Introduction
As of July 2020, 19 states and the District of Columbia have adopted Extreme Risk Protection Order laws (what we call “ERPO” or “extreme risk” laws). ERPO laws allow law enforcement, and often other statutorily defined parties, to petition a court to temporarily remove firearms from someone whom the court determines poses a threat to themselves or others. These laws are modeled on Connecticut's 1999 “risk warrant” law,1 which aimed at removing firearms from people at high risk of violence to self or others. A new wave of laws, beginning in 2014 with California's Gun Violence Restraining Order,2 has seen widespread adoption in the last few years.Reference Campbell, Yablon and Mascia3 Although these types of laws are popularly referred to as “red flag laws,” we avoid the term, except for identification purposes in the title of this article, due to potential stigma associated with it.4
One primary way in which the recent wave of ERPO laws go beyond the early Connecticut statute is by allowing not just law enforcement, but also family members and sometimes others to directly petition a judge to order temporary removal of firearms from a dangerous individual, employing a legal process similar to what all 50 states have in place for domestic violence restraining orders.Reference Booker and Rose5 In the years since the 2018 shooting at Marjory Stoneman Douglas High School, extreme risk laws have continued to evolve. Jurisdictions like California, Maryland, Hawaii, New York, and the District of Columbia have expanded the class of petitioners eligible to bring a direct petition for an ERPO, including in an initial, ex parte proceeding.6 In several of these states, co-workers, employers, healthcare professionals, and school faculty and staff now join law enforcement, family, and household members as eligible petitioners.7
The significant expansion of the ERPO petitioner classes raises questions about how far states are constitutionally permitted to expand such classes consistent with the Due Process Clause. Though the Constitution likely poses no barrier to extreme risk laws that permit law enforcement, family members, and perhaps trained professionals like healthcare workers to directly petition for a court order, the questions become harder as petitioner classes extend to those with more indirect or tangential relationships with the prospective respondent. Because the risk of erroneously depriving someone of a protected right is a key component of the due process inquiry,8 groups with less direct access to information and fewer connections to the respondent may more often petition based on mistaken perceptions, unconscious bias, incomplete information or even, rarely, outright fabrication. Broad expansions may thus, critics charge, result in additional ERPOs being erroneously issued.Reference Sullum9
In this article, we argue that, to assuage concerns with broadening the class of petitioners, states can adopt what we call a “double-filter provision,” which is exemplified by New Mexico's recent extreme risk statute.10 The main component of such a provision is a special statutory category — “reporting party” — that enables a broader social network, such as co-workers or school administrators, to request that a law enforcement officer (“LEO”) file a petition for an ERPO, but does not give reporting parties themselves a right to file a petition directly with the court. LEOs provide the first filter on these requests and, crucially for our proposal (and the New Mexico law), LEOs legally must seek an ERPO from the court if they find the reporting party's information credible. That information is then transmitted to the court (the second filter) in the form of a sworn affidavit of the reporting party.
We argue that employing LEOs and judges as double filters provides key benefits relevant to both the due process analysis and to the life-saving goals of ERPOs by funneling more reliable information to the court system and enhancing the transparency and accountability of ERPO enforcement. In these ways, expanding the sources of information through a double-filter provision maintains the appropriate balance between individual rights and government interests that the due process inquiry requires.
Employing LEOs and judges as double filters provides key benefits relevant to both the due process analysis and to the life-saving goals of ERPOs by funneling more reliable information to the court system and enhancing the transparency and accountability of ERPO enforcement. In these ways, expanding the sources of information through a double-filter provision maintains the appropriate balance between individual rights and government interests that the due process inquiry requires.
Part I of this article lays out the relevant policy landscape, including the key structural differences between original and modern ERPO statutes. Part II briefly assesses the constitutionality of expanding petitioner classes in ERPO statutes, specifically as it relates to those petitioners filing ex parte orders. Part III argues that states considering further expansion of ERPO petitioner classes to broad social networks should adopt double-filter provisions.
