Coordinated Community Responses to Domestic Violence in Six Communities: Beyond the Justice System. Chapter 2. Study Design

The purpose of this study is to understand the approaches taken to coordinating responses to domestic violence in different communities, and how each community's system developed in response to client need, and in the context of other policy influences. Specifically, the study was designed to examine the following issues:

• How model systems of comprehensive and coordinated community-based domestic violence service delivery have developed in different communities;
• The legal, policy and community contexts in which the systems operate;
• The goals of the coordination efforts and approaches used to meet the goals;
• Barriers to coordination and how they have been addressed;
• Strengths and weaknesses of the coordination effort;
• The role of laws and policies in helping or hindering coordination efforts; and
• Issues in planning, implementing and evaluating coordinated services.

To address these issues, the study used a formative evaluation approach involving case studies of model community systems. The study was not intended to evaluate the effectiveness of comprehensive service delivery systems, although it offers some preliminary ideas about issues for future evaluation efforts. Six sites were selected for in-depth study: Baltimore, Maryland; Kansas City, Missouri; Carlton County and Northern St. Louis County, Minnesota; San Diego and San Francisco, California. This remainder of this chapter discusses the process used to select these sites and the procedures used to conduct the site visits.

• Site Selection

  • We developed a list of site selection criteria to fulfill the study objectives based on a review of the relevant literature and discussions with HHS and members of the project's advisory group composed of representatives of federal agencies with responsibilities in the area of domestic violence. The selection criteria were divided into two categories--those considered essential and those to be used for obtaining variation across the sites (Exhibit 2.1).

    Three characteristics were considered essential for this study. First, we wanted to look at coordination efforts that include agencies from different service systems, rather than a single agency providing comprehensive services (i.e, one-stop shopping approach). Communities in which a larger number of agencies coordinate their efforts are considered to represent more extensive and better-developed examples of a coordinated community response. This approach also has wider applicability than one that relies on a single agency to deliver comprehensive services, since a program of that type does not exist in many communities and would be hard to develop. Second, we wanted the coordination efforts to have been in existence for several years and to be relatively stable at the time of the study. This project is considering the history anddevelopment of a coordinated response, so we thought it was important to focus on communities whose efforts have been underway for some time. Also, because it is difficult to assess the features and outcomes of a coordinated response that is undergoing major changes, we wanted the effort to be relatively stable at the time of the study. Finally, we wanted sites that were able and willing to accommodate a four-day site visit by project staff during the study period.

    There are a number of different approaches to coordinating a response to domestic violence. In order to obtain information on different models, the selection criteria included five characteristics for variation across communities. These characteristics were selected based on the overall study objectives and specific HHS interests to obtain variation in: the lead agency, the model of interagency coordination, the location and population served, and the context of the coordination effort. There was also a particular interest in community responses that include health care providers as part of the efforts.

    Exhibit 2.1: Site Selection Criteria
    Essential Characteristics
    Cross-Agency Approach The coordination must involve multiple agencies (more than two), and represent different service systems (rather than all criminal justice agencies, for example). Stability of the Coordination Efforts The efforts must have been underway for some time, and be relatively stable at the time of the visit (i.e., not in a period of flux or major growth or decline). Willingness to Host a Site Visit The community must be able and willing to accommodate a four-day site visit by project staff during the study period.
    Characteristics for Variation
    Lead Agency Include different models of program leadership which vary by (1) the degree to which leadership is centralized (i.e., leadership for the coordination comes from one partner agency, rather than being shared by several agencies or equally distributed across all partner agencies), and (2) the type of agency that leads the effort. Model of Interagency Coordination Include communities that have developed different approaches and use different activities and mechanisms to coordinate their response. Location and Population Served Select communities with different geographic and demographic characteristics including: (1) at least one rural community; (2) one or more communities with large minority populations; (3) communities from different regions of the country; and (4) communities that vary in size. Context of the Coordination Effort Include communities with different environments for coordination (i.e., those where efforts operate within a context or support system favoring coordination in addition to those which represent initial attempts in a community with little coordination experience). Involvement of Health Care Providers Include communities in which health care providers are part of the coordinated response.

• Identifying the Sites

  • After developing the selection criteria, we compiled a list of potential sites using information from Family Violence: State-of-the-Art Court Programs (National Council of Juvenile & Family Court Judges, 1992) and consultation with knowledgeable people in the field. The publication provided program descriptions and contact information for model court programs dealing with family violence, which often include joint efforts with criminal justice agencies or domestic violence service providers. We also consulted HHS, the advisory group, and several domestic violence experts in completing this list. Based on these resources, we assembled an initial list of 16 different sites for possible inclusion in the study.

