Identifying Armed Respondents to Domestic Violence Restraining Orders and Recovering Their Firearms: Process Evaluation of an Initiative in California
Shannon Frattaroli, PhD, MPH, Barbara E. Claire, Katherine A. Vittes, PhD, MPH, and Daniel W. Webster, ScD, MPHCorresponding author.G. J. Wintemute drafted the article, obtained funding, and supervised the study. G. J. Wintemute, S. Frattaroli, and B. E. Claire acquired the data. B. E. Claire performed administrative, technical, and material support. All authors designed and developed the study, analyzed and interpreted the data, and critically revised the article for important intellectual content.
Abstract
Objectives. We evaluated a law enforcement initiative to screen respondents to domestic violence restraining orders for firearm ownership or possession and recover their firearms.
Methods. The initiative was implemented in San Mateo and Butte counties in California from 2007 through 2010. We used descriptive methods to evaluate the screening process and recovery effort in each county, relying on records for individual cases.
Results. Screening relied on an archive of firearm transactions, court records, and petitioner interviews; no single source was adequate. Screening linked 525 respondents (17.7%) in San Mateo County to firearms; 405 firearms were recovered from 119 (22.7%) of them. In Butte County, 88 (31.1%) respondents were linked to firearms; 260 firearms were recovered from 45 (51.1%) of them. Nonrecovery occurred most often when orders were never served or respondents denied having firearms. There were no reports of serious violence or injury.
Conclusions. Recovering firearms from persons subject to domestic violence restraining orders is possible. We have identified design and implementation changes that may improve the screening process and the yield from recovery efforts. Larger implementation trials are needed.
Intimate partner violence is a significant threat to the public’s health and safety. Women are at greatest risk. An estimated 1127 women were murdered and some 605 000 were assaulted by their intimate partners in 2011 in the United States.1,2 The National Intimate Partner and Sexual Violence Survey suggests that 35.6% of women in the United States have experienced intimate partner violence at some time in their lives.3
Firearms figure prominently in this violence. Most intimate partner homicides involve firearms,4 and women are at least twice as likely to be murdered by intimate partners using firearms as by strangers using any weapon.5 Abusers with firearms are 5 to 8 times as likely to kill their victims as are those without firearms.6,7
Abusers with firearms also use them in nonlethal ways. About 3.5% of women report that an intimate partner has threatened them with a firearm.8 Firearm-owning abusers are 7.8 times as likely as are others to threaten their partners with firearms and no less likely to do so with other weapons.9 Among California women in shelters, two thirds of those who came from households with firearms reported that their partner used a firearm against them, most often as a threat.10
To help prevent such violence, federal statute prohibits the purchase and possession of firearms by persons subject to domestic violence restraining orders issued at hearings where both parties are present. By 2008, 10 states required and 20 states authorized courts to order respondents to surrender their firearms for the duration of the order.11 Many states extend these prohibitions to include ex parte orders.
In California, domestic violence restraining order respondents must surrender their firearms to a law enforcement agency or sell them to a licensed firearms retailer within 24 hours after the order is served. They must file a receipt with the court documenting compliance within 48 hours. Since 2007, respondents have been required to surrender their firearms immediately if a law enforcement officer makes a demand for them.12
Firearm prohibitions for domestic violence offenders, beyond the prohibition on purchases from licensed retailers,13 have rarely been enforced.14,15 In this study, we report a process assessment of a pilot initiative in San Mateo County and Butte County, California, during which domestic violence restraining order respondents were screened for firearms ownership or possession and an effort was made, at the time restraining orders were served or soon thereafter, to recover firearms from respondents believed to have them.
METHODS
The study population comprised all respondents to domestic violence restraining orders issued between May 2007 and June 2010 in San Mateo County and between April 2008 and June 2010 in Butte County.