I. The Context and History of Extreme Risk Laws
In 1999, Connecticut became the first state to pass a law allowing police to temporarily remove guns from an individual when there is “probable cause to believe … that person poses a risk of imminent personal injury to himself or herself or to other individuals[.]”11 The policy borrows from the same legal framework as domestic violence restraining orders (DVROs),12 which have now been adopted in various forms in all 50 states and the District of Columbia.13 Connecticut's statute pioneered the practice of preemptive gun removal as a civil court action, which neither requires nor generates a record of criminal or mental health adjudication. The Connecticut statute only allows LEOs or state prosecutors, after independently investigating and determining probable cause, to seek a judicial warrant authorizing them to remove firearms from an individual found to pose an imminent risk.14
Since 2014, many of the new wave of extreme risk laws have expanded the original Connecticut model by, as relevant here, broadening the class of petitioners eligible to directly petition for an ERPO. Of the 20 jurisdictions with an ERPO law as of this writing, 13 also allow a family or household member to directly petition a court for a gun removal order.15 And some of these states have expanded the class of eligible petitioners to additional categories: school administrators,16 co-workers,17 and healthcare professionals.18
The theory undergirding petitioner class expansion is that “those who teach or work with a person and have frequent interaction may see the early warning signs and be the first to know that the person is at severe risk of harming self or others with a firearm.”19 In other words, by expanding the category of persons eligible to file an ERPO petition to those in one's wider social or professional network, there is a higher likelihood that a dangerous gun owner will be identified and separated from his or her firearm(s) before a violent incident occurs.
The recent expansion of extreme risk petitioner classes may present a way to increase the frequency of pre-incident interventions that save lives. In Maryland, for example, which has extended the class of petitioners, recent statistics show that family members were petitioners in approximately 41% of cases and physicians petitioned in 1% of cases.20 Researchers who recently interviewed Maryland physicians about the state's ERPO law reported that few were familiar with the process, but that after the law was explained, most physicians (1) had interacted with patients who would be likely candidates and (2) would likely file a petition.Reference Frattaroli, Hoops and Irvin21 As this and other recent research has shown, the use and efficacy of ERPO laws are in part a function of awareness.Reference Pallin, Schleimer, Pear and Wintemute22 But expanding the category of petitioners has costs, too — a potentially higher risk of erroneous deprivation. As the categories expand, the possibility of less accurate or relevant information entering the calculation could increase as individuals have less context for the respondent's behavior or may be motivated by conscious or unconscious bias. Because such possible errors play directly into the constitutional calculus, we turn next to that question.
II. Assessing the Constitutionality of Expanded ERPO Petitioner Classes
Although many restrictions on gun possession implicate the Second Amendment, ERPOs most directly raise a different constitutional question about the procedure by which rights can be deprived. Our focus here is thus on the Due Process Clause. That Clause generally requires conditions, such as proper notice and the opportunity for a fair hearing,23 before the state deprives an individual of constitutionally-protected liberty or property.Reference Fallon24 However, the Supreme Court has often recognized exceptions in cases where an adversarial hearing is postponed until after a deprivation if a compelling government interest is at stake, as long as the risk of erroneous deprivation is not excessively high.25 Scholars (including one of us) have argued that ex parte ERPOs fit into the Court's exceptions to pre-deprivation hearings.Reference Blocher and Charles26 We are interested here in whether expanding petitioner classes in ERPO statutes jeopardizes that constitutional footing.
A court reviewing an ERPO law is likely to use the Court's balancing test laid down in Matthews v. Eldridge to determine what process is due. This test instructs courts to weigh three factors:
First, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government's interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirements would entail.27
In terms of the first factor, ERPOs implicate two private interests: (1) the Second Amendment right of a “responsible, law-abiding citizen”28 to “keep and bear arms,”29 and (2) the property interest in the firearms themselves. The last factor, the government interest in the prevention of gun violence in the community at-large, is uncontestably significant.30 Because the private and government interests at stake are both compelling, a court's evaluation will likely turn on whether the risk of erroneous deprivation has been appropriately mitigated through the law's procedures.
