an officer employs gender stereotypes. Given the extent to which this claim hinges on police use of gender stereotypes—and nothing else related to discretion—only those officers who employed gendered stereotypes would risk exposure to liability. The plaintiff would still have to convince the court that this was indeed the basis on which the officer decided to act. Thus, only those plaintiffs with some evidence that an officer treated her dissimilarly on the basis of a gender stereotype would have a credible claim. Furthermore, the additional exposure to liability created by this action would be quite small. Police officers already face the threat of § 1983 liability for violations of all other constitutional rights of a victim, so this additional avenue for liability represents only a marginal increase in the risk of liability they already face. Thus, the government interest at stake is narrower than it may appear at first blush.

Nevertheless, it is possible that a court would accept police arguments and characterize the government interest so broadly that it would be nearly impossible for a plaintiff to win. For example, the court could determine that the government interest at stake is the ability of police to do their job without being sued for every decision they make, which would effectively paralyze the police to inaction. Most broadly interpreted, the government interest here would amount to absolute police discretion and freedom from liability for all discretionary decisions. Based on the Court’s approach in DeShaney and Castle Rock, if the government interest is framed in this way, the plaintiff will be fighting an uphill battle to show a court that any classification would not substantially relate to achieving such an important interest. In this scenario, a plaintiff would almost never win.

Even with such a broad government interest at stake, however, it is possible that a plaintiff-victim could successfully argue that the government action here did indeed violate equal protection. The key for the plaintiff is to convince a court that the invocation of gender stereotypes does not substantially relate to achieving any government interest due to the inaccuracy of these stereotypes and the gravity of the risk posed when the stereotype is false and the victim is in real danger. As discussed in the second interpretation above, police liability is implicated only if officers employ gender stereotypes. This is a plaintiff’s most salient point at this juncture, as she would need to make clear that holding a police officer liable for his use of gender-based stereotypes would not inhibit his ability to do anything legitimately related to his job.

While these interpretations are based on three distinct ways of framing the government interest, the range of interests is actually a spectrum. Both plaintiff and government positions will inevitably include arguments from each interpretation. Ultimately, a crucial determinant of the success of the argument developed by this Comment will be the way in which its various pieces—the classification, the government interest, the relationship between the two—are framed by litigants and interpreted by courts.

Conclusion

In the context of the enforcement of restraining orders for gender-based violence victims, this Comment contributes to addressing the current state of police unresponsiveness and unaccountability by shifting the question presented to a court. Rather than asking the court to infer from an individual victim’s experience a policy or practice that affects all female victims similarly situated, this Comment’s equal protection argument narrows the lens and instead challenges police treatment of an individual victim. In doing so, it draws on the reasoning of well-established equal protection jurisprudence rejecting the use of gender stereotypes.

Although the exact argument of an individual gender-based violence victim will depend on the facts of her case, the model for an equal protection claim that challenges gender stereotypes in the context of intimate relationships provides a ready and useful template for male victims of intimate partner violence as well as individuals in abusive same-sex relationships. Empirically, police officers are highly unresponsive to male victims of intimate partner violence, whether they are in a heterosexual or same-sex relationship. 164 See, e.g., Carla M. da Luz, A Legal and Social Comparison of Heterosexual and Same-Sex Domestic Violence: Similar Inadequacies in Legal Recognition and Response, 4 S. Cal. Rev. L. & Women’s Stud. 251 (1994). A police officer’s hostile response is likely, at least in part, a result of the assumptions an officer might make about a man asking the police to protect him from a domestic partner—male or female. 165 See, e.g., Sheila M. Seelau & Eric P. Seelau, Gender-Role Stereotypes and Perceptions of Heterosexual, Gay and Lesbian Domestic Violence, 20 J. Fam. Violence 363 (2005). Furthermore, poor police response to lesbians abused by their partners 166 See, e.g., da Luz, supra note 164. is likely caused by the stereotypes that come into play when an individual calls for protection from violence perpetrated by a woman. 167 See, e.g., Seelau & Seelau, supra note 165. Thus, the rejection of gendered stereotypes in the context of gender-based violence could—in addition to creating police accountability where none currently exists—have the added effect of increasing the visibility of non-traditional domestic violence. 168The invisibility of same sex partner violence is not simply a problem among police officers. Legal academics have long ignored the issue and little has been written about it. See Kathleen Finley Duthu, Why Doesn’t Anyone Talk About Gay and Lesbian Domestic Violence?, 18 T. Jefferson L. Rev. 23, 24 (1996).

Finally, while relying on well-established principles, the application of this reasoning in the context of gender-based violence situations contributes to feminist legal scholarship by rejecting stereotypes in the “domestic” context—an important development since up until now, these stereotypes have been rejected in the workplace 169Price Waterhouse v. Hopkins, 490 U.S. 228 (1989). While Hopkins was a Title VII case, rather than an equal protection case, the majority made clear that gender stereotypes have no place in the workplace. Id. at 251 (“As for the legal relevance of sex stereotyping, we are beyond the day when an employer could evaluate employees by assuming or insisting that they matched the stereotype associated with their group, for ‘“[i]n forbidding employers to discriminate against individuals because of their sex, Congress intended to strike at the entire spectrum of disparate treatment of men and women resulting from sex stereotypes.”’ An employer who objects to aggressiveness in women but whose positions require this trait places women in an intolerable and impermissible catch 22: out of a job if they behave aggressively and out of a job if they do not.” (alteration in original) (citation omitted) (quoting L.A. Dep’t of Water & Power v. Manhart, 435 U.S. 702, 707 n.13 (1978))). and in education, 170 See, e.g., United States v. Virginia (VMI), 518 U.S. 515 (1996); Miss. Univ. for Women v. Hogan, 458 U.S. 718 (1984). but have remained steadfast in the domestic sphere. 171 See, e.g., Neil S. Siegel & Reva B. Siegel, Struck by Stereotype: Ruth Bader Ginsburg on Pregnancy Discrimination as Sex Discrimination, 59 Duke L.J. 771, 778–83 (2010).

