Abstract
When your residential program is full, finding emergency shelter for human trafficking and domestic violence survivors can be challenging. Advocates often spend hours calling one agency after another to locate an appropriate placement. The Safe Shelter Collaborative is an innovative technological approach to locating emergency shelter. By sending out a blast alert to certified users within a geographic network, agencies provide faster access for a greater number and diversity of survivors. This article reports on an evaluation of the Safe Shelter Collaborative and the ways it has transformed the process of finding emergency shelter.
Locating emergency shelter for human trafficking survivors can be challenging. In addition to a critical bed shortage for designated shelter space for these survivors, there are a number of systemic and technological barriers to securing appropriate and safe housing. This process can be time-consuming and frustrating for agencies that have residential facilities that are at capacity and for nonresidential organizations advocating for emergency shelter. To locate appropriate emergency shelter, advocates and other agency staff must telephone multiple agencies over many hours until a bed is identified and secured. With each phone call, it is often necessary to repeat the same case information to establish service eligibility until a placement match can be secured.
The Safe Shelter Collaborative was initially developed to address the current barriers to securing emergency shelter for survivors of all forms of human trafficking. It is a technology service that helps participating agencies identify emergency shelter in two ways: (a) locating potential beds that are currently available within their geographic network of participating organizations, and (b) providing a means to directly solicit donations from their individual donors to help cover the cost of a hotel placement for a survivor. (This article only discusses the first feature: locating potential beds that are currently available within their geographic network of participating organizations. At the time of authorship, this feature was more fully developed and adopted by agencies participating in the Safe Shelter Collaborative.)
In April 2014, the Partnership for Freedom (2017), “a public-private partnership dedicated to promoting innovative solutions to end modern day slavery in the US and around the world” awarded US$1.17 million to Caravan Studios, a division of TechSoup via a grant managed by Humanity United (www.partnershipforfeedom.org). TechSoup was founded in 1987 and is the largest nonprofit provider of technology assistance services to nongovernmental organizations and libraries worldwide. Caravan Studios, an award-winning division within TechSoup, builds apps that help communities organize, access, and apply resources to their most pressing problems. At a Partnership for Freedom conference for prospective proposals, Caravan Studios met representatives of Polaris who were also seeking to address housing shortages for human trafficking survivors. The two organizations decided to work together, and a partnership between Caravan Studios and the Washington, D.C. and New Jersey offices of Polaris was born. Polaris is an international organization that operates the National Human Trafficking Hotline, a multilingual 24/7 hotline and text line that connects human trafficking survivors with needed services. Their office in New Jersey provided direct services to trafficking survivors. During the project period, Polaris closed its New Jersey office, and the Polaris staff founded the Sanar Wellness Institute, which continued their role within the collaboration. The Sanar Wellness Institute provides transformational healing services for survivors of human trafficking and individuals and communities vulnerable to exploitation. The New Jersey Department of Children and Families also participated in the project for outreach to agencies providing residential services to minors. The Safe Shelter Collaborative was piloted in New Jersey and expanded to the San Francisco Bay Area and North Texas. The purpose of this article is to provide an overview of the project and present some results of the exploratory evaluation of the three pilot sites.
Background and Significance
Human trafficking is defined by the Victims of Trafficking and Violence Protection Act (TVPA) of 2000 as, (a) the recruitment, harboring, transporting, supplying, or obtaining a person or labor or services through the use of force, fraud, or coercion for the purpose of involuntary servitude or slavery; or (b) sex trafficking in which a commercial sex act is induced by force, fraud, or coercion, or in which the person induced to perform sex acts is under 18 years of age. (TVPA, 2000)
Precise data regarding the prevalence of human trafficking is difficult to determine. The International Labour Organization (2012) estimates that nearly 21 million people are victims of forced labor globally. Moreover, a study from the World Health Organization shows that one in three women will be affected by some form of gender-based violence, including human trafficking, during their lifetimes (World Health Organisation [WHO], Department of Reproductive Health and Research, London School of Hygiene and Tropical Medicine, South African Medical Research Council, 2013). It is estimated that 12 million individuals are trafficked at any time (U.S. Department of State, 2007). In 2016, the Polaris Hotline received 8,042 calls or texts regarding U.S. cases of human trafficking, a 35% jump over 2015 (Polaris, 2017). In addition, the National Center for Missing and Exploited Children estimates that of the 18,500 runaways in 2016, one in six or 3,083 children were victims of sex trafficking (National Center for Missing and Exploited Children, 2017).