Most existing extreme risk laws contain numerous mechanisms to mitigate the risk of erroneous orders. Judges, not administrative functionaries like court clerks,31 entertain the orders and can only grant emergency ex parte orders after a petitioner provides sufficient evidence to carry her burden (ranging from good cause to clear and convincing evidence) that an individual poses an immediate risk of harming themselves or others. All states carry a misdemeanor penalty for filing a knowingly false petition.32 Current extreme risk laws require courts to hold a full hearing in which both petitioner and respondent participate within as little as two days after the initial ex parte order is served and no longer than a few weeks.33 And even when a final order is granted after a hearing, it is usually limited to a duration of six months to a year,34 though a respondent may typically move to have an ERPO terminated sooner.35
With respect to petitioners, allowing LEOs, family, and household members to directly file petitions allows access to this legal tool by those considered best situated to detect the warning signs of potential dangerousness in armed subjects.36 The types of information these individuals possess is not likely to undermine the accuracy and fairness of the process.
Would expanding this same direct petitioning right to employers, co-workers, school employees, healthcare professionals, or others undermine the procedural safeguards that have been put in place to protect against erroneous deprivations? The ACLU argues that it does. In its opposition to California's Assembly Bill 61, enacted in October 2019, the ACLU of California argued against expanding the class of eligible GVRO petitioners to employers, co-workers, and school staff or faculty because they “lack the relationship or skills required to make an appropriate assessment” of an individual's dangerousness.37 Though the ACLU supported California's existing extreme risk law, which allowed law enforcement and family members to petition the court directly, it argued that “expanding the parties that could apply for such an ex parte restraining order to include all the parties listed above” would “upset[]” the balance between preventing gun violence and protecting civil liberties struck in the original law and “create[] significant potential for civil rights violations.”38 The ACLU chapters of Pennsylvania and Rhode Island have made similar arguments against expansive ERPO bills introduced in those states.39 The ACLU argument drives at the idea that by expanding the categories of eligible ERPO petitioners to those lacking “relationship or skills to make an appropriate assessment” policymakers will increase the risk of erroneous deprivation.
Measuring this possible risk of an erroneous deprivation is difficult. In theory, an ERPO is granted in error when (1) the respondent does not actually pose a risk to himself or others, or (2) the petitioner fails to meet her burden of proof in establishing that he does. But it is hard to translate that definition into a workable empirical analysis. For one, there is no way to objectively measure whether a standard of proof was met or whether a person was sufficiently dangerous. One imperfect substitute is to consider the percentage of ex parte ERPOs that result in full-length ERPOs. That provides some sort of statistical measurement that might show that the information leading to the ex parte order was insufficient, incomplete, or fabricated, but it could also turn out that full orders were not entered because the respondent was incarcerated or civilly committed, or that the relevant witnesses were too afraid to testify against the respondent,40 or that the crisis prompting the emergency order had ceased before the full hearing. Erroneous orders certainly exist, but it is not easy to observe their frequency.
The risk for an erroneous order depends on the type and quality of the new information made available through additional individuals and on how judges process this information. In DVRO laws, on which the basic ERPO structure was built, only family members or those with similarly close personal relationships can seek a DVRO.41 These laws have withstood due process challenges.42
Since data gathering on more expansive extreme risk statutes is still in its early stages, comparative data on petitioner classes is still hard to find. But what legislatures in California, Hawaii, New York, Maryland, and DC have concluded is that allowing trustworthy members of a community to voice concern when they witness suspicious, threatening, or dangerous behavior by armed individuals increases the likelihood that incidents of gun violence will be prevented. These are undoubtedly hard questions, and the concerns that have been raised by the ACLU and other groups are important.
But whatever the ultimate answer to the constitutional question of permitting these additional groups direct access to the courts, we propose a solution for states considering expanding the class of petitioners that, in our view, best balances governmental and individual interests while decreasing the chance for error.