Footnotes

1Town of Castle Rock v. Gonzales, 545 U.S. 748, 751 (2005). On May 21, 1999, a Colorado trial court issued a temporary restraining order prohibiting Mr. Gonzales from “molest[ing] or disturb[ing] the peace” of Ms. Gonzales and their children, ages seven, nine, and ten, and from coming within one hundred yards of the family home. Id. On June 4, 1999, with both Mr. and Ms. Gonzales present in court, the judge made the order permanent as part of their divorce proceedings after modifying it to allow Mr. Gonzales “a midweek dinner visit arranged by the parties.” Id. at 752–53 (quoting Gonzales v. City of Castle Rock, 366 F.3d 1093 (10th Cir. 2004) (internal quotation marks omitted)).

2Brief for Petitioner at 6, Gonzales v. United States, Case 12.626, Inter-Am. Comm’n H.R. (2008) [hereinafter Gonzales Brief]. Jessica and Simon Gonzales had been married for nine years, during which time they had three daughters. Id. Since 1996, Mr. Gonzales’s behavior had become increasingly “erratic.” Id. Ms. Gonzales decided to file for divorce in 1999, after Mr. Gonzales attempted to hang himself in the family garage and Ms. Gonzales determined that she needed to take steps to protect herself and her children from his behavior. Id.

3 Id. at 9.

4 Id.

5 Castle Rock, 545 U.S. at 753.

6Gonzales Brief, supra note 2, at 10, 11.

7 Castle Rock, 545 U.S. at 753; Gonzales Brief, supra note 2, at 12.

8 Castle Rock, 545 U.S. at 753.

9 Id. According to Ms. Gonzales, the officer refused to send someone to the amusement park because it was outside of the Castle Rock Police Department’s jurisdiction, and refused to put out an all points bulletin—an electronic dissemination of information about a wanted person—because it would “needlessly go statewide and would cost the state money.” Gonzales Brief, supra note 2, at 14.

10 Castle Rock, 545 U.S. at 753.

11 Id. at 754 (internal quotation marks omitted).

12 Id.

13 Id.

14 See United States v. Morrison, 529 U.S. 598, 598–99 (2000). In Morrison, the Court struck down a provision of the Violence Against Women Act, which created a federal cause of action allowing victims of gender-based violence to sue their perpetrators in federal court, as an unconstitutional exercise of congressional power under both the Commerce Clause and section 5 of the Fourteenth Amendment. Id. The Court held that such a remedy encroached too far upon states’ rights despite extensive congressional findings that gender-based violence affects interstate commerce and that pervasive bias in state justice systems against victims of gender-motivated violence was resulting in systematic underprosecution of gender-based violence crimes. Id. The attorneys general of thirty-six states had endorsed this remedy provision as “a particularly appropriate remedy for the harm caused by gender-motivated violence” in light of the “States’ own studies demonstrat[ing] that their efforts to combat gender-motivated violence, while substantial, are not sufficient by themselves.” Brief of the State of Arizona et al. as Amici Curiae Supporting Petitioners, United States v. Morrison, 529 U.S. 598 (2000) (Nos. 99-5, 99-29).

15This Comment uses the terms domestic violence and gender-based violence interchangeably. Many feminist scholars prefer the term gender-based violence because it is more inclusive and attempts to remove the idea that gender-based violence is simply a “domestic” problem not worthy of legal or community attention. However, both terms are widely used in legal scholarship and in litigation documents.

16This Comment will also use the terms restraining order, protective order, and order of protection interchangeably. State statutes generally refer to such court-issued orders using one of these terms, and the differences between them is not relevant to the issues presented here.

17Peter Finn & Sarah Colson, Civil Protection Orders: Legislation, Current Court Practice, and Enforcement, in Legal Interventions in Family Violence 43, 43 (1998) (noting that one major limitation of civil protection orders is their “[w]idespread lack of enforcement, even though such orders allow for expansion of police arrest powers and increase the ability of police to monitor repeat offenders”).

18 Nat’l Ctr. for Victims of Crime, Fact Sheet: Orders of Protection 2 (2005), available at http://www.ncvc.org/ncvc/AGP.Net/Components/documentViewer/Download.aspxnz?DocumentID=39669.

19 Id.

20 Id.

21Robert J. Kane, Police Responses to Restraining Orders in Domestic Violence Incidents: Identifying the Custody-Threshold Thesis, 27 Crim. Just. & Behav. 561, 563 (2000).

22G. Kristian Miccio, If Not Now, When? Individual and Collective Responsibility for Male Intimate Violence, 15 Wash. & Lee J. Civil Rts. & Soc. Just. 405, 434 (2009).

23Deborah L. Rhode, Social Research and Social Change: Meeting the Challenge of Gender Inequality and Sexual Abuse, 30 Harv. J.L. & Gender 11, 17–18 (2007).

24Am. Bar Ass’n Comm’n on Domestic Violence, Domestic Violence Civil Protection Orders (CPOs) by State, Am. Bar Ass’n (June 2007), http://www.abanet.org/domviol/docs/DVCPOChartJune07.pdf.