Psychological trauma resulting from these forms of abuse is frequently multigenerational and, without treatment, often results in further incidents of escalating violence (Lalor & McElvaney, 2010). Mental health symptoms including those associated with depression, anxiety, and posttraumatic stress disorder (PTSD) occur frequently in survivors, regardless of the type of exploitation to which they have been subjected (Katona, Robjant, Shapcott, & Witkin, 2015). The International Society for Traumatic Stress (2015) found that young people impacted by violence are anywhere from 2-13 times more likely to experience continued abuse throughout their adolescence and adulthood.
Advocates and researchers have identified a pressing need for additional emergency housing for human trafficking survivors (Logan, Walker, & Hunt, 2009; President’s Interagency Task Force to Monitor and Combat Trafficking in Persons, 2013). Although the covert nature of human trafficking makes it impossible to determine the exact number of beds needed for survivors per night, Polaris estimates the number to be in the tens of thousands. The number of available beds per night nationwide falls significantly short: Only 529 beds are exclusively designated for human trafficking survivors with an additional 1,115 beds available at organizations that support other population groups in need, thus limiting the actual number of beds available to human trafficking survivors (National Human Trafficking Resource Center, 2012). Nationally, there are a total of 33 residential programs aimed specifically at trafficking victims, available in only 16 states (Reichert & Sylwestrzak, 2013).
Furthermore, certain population groups, such as men and lesbian, gay, bisexual, transgender, queer, and intersex (LGBTQI) youth, have virtually no access to appropriate shelter placements. In many regions, any shelter specific to human trafficking survivors is scant to nonexistent despite a high level of need. For example, of the survivors served by Polaris’ New Jersey office from 2011-2013, approximately 70% needed shelter placement assistance, yet there were no designated human trafficking residential programs located in New Jersey (Reichert & Sylwestrzak, 2013). The Safe Shelter Collaborative addresses this need via technology that assists advocates in finding emergency shelter in a safe, survivor-sensitive and efficient manner.
Project Activities
During the grant period, Caravan Studios acted as the central entity responsible for technology design, development, and implementation. The Polaris New Jersey/Sanar office led the partner organization outreach, development of training and technical assistance for trauma-informed services to ensure all placements were appropriate for survivors, and identified unique shelter needs and eligibility requirements for survivors of all forms of human trafficking. Sanar staff was responsible for recruiting agencies to participate in the New Jersey pilot project.
Agencies from the New Jersey area who were already providing residential shelter services were recruited to participate in the project. As the majority of these agencies primarily focused on serving domestic and sexual violence survivors, they were the most likely organizations to expand their services to house and assist human trafficking survivors as well. Since domestic violence shelters were also in need of an efficient way to place survivors they did not have room for, the project was expanded to include finding emergency placements for domestic violence survivors as well. In addition to domestic violence programs, efforts were made to specifically recruit organizations providing services to homeless and LGBTQ youth, as traditional domestic violence shelters may not be able to accept underage minors in their shelter facilities. A total of 19 agencies were recruited for the New Jersey pilot. The San Francisco Bay Area added 33 agencies, and there were eight agencies certified in North Texas. The Bay Area was chosen for expansion due to anticipation for increased sex trafficking during the 2016 Super Bowl as well as for proximity to Caravan Studios headquarters. The North Texas area (Dallas & Denton counties) was selected for expansion due to participation in previous Caravan Studios initiatives.
Human Trafficking Training
To enhance the knowledge and skills of domestic violence advocates and at-risk youth agencies working with human trafficking survivors, Sanar provided participating agencies with training and technical assistance on best practices to support survivors of human trafficking. An initial 2-day face-to-face training addressed such issues as human trafficking definitions; networks and tactics; indicators and assessments; legal remedies, including criminal justice and immigration; developing partnerships with law enforcement and other service providers; the impact of trauma and trauma-informed interviewing; cultural competency, sensitivity, and language; case management services; housing; and empowerment programs. Once representatives attended the training, the agency became certified to use the Safe Shelter Collaborative. To support member organizations throughout the project implementation period, Sanar also conducted eight, 1.5-hr live webinars. Series topics included commercial sexual exploitation of children; immigration relief and legal support; working with survivors with substance abuse issues; working with federal, state, and local law enforcement; and vicarious trauma.