III. A New Model: The Double-Filter Provision
To more fully consider the ACLU's concern with expanded petitioner classes, we highlight some of the costs and benefits of expanding the pool of reporting individuals before identifying and addressing our preferred solution.
A. The Costs and Benefits of a Broader Class
Expanding the pool of petitioners could have several drawbacks. First, a larger number of requests might lead to more questionable petitions reaching LEOs and judges. This may mean that higher numbers of ERPOs are granted out of an abundance of caution for public safety, and not because there is particularly strong evidence that a respondent is a danger to themselves or others. Second, with the overall number of petitions filed increasing, the administrative costs of managing those filings will also rise. (Although the costs of a procedure are a factor in the constitutional test, a state can of course choose to create procedures that increase its own expenses.) The costs might include workload increases for state courts; additional workloads for LEOs tasked with filing more paperwork, prosecuting petitions, or retrieving weapons from the homes of dangerous individuals; law enforcement agencies or gun dealers mandated to store weapons for extended periods; and the data management needed to keep track of all this information.43
The potential payoff in exchange for these costs is the same paramount goal for all weapon restrictions — lives saved. A greater number of community members capable of requesting ERPOs means a greater number of eyes and ears employed to aid law enforcement in the pre-incident intervention of dangerous individuals. This is the same idea behind enlisting the support of the public in thwarting domestic terrorist plots.44
There are aspects of our identities, attitudes, and intentions that we do not share with family or household members, much less a LEO. There are things we might tell a co-worker, a healthcare professional, or a friend in our social network that we might never say to our significant other, or our parents. If that information could save a life, shouldn't the system make it easy to use that information to take action? Present estimates indicate that for every ten to twenty ERPOs granted, one suicide is averted.Reference Swanson and Swanson45 And that estimate is based on data derived from Connecticut and Indiana, states that only allow LEOs to file ERPO petitions. Giving categories of people with potentially better knowledge about the character or emotional status of a respondent the ability to raise that information may make the risk of erroneous deprivation smaller, not greater.
B. The Double-Filter Proposal
A state considering expanding the class of those eligible to seek an ERPO need not choose between providing these individuals direct access to court and shutting them out of the ERPO process altogether. We propose that states inclined to expand the pool out to a broader social network adopt what we call a “doublefilter provision,” like New Mexico's recent adoption of a “reporting party” category.46 This provision would allow a wide array of specifically listed groups, like co-workers, employers, and school administrators, to request that LEOs file a petition for an ERPO. Unlike in states where worried citizens can only raise concerns with LEOs, who are the only party allowed to petition directly, reporting parties are not passive conveyers of information. Instead, they must file a sworn affidavit supporting the LEO's petition that goes to the court.47 Such sworn documents subject reporting parties to the same civil and criminal penalties that similarly-situated petitioners in other states are for false statements. Of course, in such a process, LEOs could still directly petition without a reporting party if they perceive threatening or dangerous behavior themselves.
Under this framework, LEOs are the first filter with judges serving as the second. In this way, a broader number of individuals with relevant information have statutory rights to request that LEOs file for a petition against a specific person. Two crucial parts of this framework help distinguish it from other possible ERPO designs: (1) LEOs are required to petition for an ERPO if they are presented with “credible information from a reporting party that gives the agency or officer probable cause to believe that a respondent poses a significant danger of causing imminent personal injury to self or others,”48 and (2) if they decline to seek a requested petition from a reporting party, LEOs must inform the chief law enforcement officer of the respondent's county.49 We would also add that declination information should be reported to the state's Attorney General and that as much information as possible should be publicly released about declination decisions, rates, and rationales on a periodic basis.
In some ways, our proposal resembles the original Connecticut risk warrant law, because both rely on several stages of review before a petition can be filed. But the double-filter provision here is importantly different. Most notably, reporting parties are a statutory category with respect to whom LEOs owe special duties. If someone in that category requests that the LEO file an ERPO, the LEO either must do so if she finds the information credible or report why she has declined to do so. And that information on declinations must be reported, tracked, and publicly disclosed. There are no similar obligations under Connecticut's original law.