25 Nat’l Ctr. for Victims of Crime, Legal Ser. Bulletin No. 4, Enforcement of Protective Orders 1 (2002), available at http://www.ncjrs.gov/ovc_archives/bulletins/legalseries/bulletin4/ncj189190.pdf. The consequences of violating an order of protection vary by jurisdiction; they often include arrest and charges of civil or criminal contempt, misdemeanor, or even felony. Id. at 1–2.

26 See Eve S. Buzawa & Carl G. Buzawa, Domestic Violence: The Criminal Justice Response 235 (3d ed. 2003).

27 See Miccio, supra note 22, at 420.

28The text of the Colorado statute reads: “A peace officer shall arrest, or, if an arrest would be impractical under the circumstances, seek a warrant for the arrest of a restrained person” when the restrained person has violated a provision of the order. Colo. Rev. Stat. § 18-6-803.5(3)(b) (1999).

29Gonzales Brief, supra note 2, at 7.

30The mandate on the back of the restraining order states, “YOU SHALL USE EVERY REASONABLE MEANS TO ENFORCE THIS RESTRAINING ORDER. YOU SHALL ARREST, OR, IF AN ARREST WOULD BE IMPRACTICAL UNDER THE CIRCUMSTANCES, SEEK A WARRANT FOR THE ARREST OF THE RESTRAINED PERSON WHEN YOU HAVE INFORMATION AMOUNTING TO PROBABLE CAUSE THAT THE RESTRAINED PERSON HAS VIOLATED OR ATTEMPTED TO VIOLATE ANY PROVISION OF THIS ORDER . . . .” Id. (emphasis omitted).

31 Erwin Chemerinsky, Federal Jurisdiction 405–06 (5th ed. 2007).

32Monell v. Dep’t of Soc. Servs., 436 U.S. 658 (1978).

33Miccio, supra note 22, at 429–30.

34For further discussion of this possibility, see infra text accompanying notes 61–65.

35Miccio, supra note 22, at 428. The Supreme Court’s description of the public duty doctrine has been adopted by most state legislatures. Id. at 429.

3642 U.S.C. § 1983 (2006) (“Every person who, under color of any statute, ordinance, regulation, custom, or usage, of any State . . . subjects, or causes to be subjected, any . . . person . . . to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress.”)

37Monell v. Dep’t of Soc. Servs., 436 U.S. 658 (1978). Section 1983 creates a cause of action against individual officers who violate an individual’s constitutional right while acting under the color of law. Chemerinsky, supra note 31, at 492. Historically, the cause of action was limited to individual officers, but the Court has since broadened its scope to allow a municipality (but not a state government) to be considered a person for § 1983 purposes. Id. at 493, 508–09. However, municipalities may not be held vicariously liable for the actions of individual officers; municipal liability extends only to official policies or customs that cause violations of the Constitution or federal law. Id. at 493.

38Although a federal statute could theoretically provide a cause of action under § 1983, the Supreme Court’s decision in Morrison forecloses this avenue for relief. See supra note 14.

39However, even if a plaintiff is successful in bringing a § 1983 claim, her options for relief are limited. While the text of § 1983 contemplates monetary damages as well as injunctive and declaratory relief, plaintiffs must clear an additional hurdle if they wish to recover damages from a police officer. Police officers enjoy qualified immunity from civil suit for injuries caused by their discretionary conduct. Pearson v. Callahan, 129 S. Ct. 808, 815 (2009). The purpose of the qualified immunity defense is to protect state actors from the perpetual fear of liability for every exercise of discretion, which would hinder their ability to fulfill their public function. Kit Kinports, Qualified Immunity in Section 1983 Cases: The Unanswered Questions, 23 Ga. L. Rev. 597, 600–01 (1989). The protection provided by qualified immunity is vast: it is an affirmative defense that insulates officers from being sued at all, permitting complete dismissal of a claim against them. Pearson, 129 S. Ct. at 815. The shield of qualified immunity can be lifted—and an officer exposed to liability for damages—only if the constitutional right violated by the officer was “clearly established.” Harlow v. Fitzgerald, 457 U.S. 800, 818 (1982). The field of jurisprudence surrounding § 1983 litigation and particularly the qualified immunity defense is extremely complicated and in constant flux. Sheldon Nahmod et al., Constitutional Torts 461 (3d ed. 2010). A complete discussion of these issues is beyond the scope of this Comment. However, it is important to note that a plaintiff-victim would have to show that an officer violated her right to equal protection, and that this particular equal protection right was clearly established, before she could recover any compensatory or punitive damages from an individual officer.

40 See Miccio, supra note 22, at 416.

41The third is an individual’s right to equal protection of the law. This will be discussed in much further detail in Parts II and III infra.

42489 U.S. 189 (1989).

43 Id. at 197.

44This notion has become known as the public duty doctrine, which asserts that the police owe a duty to protect the public in general, as opposed to individual victims. Miccio, supra note 22, at 428.

45 DeShaney, 489 U.S. at 197–99.

46 Id. at 197 n.4.

47 Id. at 200; Miccio, supra note 22, at 428.

48 DeShaney, 489 U.S. at 201. For a more detailed discussion of state tort liability, see also infra text accompanying notes 61–65.

49Miccio, supra note 22, at 428–29; see also Hudson v. Hudson, 475 F.3d 741, 745 (6th Cir. 2007) (rejecting, on similar facts, the plaintiff’s arguments that the police could be liable to the victim under either the special relationship or the state-created-danger theory).

50Town of Castle Rock v. Gonzales, 545 U.S. 748, 755 (2005).

