Specialization and the Use of GPS for Domestic Violence by Pretrial Programs: Findings from a National Survey of U.S. Practitioners

ABSTRACT

Tools that facilitate the electronic monitoring of criminal justice populations are becoming widespread and multifaceted as they are adapted for a range of purposes and offender categories. In the past two decades, justice agencies across the United States have incorporated global positioning systems (GPS) to enforce no-contact orders in cases involving domestic violence (DV) or intimate partner violence (IPV). The current study surveyed a national (U.S.) sample of representatives (N = 114) from agencies administering pretrial programs that use GPS following DV-related charges. While all respondents are involved in using GPS for DV, analysis shows that some also use a range of other tools and monitor diverse portfolios of offenders; we report on relationships between the number of technologies used, populations monitored, attitudes, and practices. The article discusses the importance of giving due attention to the role of specialization in remotely supervising clients and providing them with services.

KEYWORDS:

• Domestic violence
• electronic monitoring
• global positioning systems
• national survey
• pretrial
• specialization
• supervision

View correction statement:

Erratum

Notes

For brevity, we indicate “DV,” but survey questions asked about DV and Intimate Partner Violence (“DV/IPV”).

Representing the “second generation” of EM-based technologies, GPS has become more prevalent in North America, Western Europe, Australia, and New Zealand (Gibbs & King, 2002, 2003; John Howard Society of Alberta, 2000, 2006; Mainprize, 1996; Nellis, 2014), and is being incorporated elsewhere, including Chile (Peillard, 2013), Argentina (Paterson & Clamp, 2014), and some Eastern European countries (Nellis, 2014).

GPS also has limitations; it is expensive, potentially creates more work for employees, has issues associated with satellite “drift” (cf. Erez et al., 2012, 2013), and cannot detect all contact attempts (e.g., via e-mail, text messages, or if the victim goes to meet the defendant).

More programs use risk assessment for other issues, for example, “substance use” (42%) and mental health (27%; PJI, 2009, p. 36).

Erwin’s (1990) prescient observation came at a time before the Internet, smartphones, and criminal justice “gadgets” (e.g., tasers, dashboard and body cameras, console-mounted computers) became prevalent; probation has presumably also been affected by these innovations.

Writing from a U.S.-based perspective, Burrell (2006) argues:

Offender supervision is a human capital intensive activity. There is no technological or automated solution to this problem. While technological innovations have certainly transformed the work of the PPO [probation and parole officer], they primarily have improved the monitoring capability of the officers and their access to information, but have done little to change the core correctional practices that comprise case management. People, in the form of PPOs are the core correctional resource (p. 2).

For example, Burrell (2006) reports adult caseloads standards for the APPA: intensive caseload, 20:1; moderate-to-high risk, 50:1; and low risk, 200:1. DeMichele and colleagues (2011) report the same figures, adding 1,000:1 or “No limit?” for administrative cases (p. 17). Compared to earlier guidelines (APPA, 1991), more recent distributions allocate fewer cases to those responsible for high and medium priority cases, and more to those tasked with low-priority cases.

While seven of 36 officers monitor more than 62 EM cases between them (roughly 8.9 cases each), 29 of 36 officers are responsible for only 78 EM cases (2.7 each; Bales et al., 2010, p. 84).

DeMichele and colleagues (2011) do not explore workload dynamics for those who use GPS/EM with pretrial clients facing DV charges.

Few studies have explored practitioner attitudes toward EM (cf. Clear & Latessa, 1993, pp. 461–462; Seiter & West, 2003, p. 75; Erez et al., 2012, p. 51). One early effort (Johnson et al., 1989) portrays local sentiments in a single jurisdiction, though the questions asked of practitioners were not geared toward documenting officers’ perceptions of EM’s impact on their performance. More recently, Bales and colleagues (2010, p. 109) report that only 2.9% of practitioners in one state thought EM was most effective for supervising DV offenders; this may be related to the early passage of sex offender legislation in that local context.

The survey represents one prong of a multimethod study of GPS for DV during pretrial (Erez et al., 2012); the other prongs consisted of, first, a quasi-experimental design across three U.S.-based sites measuring the impact of GPS on offender behavior in the short- and long-term, and, second, an in-depth interview component with participants—victims, defendants, justice personnel, and social service providers (N = 201)—drawn from programs at the three impact sites, and three additional U.S. locations.

While such high-profile events can be the impetus for the creation of such programs, only a minority of agencies (8.6%) had changed policies about the use of GPS because of a high profile incident reported in the media (e.g., newspaper, TV, local news), suggesting that once the programs are in place and become embedded in local agencies and practices, they become more difficult to change. Therefore, proper planning before establishment is key, including a pilot program or feasibility study (cf. Court Services and Offender Supervision Agency for the District of Columbia, n.d., p. 53; John Howard Society of Alberta, 2006; Lemke, 2009).

Not all respondents answered each item.

Only definitive responses (i.e., “Yes” or “No”) to items regarding victims were included in this analysis. On average, less than 10% of respondents did not know whether their program used these various features with victims.

The question choices for type of defendant were: domestic violence or intimate partner violence, other violence offense (not DV/IPV), sexual offense, DWI/DUI, other alcohol/drug offense, juveniles, gang members, and other.

The question choices for type of technology were: GPS, radio frequency, SCRAM, remote alcohol monitoring (not SCRAM), equipment-free monitoring, and other.

Considering how technologies were used across offense categories, on average, respondents monitored 4.9 of a possible seven offense categories with some technology. In this sample of respondents who all use GPS for DV, those using GPS apply it to an average of 4.4 offense types, SCRAM users average 3.3 offense types, yet some also use equipment-free monitoring with 5.1 offense categories. These averages show that respondents tend to have experience using technologies with multiple offense types; some technologies, like SCRAM and GPS, are used by some practitioners with fewer offense categories, while others are used with a broader range.

This group (n = 43) contains a subset of respondents (n = 5) who restrict use of GPS to the DV offense category. The others in the large group also rely only on GPS in cases of DV, but in addition use GPS for other offense categories.

Over 86% of respondents (N = 114) were either “very satisfied” or “somewhat satisfied,” so this analysis seeks to distinguish between the former and any other attitudes. “Other than very satisfied” respondents include those who were “somewhat satisfied,” “somewhat dissatisfied,” and “very dissatisfied”; five respondents indicated they “don’t know,” and were excluded from the analysis.

Available upon request.

For this calculation, “neutral” responses were combined with “negative” responses. These results do not hold if “neutral” responses are instead counted as “positive” responses. Given that the majority of respondents were positive in regards to EM throughout the survey, recoding neutral as negative was helpful in noticing small attitudinal or experiential differences.

Satterthwaite correction was used because of unequal variances. A parallel listwise analysis (N = 71) shows similar patterns, but is not reported here because it would omit smaller highly-specialized programs that do not report non-GPS caseloads because they monitor all defendants with GPS.

Satterthwaite correction was used because of unequal variances.

The measure is based on responses to the question: “What is your program’s schedule of operation for the interviewing staff? (e.g., 24 hr/day, Monday to Friday during court hours, 7 days/week, etc.).” Such a question is likely to result in responses that under report the total number of hours an agency is open each week, and hence provides a conservative estimate of agency hours of operation.