San Mateo County, south of San Francisco, had a population of 718 451 in 2010: 42.3% non-Hispanic White, 25.4% Hispanic or Latino, 24.8% Asian, 2.8% Black, and 4.7% other, with 98.1% living in census-defined urban areas.16,17 The 2007–2011 median household income, $87 633, was 42.2% higher than that for California as a whole.16 The county reported 2766 domestic violence calls for assistance in 2010.18 The county sheriff had primary jurisdiction over unincorporated areas of the county; 22 other jurisdictions had their own law enforcement agencies.
Butte County, in central Northern California, had a population of 220 000 in 2010: 75.2% non-Hispanic White, 14.1% Hispanic or Latino, 4.1% Asian, 1.4% Black, and 5.2% other, with 81.1% living in urban areas.16,17 The 2007–2011 median household income, $42 971, was 30.3% lower than that for the state as a whole.16 The county reported 1098 domestic violence calls for assistance in 2010.18
The evaluators also served as advisors during the design of the intervention, but law enforcement personnel made all decisions regarding design and execution. Evaluators and detectives assigned to the initiative developed a detailed data form for cases involving firearms. We tabulated and reviewed the data with each county monthly, discussing ambiguous cases and outliers. San Mateo detectives recorded each order sent to the sheriff’s office, compiled identifiers for respondents, and provided those data to us. They completed the data form when firearms were recovered by the sheriff’s office and in selected cases when recovery was performed by a local police department. (In those cases, detailed firearm information was generally not available.)
In Butte County, neither the detectives nor the court kept counts of restraining orders or respondents. We estimated the number of orders detectives reviewed to be the court’s estimate of the average number issued each month. Detectives completed the data form when they served an order, whether firearms were recovered or not, and when they recovered firearms by other means: at follow-up visits after orders had been served by private parties or when respondents brought firearms voluntarily to the sheriff’s office.
We summarized the results using descriptive measures, presenting separate results for each county. We calculated the percentage of San Mateo County respondents who were linked to firearms on the basis of information for all respondents. For Butte County, we derived this calculation from information from the data forms. We used SAS version 9.1.3 (SAS Institute, Cary, NC) to conduct the analysis.
RESULTS
The initiative and study period began in May 2007 in San Mateo County and April 2008 in Butte County and ended in June 2010. San Mateo County detectives reviewed 6024 restraining orders involving 2973 respondents. Butte County detectives reviewed an estimated 1978 restraining orders on an unknown number of respondents.
Operations of the Initiative
Two detectives in each county, who were assigned solely to the initiative, received copies of restraining orders from their county courts. To screen respondents for links to firearms, they reviewed petitioners’ restraining order applications—these included a check box to indicate that firearms were involved, and client advocates encouraged petitioners to provide specifics. A form with photographs of representative firearms was developed to facilitate reporting. Detectives determined whether respondents had records in California’s Automated Firearms System, which contains records for handgun purchases and denied purchases of firearms of all types, both since 1996; assault weapon registrations; and concealed weapon permit applications. They searched other relevant California Department of Justice databases. When they judged it necessary, they interviewed petitioners. Information linking respondents to firearms ranged from specific knowledge derived from records in the Automated Firearms System or a petitioner's detailed eyewitness report to a petitioner's report that a respondent had mentioned buying or having access to a firearm.
When private parties served orders on behalf of San Mateo County petitioners, detectives usually received notice after the fact when the court received proof of service. If respondents were linked to firearms, detectives contacted them to explain the prohibition and their options for compliance. If respondents acknowledged having firearms, the detectives, sheriff’s deputies, or local police officers took custody of the weapons or facilitated their sale to a retailer.
If a San Mateo County petitioner requested law enforcement service, this was usually performed by deputies in the sheriff’s civil division. Detectives advised the deputies beforehand whether the respondent had been linked to firearms. Civil deputies explained the prohibition when serving the order, but their scope of practice did not allow them to recover firearms. If there were firearms at the scene, deputies were expected to wait for a detective or another officer to recover them. Because of the deputies’ heavy workload, waiting was not always possible. Service by the detectives themselves was generally limited to respondents who were in custody.