51 Id. at 756 (quoting Bd. of Regents of State Colls. v. Roth, 408 U.S. 564, 577 (1972)) (internal quotation mark omitted).

52 Id. (quoting Roth, 408 U.S. at 577).

53 Id.

54 Id. at 760.

55 Id.

56 Id. The Court concluded that the Colorado statute failed to indicate strongly enough that police were required to act to protect victims because the statute gave officers a choice of actions: they could either arrest the individual whom they had probable cause to believe violated the restraining order, or they could seek a warrant if arrest was impracticable. This choice, the majority reasoned, left room for police discretion and fell short of creating an entitlement to protection. Id.

57 Id.

58Curt Stokes, Note, Restraining Orders and Law Enforcement Liability After Town of Castle Rock, Colorado v. Gonzales, 26 St. Louis U. Pub. L. Rev. 99, 120–21 (2007).

59 Castle Rock, 545 U.S. at 766.

60Amber Fink, Every Reasonable Means: Due Process and the (Non)enforcement of a Restraining Order in Gonzales v. Town of Castle Rock, 24 Law & Ineq. 375, 395 (2006). Many critics of the Castle Rock decision have argued that Justice Scalia’s opinion goes against extensive legislative history documenting the intent of the Colorado legislature to remove police discretion in domestic violence situations. See, e.g., Castle Rock, 545 U.S. at 779–83 (Stevens, J., dissenting). Critics and feminist scholars have also argued that Justice Scalia’s conclusion disregards the plain meaning of the word “shall” in the statute and reflects the Court’s continuing indifference to victims of gender-based violence. See, e.g., Miccio, supra note 22, at 420. After the Supreme Court ruled against her, Jessica Gonzales filed suit against the United States in the Inter-American Commission on Human Rights, alleging that the United States had violated her human rights by failing to protect her and her children from Mr. Gonzales’s actions. Gonzales Brief, supra note 2. That decision is pending; if Ms. Gonzales wins, it will be a strong critique of the Court’s historical resistance to require police officers to protect women from gender-based violence. The European Court of Human Rights recently heard a similar case, Opuz v. Turkey, in which a Turkish victim of domestic violence sued the Turkish government for its failure to protect her and her family members from violence by her husband. In a landmark decision, the court ruled for the plaintiff-victim and, for the first time, held a government accountable for failure “to take adequate steps to protect victims of repeated domestic violence, even absent any active malfeasance on the state’s part.” Tarik Abdel-Monem, Opuz v. Turkey: Europe’s Landmark Judgment on Violence Against Women , 17 Hum. Rts. Brief 29, 29 (2009). Especially interesting is that the court agreed with the plaintiff that the Turkish government violated, inter alia, the prohibition on discrimination contained in the European Convention on Human Rights. Id.

61 Castle Rock, 545 U.S. at 768–69; DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 201–03 (1989).

62 See, e.g., Joshua J. Bennett, Case Note, Town of Castle Rock v. Gonzales and the Awful Truth: Utah’s Mandatory Enforcement Laws Make Police Enforcement of a Victim’s Protective Order Optional, 8 J.L. & Fam. Stud. 405, 422–23 (2006); Danielle Lynn Lordi, Comment, Police Liability Under State Tort Law for Failure to Enforce Protection Orders: The Last Demand for Accountability, 85 Or. L. Rev. 325, 346–49 (2006); Stefanie T. Scott, Note, Trying to Touch the Untouchables: The Challenges Faced by Texas Plaintiffs Asserting Failure-to-Protect Suits Against Police Departments, 27 Rev. Litig. 539, 557–60 (2008); Stokes, supra note 58, at 125.

63 Kristian Williams, Our Enemies in Blue: Police and Power in America 2, 139 (2004) (“Few changes in public safety or security policies can be made without the tacit approval of the police unions, and the officers’ associations are routinely consulted on changes in the criminal code, or in city policies that might indirectly affect police work.”).

64Lordi, supra note 62, at 348.

65Miccio, supra note 22, at 428.

66Victoria L. Holt et al., Do Protection Orders Affect the Likelihood of Future Partner Violence and Injury?, 24 Am. J. Preventative Med. 16, 18–21 (2003).

67Adele Harrell & Barbara E. Smith, Effects of Restraining Orders on Domestic Violence Victims, in Do Arrests and Restraining Orders Work? 214, 234 (Eve S. Buzawa & Carl G. Buzawa eds., 1996) (“The odds of severe violence in cases in which an arrest had been made were less than half that of cases in which no arrest had been made [in the year following the arrest].”).

68Jacquelyn C. Campbell et al., Risk Factors for Femicide in Abusive Relationships: Results from a Multisite Case Control Study, 93 Am. J. Pub. Health 1089, 1090 (2003). Moreover, there is evidence that suggests that obtaining a restraining order itself can increase the likelihood of retaliatory violence, making its enforcement even more important to avoid endangering the victim rather than protecting her. Fink, supra note 60, at 379–80 (citing Brief of National Network to End Domestic Violence et al. as Amici Curiae Supporting Respondents at 20, Town of Castle Rock v. Gonzales, 545 U.S. 748 (2005) (No. 04-278)).

69This Comment recognizes that gender-based violence occurs against both men and women, but it refers to victims as women because of the endemic and disproportionate prevalence of gender-based violence by men against women. Similarly, not all police officers are men, and the use of gendered pronouns runs the risk of incurring the same criticism that forms the premise of the argument advanced by this Comment. However, for the sake of differentiation between officers and victims, and having made this concession, this Comment will refer to police officers using the male pronoun where necessary for clarity.