The double-filter provision enables LEOs to provide an extra check on misleading, baseless, or incomplete requests. It may also help assuage concerns about expanded petitioner classes, including some of the ACLU's objections. LEOs' first-level review also helps decrease the possibility that expanded petitioning rights may result in the biased use of the procedure, such as by those who are more likely to code (unconsciously or not) behavior as threatening when it comes from Black or Brown students, neighbors, or co-workers. To be sure, these concerns are not fully eliminated, but inserting an extra filter before views are aired in court filings should reduce the likelihood of petitioners' potential misuse of the tool.
On the other hand, requiring a LEO filter can be problematic if, for example, a member of law enforcement is the individual who poses a risk of imminent personal injury to himself or herself or to other individuals. Other LEOs might be hesitant or resistant to filing a petition against their colleague. It can also generate difficulty if — as is the case in some New Mexico jurisdictions — LEOs have stated their refusal to seek or enforce such orders by proclaiming their jurisdictions “Second Amendment sanctuaries.” But New Mexico's law adds several counterbalancing features to mitigate these concerns, which we think can be slightly modified for states that want to adopt broader classes of requesters. In New Mexico, if a LEO is the potential petition respondent, then a local prosecutor can file the petition. And if a LEO declines to file a requested petition, the officer must inform the sheriff of the respondent's county.50 (The law went into effect on July 1, 2020, and no data is yet available on the frequency of the practice.) We would modify this requirement to mandate LEOs, or the county sheriffs, to provide this information to the state's chief law enforcement officer, typically the Attorney General, so that recalcitrant local LEOs, like sheriffs refusing to enforce gun laws, are not the only repository of information about declined requests; they should also provide a rationale for why they have declined to seek such an order.
New Mexico's law also strips LEOs of their state immunity for harm that results if they refuse to follow the law.51 Although such a provision would undoubtedly provide extra incentive for LEOs to seek appropriate orders, we are agnostic about whether other states should include similar provisions in their laws. The key point to maintain the efficacy of the double-filter provision is the transparency and accountability that comes from a legal obligation to file warranted ERPOs and a reporting provision for declinations. The state should provide a mechanism to publicly report as much information as possible on declination decisions, such as rates, rationales, and case outcomes, while respecting the privacy rights of those against whom no court order was ultimately sought.
Not only does this mechanism provide an initial filter of potentially spurious or unreliable claims made against a respondent, but it keeps LEOs accountable to reporting parties and the public as well. Unlike the role that everyday citizens play in alerting law enforcement to troubling behavior that might lead to a LEO-initiated ERPO in other states, such as by calling 911 or showing up to a station, reporting parties are more central. They play a formal role in court as affiants, and LEOs are legally required to act on credible information from reporting parties and to report and disclose when they decline to do so. This combination of reduced LEO discretion and enhanced accountability helps expanded extreme risk laws more easily satisfy the due process analysis because it lowers the marginal risk for error. It adds a disinterested, neutral observer into the equation before a judicial officer reviews the petition and erects an additional burden on the reporting party to satisfy that LEO. This kind of error-reducing measure allows the state both to vindicate the individual interests in correctly-adjudicated petitions and, because of the expanded class of requesters, to increase the likelihood that more lives will be saved with additional information coming into the process.
Conclusion
Extreme risk laws can save lives without unduly affecting the right of responsible, law-abiding citizens to keep and bear arms. The laws impose only a temporary burden on those whom a court finds to be at serious risk to himself or others. We offer a balanced model that has the additional benefit of expanding the reporting circle more widely but that also reduces the possibilities for errors as the circles spread. The double-filter provision increases accountability, adds transparency, decreases error, and filters a greater amount of credible information on serious risks to the court system.
Initial data in some jurisdictions with expanded petitioner classes demonstrate that there are a set of people with relevant risk information willing to make those declarations under oath in hopes of averting harm.52 We suspect that adding a layer of review will encourage more individuals to come forward, fewer mistaken claims to slip through, and more lives to be saved.