In Butte County, the initiative’s detectives served the order if the respondent was linked to firearms and the petitioner requested law enforcement service. They asked respondents if they possessed firearms, reviewed the terms of the order and the firearms prohibition, and explained the respondents’ options. If a private party served an order on a respondent who was linked to firearms, the detectives conducted a “knock and talk” to explain the prohibition and compliance options. They did not routinely make a demand for firearms, preserving the option of sale to a retailer.
Firearms in law enforcement custody were stored by agency property departments. Respondents could retrieve firearms after restraining orders expired if no other orders were active and background checks found them not to be prohibited for other reasons.
Respondents often denied possessing firearms. They could be asked to certify this under penalty of perjury but could not be compelled to do so and faced no penalty for refusing. Corroborating evidence for such certifications was not required until early 2009. Search warrants were generally not used in San Mateo County before January 2010, when new state statutes took effect.19 In Butte County, where a judge was closely involved with the initiative, detectives could obtain search warrants when evidence supported their use.
Screening Results
A link to firearms was made for 525 San Mateo County respondents (17.7% overall, 19.7% for men, 8.3% for women) and 88 in Butte County (31.1% overall, 33.3% for men, 16.3% for women; Figure 1). With 1 exception, each source of information identified less than half the respondents who were linked to firearms. Only 115 San Mateo County respondents (21.9%) and 31 Butte County respondents (35.2%) were linked to firearms by multiple sources, and only 18 (3.4%) in San Mateo County and 9 (10.2%) in Butte County were linked by all 3.
Restraining orders processed, firearms identified, and source of firearms information in (a) San Mateo County, CA, May 2007–June 2010, and (b) Butte County, CA, April 2008–June 2010.
Note. AFS = Automated Firearms System; CPO = criminal protective order; EPO = emergency protective order; OAH = order issued after a hearing; TRO = temporary restraining order. Of 525 respondents linked to firearms in San Mateo County, 236 (45.0%) were identified by AFS, 220 (41.9%) by the petitioner declaration or other court documents, and 180 (34.3%) by interviews with petitioners. Of 88 such respondents in Butte County, 45 (51.1%) were identified by AFS, 42 (47.7%) by the petitioner declaration or other court documents, and 28 (31.8%) by interviews with petitioners.
For 22 respondents (4.2%) in San Mateo County and 13 respondents (14.8%) in Butte County, a link to firearms was made only outside the screening process, such as by police departments or respondents themselves.
Firearm Recoveries
Firearms were recovered from 119 (22.7%) respondents who were linked to firearms in San Mateo County and 45 (51.1%) in Butte County (Figure 2). For 29 (24.4%) of these respondents in San Mateo County and 21 (46.7%) in Butte County, firearms were recovered that were not recorded in the Automated Firearms System.
Firearm recoveries in (a) San Mateo County, CA, May 2007–June 2010, and (b) Butte County, CA, April 2008–June 2010.
Note. A single firearm was recovered from 44 (37.0%) of 119 San Mateo County respondents from whom any firearms were recovered, 2–3 firearms from 35 (29.4%), and 4 or more firearms from 40 (33.6%). One firearm was recovered from 15 (33.3%) of 45 Butte County respondents from whom any firearms were recovered, 2–3 firearms from 9 (20.0%), and 4 or more firearms from 21 (46.7%).
Altogether, 665 firearms were recovered: 305 handguns, 291 long guns, and 69 of unknown type. Of the handguns, 109 (35.7%) were not in the Automated Firearms System. Maximum recoveries in individual cases were 21 firearms in San Mateo County and 36 in Butte County. The mean number recovered, among cases with any firearms recovered, was 3.4 in San Mateo County and 5.8 in Butte County. Almost all recovered firearms (622 of 665, 93.5%) were taken into custody by law enforcement agencies; respondents sold the remainder to licensed retailers.