70For example, the Gonzales tragedy occurred one month after Ms. Gonzales obtained a divorce and permanent restraining order against her husband. Castle Rock, 545 U.S. at 753.

71 See, e.g., id.; Estate of Macias v. Ihde, 219 F.3d 1018, 1026 (9th Cir. 2000) (ex-husband shot and killed victim, as well as victim’s mother and himself, after police officer told victim there was probable cause that ex-husband had violated restraining order and would be arrested); Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026 (3d Cir. 1988) (estranged boyfriend murdered victim at her place of work after stalking her and breaking into her home in the middle of the night in violation of a recently expired emergency restraining order).

72 Ihde, 219 F.3d at 1021–25.

73Estate of Macias v. Lopez, 42 F. Supp. 2d 957, 960 (N.D. Cal. 1999) (internal quotation mark omitted), rev’d sub nom. Estate of Macias v. Ihde, 219 F.3d 1018, 1026 (9th Cir. 2000).

74Jennifer Rios, Note, What’s the Hold-Up? Making the Case for Lifetime Orders of Protection in New York State, 12 Cardozo J.L. & Gender 709, 728–29 (2006).

75Jessica Gonzales, Statement Before the Inter-American Commission on Human Rights (Mar. 2, 2007), available at http://www.aclu.org/human-rights-womens-rights/jessica-gonzales-statement-iachr.

76Press Release, Nat’l Crime Victim Bar Ass’n, National Crime Victim Bar Association Asks Supreme Court to Rule in Favor of Victims on Protective Orders (Feb. 11, 2005), available at http://www.ncvc.org/vb/AGP.Net/Components/documentViewer/Download.aspxnz?DocumentID=39534.

77Miccio, supra note 22, at 416. In cases such as Castle Rock, where the abuser kills himself or is otherwise killed, neither recovery nor individual accountability is possible, highlighting the need for law enforcement accountability and prevention.

78In DeShaney, the majority recognized that “[t]he State may not, of course, selectively deny its protective services to certain disfavored minorities without violating the Equal Protection Clause.” DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 197 n.3 (1989).

79U.S. Const. amend. XIV, § 1.

80 Erwin Chemerinsky, Constitutional Law: Principles and Policies 676 (3d ed. 2006) (noting that equal protection principles protect individuals, not groups, and that an individual can bring an equal protection claim for discrimination in the form of arbitrary government treatment even against a “class of one” (quoting Village of Willowbrook v. Olech, 528 U.S. 562, 564 (2000)) (internal quotation marks omitted)).

81 Id. at 669–70.

82 Id. at 670.

83 Id. at 671.

84 Id.

85 Id.

86 Id. at 671–72.

87 Id. at 671.

88Miss. Univ. for Women v. Hogan, 458 U.S. 718, 724 (1982).

89United States v. Virginia (VMI), 518 U.S. 515, 533 (1996) (internal quotation marks omitted).

90 Hogan, 458 U.S. at 725–26.

91The Supreme Court has invalidated many gender-based classifications challenged under the Equal Protection Clause, but has upheld some. Chemerinsky, supra note 80, at 672; see also infra Part III.B.1.b.

92 Chemerinsky, supra note 80, at 672.

93 Id.

94 Id.

95 Id.

96 Id. at 677–78.

97 Id.

98 See, e.g., Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026 (3d Cir. 1988); Watson v. City of Kansas City, 857 F.2d 690, 693 (10th Cir. 1988) (husband, a police officer himself, raped, beat, and stabbed victim-spouse in the presence of her children, in violation of a restraining order, before committing suicide); Thurman v. City of Torrington, 595 F. Supp. 1521, 1527–29 (D. Conn. 1984) (ex-husband repeatedly stabbed victim in the chest, neck, and throat, in violation of a restraining order, and continued assault after police arrived, kicking victim in the head multiple times, grabbing her child from inside the residence, and dropping child on wounded victim before being arrested).

99 864 F.2d at 1030.

100 Id. at 1031.

101All of these plaintiff-victims had obtained a civil protection order before their attacks. Most sued police departments on both due process and equal protection grounds. This Comment will only discuss the courts’ equal protection analyses since the Supreme Court has rejected both procedural and substantive due process theories as a basis for bringing a § 1983 action on these facts.

102Many of the equal protection claims discussed in this section consisted of both an action against an individual officer and an action against the police department itself for a municipal liability claim. See supra note 32 and accompanying text. A municipal liability claim differs from a claim against an individual officer because a victim can sue the police department only if the entity itself caused the constitutional violation. Monell v. Dep’t of Soc. Servs., 436 U.S. 658 (1978). Municipal liability requires the plaintiff to show that the municipality had a policy or custom that caused the violation of the plaintiff’s constitutional right. Id. at 694. However, claims of this type are difficult to prove because there must be “a direct causal link between a municipal policy or custom and the alleged constitutional deprivation.” City of Canton v. Harris, 489 U.S. 378, 391 (1989). However, municipal liability claims provide deeper pockets for damage awards, as the city itself must pay the damages, whereas a claim against an individual officer will expose only that officer to personal liability. Id. For the argument advanced by this Comment, however, additional distinctions between municipal liability and individual officer liability are ancillary and will not be discussed in detail.