Development of the Sharable Eligibility Assessment Tool
To help decrease the number of times survivors have to repeat information about their circumstances, a sharable intake assessment tool called the “Common Eligibility Assessment Tool” (CEAT) was incorporated into the online technology. The CEAT was developed in collaboration with Polaris New Jersey, Caravan Studios, and eight New Jersey agencies that were identified as key stakeholders participating in the initial pilot stage of the project. Each of the variables included in the CEAT was identified as necessary information for shelter personnel to determine eligibility, space availability, and individualized service needs.
Once the CEAT was integrated into the Safe Shelter Collaborative technology, the online technology was used to simultaneously send a text message, email, or phone call that acted as a shelter placement alert to certified users within the geographic network. The shelter placement alert can be sent via desktop, computer tablets, landline phones, and mobile phones to meet the unique needs of each participating Safe Shelter Collaborative agency.
Program Evaluation
In April 2016, 2 years after the project began, the project conducted an exploratory evaluation of the Safe Shelter Collaborative. With only three months before an evaluation was due to the funders, a post-test-only design was used. Post-test only designs are limited to assessing what has occurred after an intervention has been implemented. The initial evaluation questions were as follows: (a) Has there been an increase in access to potential emergency beds? and (b) Has the Safe Shelter Collaborative transformed the process of finding emergency shelter? The evaluation covers the period of May 7, 2015 to August 9, 2016. However, since data collection regarding placement dispositions did not begin until October 2015, there was less than a year of data available on that important variable.
Data Collection and Analysis
Both administrative data and a user feedback survey were used for data collection. Administrative data includes information automatically collected each time an agency uses the Safe Shelter Collaborative to find a shelter bed. Data collected includes the initiating agency, demographics of the survivor as collected by the CEAT, date and time of the request, time to first positive response, and placement disposition. After a housing alert, the first agency to reply with a possible placement is considered the “time to first positive response.” This is coded as a “maybe” and signals that a potential bed may be available. It is the responsibility of the originating agency to contact the responding agency directly to discuss whether the potential placement is compatible with the survivor’s needs and to make transfer plans. With regard to disposition of placement requests, users are responsible for updating their housing requests with information regarding whether a placement had been made or if no placement had been made. There are four options: “This request resulted in a successful placement,” “Referral information was given to the client,” “An appropriate referral could not be made,” and “Other” (with a required input box).
CEAT demographic variables included age (18-24, 25 and older, under 18 and legally emancipated), type of survivor (sex or labor trafficking, or domestic violence), gender identity (female, male, or transgender), number of accompanying children or relatives, language, physical or intellectual disability, substance use (recreational use only; Diagnostic and Statistical Manual of Mental Disorders [5th ed.; DSM-5; American Psychiatric Association, 2013] diagnosis; or use while trafficked), immigration status (foreign national, U.S. citizen or naturalized U.S. citizen), government identification (birth certificate, passport, state identification card, social security card), and history of violence usage (self-injury, aggression toward others, suicide attempts). Questions about disability, substance use, violence usage, and immigration status and governmental identification were used to assess the individual service needs of the survivor. Other questions, such as the number of accompanying children or relatives, helped identify space needs. The project team decided not to include information about race/ethnicity and sexual orientation in the CEAT. This decision was based on the belief that an individual’s race or sexual orientation should not serve as a barrier or a facilitating factor in deciding to provide services. Data analysis was conducted using Excel or SOFA, an online data analysis software.
The second data collection strategy involved an online user feedback survey with certified Safe Shelter Collaborative users. The majority of users were frontline or program directors at agencies that provide residential services and have responsibility for requesting emergency shelter for survivors when their facilities were full. The purpose of the user feedback survey was to explore overall satisfaction with using the Safe Shelter Collaborative, the impact participating in the project has had on agencies and direct service personnel, and to solicit some anecdotal evidence of how the tool has made a difference in providing emergency shelter services to survivors. For example, to understand how using the Safe Shelter Collaborative might have transformed the process of identifying shelter, users were asked two questions: How long did it take to find housing before the Safe Shelter Collaborative, and How long did it take to find housing after use of the Safe Shelter Collaborative began?
To understand how participating in the Safe Shelter Collaborative had impacted their agencies, respondents were asked to check off the types of service expansion related to human trafficking that their organization undertook as a result of participating in this project. There were six potential service expansions available for selection. This included receiving additional training for working with human trafficking survivors, joining established human trafficking community task forces, adopting agency intake procedures/protocol to identify trafficking survivors, adding trafficking survivors to the agency’s list of populations served, applying for funding specifically for human trafficking services, and adding human trafficking to the agency’s mission statement or goals and objectives. Respondents were able to check all categories that applied to them.