    We conducted a phone interview with a key person in each site to assess how well the community's efforts met the various selection criteria. The sites were then grouped according to their strength as candidates, and final sites were chosen in consultation with HHS and members of the federal advisory group.

    Six sites were selected for in-depth study: Baltimore, Maryland; Kansas City, Missouri; Carlton County and Northern St. Louis County, Minnesota; San Diego and San Francisco, California. Initially, Duluth, Minnesota had been selected as one of the study communities. However, since the Duluth model has been extensively studied, we decided to include two rural communities in Northern Minnesota (Carlton and Northern St. Louis Counties) whose response to domestic violence has been influenced by the Duluth model. All of the sites selected have established links between criminal justice agencies and service providers in the community and, in every site, the coordination effort has been in existence for at least five years (in some cases, for more than a decade).

    The selected sites provide a number of interesting examples of coordination efforts and contexts for the coordinated response. Baltimore is a large urban area with a significant African American population. It reflects a model that is dominated by one primary domestic violence service provider, the House of Ruth, which provides comprehensive services for battered women. There are also several examples of health care responses in Baltimore.

    Kansas City does not have a long tradition of interagency collaboration, but the community has made a number of significant changes in the criminal justice system, particularly in the courts. The city does not have a longstanding coordinating committee for domestic violence, but it has strong leadership on the issue from within the criminal justice system. Advocates are currently less active in Kansas City, although they heavily influenced the earlier changes in the criminal justice system. Two hospitals in Kansas City have recently developed on-site programs for battered women.

    Carlton and Northern St. Louis Counties provide examples of coordination in rural communities. Northern St. Louis County has a service network that has been in place and evolved since 1978. Carlton County includes the Fond du Lac Indian reservation and has a sizable Native American population.

    San Diego County has a well-developed coordinating council and several examples of interagency collaboration including a joint program between probation and child protection. The Children's Hospital operates the Family Violence Program which provides primary prevention and intervention services for battered women and their children.

    San Francisco has a long history of interagency collaboration on a number of issues including domestic violence. Advocacy efforts are particularly strong and the community features several coordinating committees with different objectives. The population is very diverse with large numbers of immigrants, non-English speaking persons and gays and lesbians. The city also was selected because of recent initiatives in the health care community.

• Site Visit Procedures

  • Information on the coordinated response was collected during in-depth site visits to each community. In each site, we conducted semi-structured interviews with key staff in relevant agencies and programs. An initial list of respondents was developed through discussions with the key contact person in each site. As we contacted each person, we asked them to identify additional people who were involved in the community's efforts. In each community, we interviewed people from the following agencies, programs and organizations:

    • Members of Domestic Violence Coordinating Council(s)
    • Battered Women's Services - Shelters, Counseling, Advocacy, Legal Services
    • Batterer Intervention Programs
    • Child Welfare Agencies
    • Criminal Justice Agencies -Police, Prosecutors (Felony and Misdemeanor)
    • Courts - Criminal Courts (Superior/District/Municipal), Civil Courts
    • Health Care - Hospitals, Community Health Centers, Healthy Start Programs
    • Other Services - Chemical Dependency and Mental Health Agencies

    We tried to identify links with programs that provide public assistance, housing assistance, employment and training services, and child support enforcement services. In a couple of cases, individual service providers had worked with these agencies on occasion. However, we did not find any well-developed links or routine coordination with these agencies in any of the communities that we visited.

    In most cases, we interviewed the person at each agency or organization who was most directly involved in the coordination efforts. In cases where more than one person from the agency was actively involved, we tried to interview staff at different levels who were involved either individually or in groups. However, given the time constraints, it was not always possible to interview both senior management and front-line staff from every program and agency.

    We also made an effort to schedule site visits to overlap with a meeting of the site's coordinating committee or other key events when possible. In two sites (Baltimore and Kansas City) we attended and observed a coordinating committee meeting during the site visit. In San Diego, we attended a full-day summit on violence against women that was also attended by many of the coordinating council members. We also observed the docket for protection order hearings in Kansas City and accompanied a child protection worker on home visits in San Diego. These activities provided useful information about each community's coordination efforts.