103Prior to DeShaney, one district court allowed a plaintiff to bring a § 1983 equal protection claim against the police on the basis of gender because it found that the police department’s practice of affording inadequate protection to domestic-violence-related calls operated as an administrative classification that implemented the law in a discriminatory fashion on the basis of sex. Thurman, 595 F. Supp. at 1527–29 (denying city’s motion to dismiss plaintiff’s equal protection claim). Tracey Thurman was awarded $2.3 million by a federal jury. Anne Sparks, Feminists Negotiate the Executive Branch: The Policing of Male Violence, in Feminists Negotiate the State: The Politics of Domestic Violence 35, 42 (Cynthia R. Daniels et al. eds., 1997). However, the Thurman court’s rationale is no longer good law. The district court based its decision on a finding that the police had an affirmative duty to protect the plaintiff-victim because they were aware of her situation. Thurman, 595 F. Supp. at 1527. Four years later, in DeShaney, the Supreme Court rejected this notion, stating that the police never have an affirmative duty to protect an individual unless the individual is in state custody. It explicitly rejected the argument that state knowledge of the risk or danger faced by the victim is sufficient to create a special relationship that would trigger an affirmative duty to protect. DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 197–200 (1989).

104The court stated that its rationale for this three-part standard was “to eliminate the all too common shotgun pleading approach to these equal protection claims.” Hynson, 864 F.2d at 1031 n.13. After laying out this standard, the appeals court remanded the case to the district court to apply the test. On remand, the district court in Hynson found sufficient evidence for a jury to infer an intent to discriminate against women, but still granted summary judgment for the police officers on the basis of qualified immunity. Hynson v. City of Chester, 731 F. Supp. 1236, 1240–41 (E.D. Pa. 1990). In its decision, the district court did not give any indication as to how it would have decided had this immunity not been available.

105 Hynson, 864 F.2d at 1031. The second and third factors of this test require that discrimination against women was a motivating factor in this policy and that the plaintiff was injured by this policy or custom. Id. See infra Part II.B.1.b. for discussion of the second factor. Many circuit courts have adopted the Hynson test or one that is analogous. See, e.g., Soto v. Flores, 103 F.3d 1056 (1st Cir. 1997); Navarro v. Block, 72 F.3d 712 (9th Cir. 1995); Eagleston v. Guido, 41 F.3d 865 (2d Cir. 1994); Ricketts v. City of Columbia, 36 F.3d 775 (8th Cir. 1994); McKee v. City of Rockwall, 877 F.2d 409 (5th Cir. 1989).

106 See, e.g., Scott, supra note 62, at 559–62.

107Burella v. City of Phila., 501 F.3d 134 (3d Cir. 2007). In addition to her equal protection claim, Jill Burella also brought § 1983 challenges on the basis of substantive and procedural due process. The Third Circuit denied both due process claims based on the precedents in DeShaney and Castle Rock.

108 Id. at 148.

109 Id.

110 Id. at 148–49.

111 Id. at 149.

112 Id. at 150 (Ambro, J., concurring in part).

113442 U.S. 256, 279 (1979).

114The Court effectively stated that a showing of disparate impact alone, no matter how extreme, was not an actionable violation of equal protection. Id.

115 Id.

116Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026, 1031 (3d Cir. 1988).

117857 F.2d 690, 694 (10th Cir. 1988).

118 Id. at 695.

119 Id. at 696–97 (citations omitted).

120577 F.3d 415, 438 (2d Cir. 2009).

121Prior to Castle Rock but after DeShaney, the Ninth Circuit heard a case in which the district court granted summary judgment to police officers against a plaintiff who alleged violation of her right to equal protection on the basis of both domestic-violence-victim status and gender. Estate of Macias v. Ihde, 219 F.3d 1018, 1028–29 (9th Cir. 2000). On appeal, the Ninth Circuit reversed the grant of summary judgment and remanded to the district court. Id. The police department then settled with the victim’s family for one million dollars. Lordi, supra note 62, at 336 n.71 (citing Sonoma County, California, Sheriff’s Department Settles Domestic Violence Murder Case for One Million Dollars, 8 Domestic Violence Rep. 1, 15 (2002)). This case, while encouraging, still does not provide a model for a successful equal protection challenge because it did not reach trial.

122 See, e.g., Hynson v. City of Chester, Legal Dep’t, 864 F.2d 1026, 1026 (3d Cir. 1988); Watson, 857 F.2d at 690.

123 See, e.g., Watson, 857 F.2d at 696.

124 See, e.g., Scott, supra note 62, at 559–62.

125 Chemerinsky, supra note 80, at 672.

126 Id.

127 Id.

128 Id. at 683.

129In Okin, the Second Circuit noted the district court’s finding that the plaintiff “relie[d] exclusively on her own story—which, while heartrending, does not get her where [the test] says she must go.” Okin v. Cornwall-on-Hudson Police Dep’t, 577 F.3d 415, 438 (2d Cir. 2009) (quoting Okin v. Cornwall-on-Hudson Police Dep’t, No. 04 Civ. 3679 (CM), 2006 U.S. Dist. LEXIS 75881, at *58 (S.D.N.Y. Oct. 13, 2006)) (internal quotation mark omitted).

130 See supra text accompanying notes 50–60.

131There is a key distinction between this discussion of discretion and the role of discretion in the procedural due process line of cases, as discussed by Justice Scalia in the Castle Rock opinion. In that case, in order to show that Ms. Gonzales’s procedural due process rights were violated, it was necessary to show that she was entitled to enforcement of her restraining order, and the prerequisite to entitlement was a lack of police discretion. Therefore, she could not make a successful claim unless the mandatory arrest statute eliminated any and all police discretion, an argument that Justice Scalia rejected. For this type of equal protection challenge, in which only an officer’s use of gender stereotyping is at issue, there is no need to eliminate discretion entirely; the only requirement is that the discretion be exercised for an appropriate reason, and not on the basis of an impermissible, unconstitutional stereotype. This claim does not invalidate discretion; it simply challenges how the police use their discretion. Therefore, this strategy is more moderate than both a procedural due process-based claim and a state tort remedy, each of which is premised on violation of a mandatory arrest statute. See supra text accompanying notes 50–59 and 61–62. In fact, this equal protection claim does not require states to have enacted mandatory arrest statutes at all, because it does not question officers’ actual ability to exercise their discretion—it simply requires them to do so on an appropriate basis.