The likelihood of firearm recovery varied with the source of information linking the respondent to firearms and with the method of service (Table 1). In San Mateo County, the most common cause of nonrecovery from a respondent who had been linked to firearms was that the order was never served and the prohibition never took effect (162 respondents, 39.9%; Table 2). In Butte County, detailed data were largely limited to cases in which the order was served; nonrecovery most commonly occurred when the respondent claimed or was determined not to possess firearms (46.5% of nonrecoveries). Such cases were also common in San Mateo County (Table 2).
TABLE 1—
Respondents With Firearms Recovered by Source of Information Linking Them to Firearms and Method of Service: San Mateo County, CA, May 2007–June 2010; Butte County, CA, April 2008–June 2010
| Respondents With Firearms Recovered | ||
| Characteristics | San Mateo County, No. (%)a | Butte County, No. (%)a |
| Source of information | ||
| Automated Firearms System only | 42 (40.8) | 8 (44.4) |
| AFS and declaration | 17 (89.7) | 7 (77.8) |
| AFS and contact with petitioner | 17 (48.6) | 4 (57.1) |
| Petitioner declaration only | 17 (17.7) | 6 (35.3) |
| Declaration and contact with petitioner | 0 (0.0) | 1 (25.0) |
| Contact with petitioner only | 0 (0.0) | 1 (12.5) |
| All 3 sources | 15 (88.2) | 5 (62.5) |
| Other | 11 (57.9) | 13 (100.0) |
| Method of serviceb | ||
| Civil deputies | 43 (44.3) | NA |
| Detectives or in court | 19 (20.4) | NA |
| Private party | 14 (63.6) | NA |
| Unrecorded | 43 (28.9) | NA |
Note. AFS = Automated Firearms System; NA = not available. Data are limited to cases in which orders were served.
TABLE 2—
Reasons for Nonrecovery of Firearms for Respondents Who Were Linked to Firearms: San Mateo County, CA, May 2007–June 2010; Butte County, CA, April 2008–June 2010
| Reason for Nonrecovery | San Mateo County (n = 406), No. (%) | Butte County (n = 43), No. (%) |
| Order issued but not served | 162 (39.9) | NA |
| Protected party claims possible firearms possession but not found in AFS | 100 (24.6) | NA |
| Respondent determined to be no longer in possession | 66 (16.3) | 9 (20.9) |
| Respondent denies possession | 21 (5.2) | 11 (25.6) |
| Firearms hidden or in another county, or searched for but not found | 24 (5.9) | 13 (30.2) |
| Order denied by judge or expired before service (therefore not served) | 0 (0.0) | 4 (9.3) |
| Respondent in jail or deceased | 6 (1.5) | 2 (4.7) |
| Protected party has possession | 3 (0.7) | 0 (0.0) |
| Other | 6 (1.5) | 4.0 (9.3) |
| Unknown | 18 (4.4) | NA |
Note. AFS = Automated Firearms System; NA = not available.
Firearm recoveries proceeded largely without incident. In 73 San Mateo County cases with detailed data, detectives rated the respondent as cooperative in 69 (94.5%). There was no report of violence or threatened violence, and no arrests were made. In Butte County, 2 of 45 cases involved violence, 3 involved threats, and another 5 respondents were considered uncooperative. Five cases involved an arrest, and 1 resulted in a minor injury to an officer. Once in San Mateo County and 4 times in Butte County, detectives reported that enforcing the firearms prohibition increased the risk they faced when serving an order. These cases involved initial refusals to surrender multiple firearms or threats against officers.
The counties did not provide information on firearm retrievals once orders had expired.
DISCUSSION
This evaluation suggests that it is possible to enforce prohibitions on firearm possession among persons subject to domestic violence restraining orders. Hundreds of firearms were recovered without significant adverse events. It is not clear whether such efforts reduce respondents’ risk for criminal activity.20,21 Studies of prohibitions on firearm purchases by domestic violence offenders,22–24 felons,25 violent misdemeanants,26 and persons with serious mental illness27 have yielded positive results.