132J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 135 (1994) (quoting Schlesinger v. Ballard, 419 U.S. 498, 508 (1975)).

133 Id. at 130–31. In Mississippi University for Women v. Hogan, the Court stated that “[a]lthough the test for determining the validity of a gender-based classification is straightforward, it must be applied free of fixed notions concerning the roles and abilities of males and females.” 458 U.S. 718, 724–25 (1982).

134511 U.S. at 139–40.

135458 U.S. at 725, 730.

136518 U.S. 515, 541–42 (1996).

137458 U.S. at 726 (quoting Craig v. Boren, 429 U.S. 190, 198 (1976)).

138518 U.S. at 540.

139 Id. at 541 (quoting United States v. Virginia, 766 F. Supp. 1407, 1434 (W.D. Va. 1991)).

140 Id. (alterations in original) (quoting Virginia, 766 F. Supp. at 1434) (internal quotation marks omitted).

141 Id.

142 Id. at 540.

143 Id. at 541–42 (alteration in original) (quoting and citing J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 139 n.11 (1994)).

144 Id. at 550.

145Two other important cases in the Court’s gender-based stereotyping jurisprudence, mentioned above, are Mississippi University for Women v. Hogan and J. E. B. ex rel. T. B. v. Alabama. In Hogan, a state-run university restricted admission to its nursing program to women. 458 U.S. 718, 720 (1982). A male applicant challenged the policy under equal protection law. Id. at 721. Mississippi argued that the admission policy was an instance of educational affirmative action. Id. at 727. The Court rejected this argument, holding, inter alia, that the State “failed to establish that the alleged objective is the actual purpose underlying the discriminatory classification.” Id. at 730. Instead, the majority stated that the policy of excluding males from the nursing program “tends to perpetuate the stereotyped view of nursing as an exclusively woman’s job.” Id. at 729. In J. E. B., the defendant in a paternity suit challenged on equal protection grounds the government’s use of its peremptory strikes to empanel an all-female jury. 511 U.S. 127, 129 (1994). He argued that the exclusion of men from the jury on the basis of their gender violated equal protection principles. Id. The government defended its use of gender-based peremptory strikes by arguing that men and women have different attitudes and that women would be more sympathetic to a paternity suit and more likely to find in favor of paternity. Id. at 137–38. The government contended that its interest was establishing the paternity of a child born out of wedlock, which justified the use of the gender-based discrimination. Id. at 137 n.8. The Court rejected both the interest advanced by the government and the proffered justification. Id. at 137–38. Discussing the actual interest of the state, the Court stated, “What respondent fails to recognize is that the only legitimate interest it could possibly have in the exercise of its peremptory challenges is securing a fair and impartial jury.” Id. at 137 n.8. Holding that the classification did not substantially relate to advancing the state’s interest under intermediate scrutiny, the Court stated it would “not accept as a defense to gender-based peremptory challenges ‘the very stereotype the law condemns.’” Id. at 138 (quoting Powers v. Ohio, 499 U.S. 400, 410 (1991)). The Court further noted:Respondent offers virtually no support for the conclusion that gender alone is an accurate predictor of juror’s attitudes; yet it urges this Court to condone the same stereotypes that justified the wholesale exclusion of women from juries and the ballot box. . . . We have made abundantly clear in past cases that gender classifications that rest on impermissible stereotypes violate the Equal Protection Clause.Id. at 138–39, 139 n.11. These cases, as well as VMI, demonstrate the Court’s commitment to eradicating the use of gendered stereotypes under equal protection jurisprudence.

146Balistreri v. Pacifica Police Dep’t, 855 F.2d 1421, 1427 (9th Cir. 1988), superseded by 901 F.2d 969 (9th Cir. 1990).

147The officer’s own statement shows that he believed Ms. Balistreri when she told him that she had been attacked by her husband. Furthermore, the facts of the case provide sufficient evidence that Mr. Balistreri was in violation of his restraining order. Id. at 1423, 1426.

148 Id. at 1423.

149 Id. at 1427.

150Gonzales Brief, supra note 2, at 10–15.

151 Id. at 12.

152Town of Castle Rock v. Gonzales, 545 U.S. 748, 753 (2005).

153Gonzales Brief, supra note 2, at 24 (footnotes omitted).

154 Id. at 9.

155See supra note 11 for an instance in which the officer had neither a factual basis for ignoring Ms. Gonzales’s request, nor another task to which he needed to attend; he simply declined to assist her and went to dinner instead.

156 See, e.g., Emily C. Rutledge, Comment, Healing Jurisprudential “Bruises”: A Critique of the Failure of Due Process to Account for the Intersection of Race, Class, and Gender in the Post-Castle Rock Era, 53 How. L.J. 421, 434–35 (2010) (discussing a variety of stereotypes that police hold about battered women generally, and poor, minority victims in particular).