These results parallel those for other programs that recover firearms from prohibited persons. Federal agencies intervene more than 3000 times per year when prohibited persons purchase firearms because the 3-day waiting period expires before the background check is completed.28 Since its inception in late 2006, the California Department of Justice’s Armed and Prohibited Persons System has recovered more than 10 000 firearms from legal purchasers who subsequently became prohibited persons.29,30 Neither program has reported significant violence associated with firearm recoveries.
Policies that other states might adopt facilitated the implementation of this initiative. These include prohibiting the purchase and the possession of firearms by restraining order respondents, requiring that this prohibition be included in restraining orders, and requiring that firearms be transferred or surrendered within 24 hours of service of the order or immediately if requested by a law enforcement officer. Evidence of compliance should be required.
Other complementary policies help prevent firearm purchase and possession by prohibited persons. California’s archive of firearms transactions facilitated screening, and its sensitivity will increase when rifle and shotgun transactions are added in 2014. Firearm recovery was particularly likely when this archive linked respondents to firearms. The background check and recordkeeping requirements for firearm transfers by licensed retailers provide benefits when extended to transfers by private parties.31,32
We identified several opportunities for improvement. Screening missed respondents who were linked to firearms by other means, and others with firearms were likely never identified. The undercount may be small, however, as illustrated for men. In 2004, 26.0% of men in California lived in a household with firearms.33 Nationwide, 88.3% of men in households with firearms nationwide own firearms themselves,34 from which the estimated prevalence of ownership among California men is 23.0% (26.0 × 0.883). Screening linked 19.7% of male respondents to firearms in urban San Mateo County, where below average firearm ownership might be expected, and 33% in Butte County, where firearm ownership was reportedly common.
In San Mateo County, at least, restraining orders were often never served; this was the chief reason firearms were not recovered from respondents who were thought to have them. High rates of nonservice are common.35–37 Some instances reflect a petitioner’s decision not to proceed, but procedural failures also occur for many reasons.
Civil deputies did not recover firearms and, because of time constraints, could not always wait until other officers arrived. The unintended effect of this practice may sometimes have been to provide respondents an opportunity to dispose of firearms before they could be recovered. When firearms are involved, personnel who can recover firearms immediately should serve domestic violence restraining orders. San Mateo County, which has continued its intervention beyond the study period, now emphasizes service by detectives.38
Denials of possession are problematic. Beginning in 2009, the California Department of Justice stopped accepting certifications of nonpossession unless corroborating evidence was provided. Search warrants should be available when a respondent believed to possess firearms does not surrender them.19
Future enforcement efforts in California or elsewhere will need to tap all available sources of information to screen respondents for links to firearms. The widespread lack of data needed for such screening currently limits the generalizability of this initiative; only 10 states archive any firearm purchase or permit-to-purchase records for 10 years or longer.39 Having victim advocates encourage petitioners to provide firearm information will be particularly helpful in states without transaction records. In Wisconsin, one such state, low rates of firearm recovery during a similar initiative were considered surprising.40
Our evaluation itself is subject to several important limitations. This is a small-scale, pilot initiative. Important process data were not available, and we do not present a cost analysis. Larger-scale replications will be needed to determine optimal procedures for screening and firearm recovery, the incidence of adverse events, and effects on rates of violence.
Acknowledgments
This study was supported by contracts with the California Department of Justice (CalDOJ; contract 06-7816).
We gratefully acknowledge the assistance of Vanessa McHenry of the Violence Prevention Research Program, sheriff’s detectives, and others in San Mateo and Butte counties who implemented the initiatives, as well as the participating staff of the Bureau of Firearms and the Crime and Violence Prevention Center at CalDOJ.
Human Participant Protection
This study was reviewed and approved by the University of California, Davis institutional review board.
References
Articles from American Journal of Public Health are provided here courtesy of American Public Health Association
Facebook
Twitter
Google+