157It is important to note here that gendered stereotypes are not appropriate even if they are true of most women. See supra note 141; J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 139 n.11 (1994) (“We have made abundantly clear in past cases that gender classifications that rest on impermissible stereotypes violate the Equal Protection Clause, even when some statistical support can be conjured up for the generalization.”). Even if many women are hysterical, or do exaggerate the dangerousness of their situations, and even if the officer has personally experienced this, it would still be unconstitutional for that stereotype to influence the treatment given to an individual woman. Obviously, there are examples of women who do not fit the stereotype, such as Jessica Gonzales, whose fear and panic were well-founded. Treating her based on the stereotype both was unconstitutional and resulted in horrific but avoidable consequences.

158By promoting police responsiveness to gender-based violence victims without advocating for, or relying on, mandatory arrest provisions, this argument avoids the many feminist criticisms of mandatory arrest laws. See, e.g., Leigh Goodmark, Autonomy Feminism: An Anti-Essentialist Critique of Mandatory Interventions in Domestic Violence Cases, 37 Fla. St. U. L. Rev. 1 (2009); G. Kristian Miccio, A House Divided: Mandatory Arrest, Domestic Violence, and the Conservatization of the Battered Women’s Movement, 42 Hous. L. Rev. 237 (2005).

159 See, e.g., Laurence H. Tribe, Lawrence v. Texas: The “Fundamental Right” That Dare Not Speak Its Name, 117 Harv. L. Rev. 1893, 1900 (2004) (noting that in Bowers v. Hardwick, which involved a statute banning any act of sodomy, “the Supreme Court went out of its way to recast the plaintiff’s claim to substantive protection under the Due Process Clause for [the plaintiff’s] private sexual relationships as an asserted ‘fundamental right to engage in homosexual sodomy’” (quoting Bowers, 478 U.S. 186, 191 (1986))); supra note 141 (describing the J. E. B. Court’s rejection of the government’s argument for the use of peremptory strikes based on gender and the majority’s subsequent reinterpretation of the actual interest at stake).

160In addition to demonstrating that the classification substantially relates to achieving the government interest, the government must also show that the interest advanced during litigation is the actual purpose behind the classification, not a post hoc justification. United States v. Virginia (VMI), 518 U.S. 515, 535–36 (1996) (“[A] tenable justification must describe actual state purposes, not rationalizations for actions in fact differently grounded.”).

161 See supra text accompanying notes 53–57; see also supra note 131.

162A simple failure to respond or protect, without any allegation of dissimilar treatment of similarly situated persons, does not violate equal protection. This is because individual citizens hold no absolute right to police protection. See supra text accompanying notes 33 and 43–44. Equal protection simply ensures a citizen’s right to be protected equally to others that are similarly situated; that is, police may decline to protect everyone, or they may provide poor responses across the board. In this situation, there is no allegation of inappropriate classifications on the basis of a suspect or quasi-suspect class and, thus, no equal protection claim.

163 Brief for United States as Amicus Curiae Supporting Petitioner at 11–12, Town of Castle Rock v. Gonzales, 545 U.S. 748, 751 (2005) (No. 04-278).

164 See, e.g., Carla M. da Luz, A Legal and Social Comparison of Heterosexual and Same-Sex Domestic Violence: Similar Inadequacies in Legal Recognition and Response, 4 S. Cal. Rev. L. & Women’s Stud. 251 (1994).

165 See, e.g., Sheila M. Seelau & Eric P. Seelau, Gender-Role Stereotypes and Perceptions of Heterosexual, Gay and Lesbian Domestic Violence, 20 J. Fam. Violence 363 (2005).

166 See, e.g., da Luz, supra note 164.

167 See, e.g., Seelau & Seelau, supra note 165.

168The invisibility of same sex partner violence is not simply a problem among police officers. Legal academics have long ignored the issue and little has been written about it. See Kathleen Finley Duthu, Why Doesn’t Anyone Talk About Gay and Lesbian Domestic Violence?, 18 T. Jefferson L. Rev. 23, 24 (1996).

169Price Waterhouse v. Hopkins, 490 U.S. 228 (1989). While Hopkins was a Title VII case, rather than an equal protection case, the majority made clear that gender stereotypes have no place in the workplace. Id. at 251 (“As for the legal relevance of sex stereotyping, we are beyond the day when an employer could evaluate employees by assuming or insisting that they matched the stereotype associated with their group, for ‘“[i]n forbidding employers to discriminate against individuals because of their sex, Congress intended to strike at the entire spectrum of disparate treatment of men and women resulting from sex stereotypes.”’ An employer who objects to aggressiveness in women but whose positions require this trait places women in an intolerable and impermissible catch 22: out of a job if they behave aggressively and out of a job if they do not.” (alteration in original) (citation omitted) (quoting L.A. Dep’t of Water & Power v. Manhart, 435 U.S. 702, 707 n.13 (1978))).

170 See, e.g., United States v. Virginia (VMI), 518 U.S. 515 (1996); Miss. Univ. for Women v. Hogan, 458 U.S. 718 (1984).

171 See, e.g., Neil S. Siegel & Reva B. Siegel, Struck by Stereotype: Ruth Bader Ginsburg on Pregnancy Discrimination as Sex Discrimination, 59 Duke L.J. 771, 778–83 (2010).

Notes and Comments Editor, Emory Law Journal; J.D. Candidate, Emory University School of Law (2011); B.A., Stanford University (2008). I would like to thank Professor Victoria Nourse for her guidance in developing the idea for this Comment, Professor Kay Levine for her mentorship and help in clarifying my writing, and Professor Martha Fineman for her advice and support. My sincerest appreciation also goes to the executive editorial board of the Emory Law Journal, particularly to Laura Pisarello and Jena Sold, for their wisdom, editorial expertise, and patience. Finally, I could never have completed this project without the love and confidence of my family and friends.