The survey was developed and deployed online using the Google Forms application. A list of 92 certified users and their email contact information was provided by Caravan Studios and reviewed by Sanar personnel. As an incentive for participation, respondents were offered an opportunity for inclusion in a random drawing for four US$25.00 gift cards. Over a 10-day period, with frequent email reminders, a 46% (n = 42) return rate was achieved.
Findings: Administrative Data
Administrative data collected for the project period of May 7, 2015 to August 9, 2016 showed 78 requests for emergency shelter sent out on the Safe Shelter Collaborative. Of the 78, 40 were domestic violence survivors, 34 were sex trafficking survivors, and three were labor trafficking survivors. One was submitted as other; however, specifics were not included. The majority of requests came from the New Jersey region (70%, n = 55), followed by North Texas (17%, n = 13), and the San Francisco Bay Area (13%, n = 10).
CEAT Demographics
The following demographic information was collected through the CEAT data included on the emergency shelter request.
Age, gender identity, and type of survivor
The survivors needing emergency shelter were most often women (91%), 25 years of age or older (54%), who were escaping domestic violence (51%). This description applies to the individuals assisted in the San Francisco Bay area and North Texas. However, the New Jersey survivors were different. The typical survivor in New Jersey was female (93%) but younger, as 53% were between the ages of 18-24. Furthermore, 60% of New Jersey requests were for survivors of sex trafficking.
Accompanying children
Although there were 78 adult requests for emergency housing, 27 of these women (64%) were responsible for a total of 73 children and other family members with them. Therefore, the total number of persons for whom emergency housing was requested was 151.
Language
Overwhelmingly, across all regions, a high percentage (90%) of women were proficient in English. One user’s primary language was Tagalog, primarily spoken in the Philippines, and was reported in New Jersey. The Tagalog speaker was a female labor trafficking survivor.
Physical or mental disability
Only 10% (n = 8) had physical or intellectual disabilities. However, 26% (n = 20) had self-disclosed disabilities related to mental illness.
Substance use
Only 5% (n = 4) of survivors seeking help were considered as having a diagnosable substance abuse addiction. One individual indicated she used substances while trafficked.
Immigration status and government identification
Of the data collected from New Jersey, Texas, and California, the overwhelming majority of persons seeking shelter were U.S. citizens (78%); only four persons (5%) were foreign nationals. All of the foreign nationals were identified through New Jersey agencies. Of the four New Jersey foreign nationals, three were survivors of labor trafficking and one was a sex trafficking survivor. Three were primary Spanish speakers and one a native Tagalog speaker. Two were males and two were females. Seventeen percent (n = 13) were marked as “unclear” immigration status. Only 51% (n = 40) were able to disclose any type of government identification. Not surprisingly, 25 individuals (32%) were either not asked or agency personnel were not sure if the survivor had any ID. Thirteen individuals (17%) did not have any ID. Of all persons without any ID, 85% were in the New Jersey region. All but one of the North Texas domestic violence survivors had ID with them.
History of violence use
The overwhelmingly majority of survivors (94%) disclosed no history of violence usage. The two survivors with a history of aggression against others were offered emergency housing. Additionally, placements were found for the two survivors with previous suicide attempts and the one individual involved in self-inflicting injuries.
Table 1 provides another way to view the CEAT data as they highlight the differences and similarities found in the demographics by survivor type (trafficking or domestic violence). For example, sex trafficking survivors were more likely to be younger, were less likely to have access to government identification, and had fewer children. Domestic violence survivors were more likely to have children as well as a reported disability. There were also many similarities between both groups. Both groups had a high level of English proficiency and equivalent numbers of reported mental illness and substance use.
Characteristics of Survivors by Survivor Type.
Note. DSM-5 = Diagnostic and Statistical Manual of Mental Disorders (5th ed.; APA, 2013).
Time to Response
Table 2 demonstrates how quickly the response time to a first maybe was achieved. There were 59 valid responses recorded. Of these, 45 requests for emergency shelter received an encouraging “maybe” within 5 minutes or less (76%). Only five requests had to wait more than 20 minutes.
Time to First Maybe.
Disposition of Placement Requests
While there were 78 total requests sent out, only 45 (58%) of the requests were updated by the requesting agency. Although these data were not collected for the first 5 months of the project, the last month of data collection also had missing placement data. Table 3 highlights the responses by agency regarding the disposition of their request for emergency housing and highlights regional differences in dispositions reported. While only 29% (n = 13) resulted in a successful placement, another 33% (n = 15) were provided a referral for another housing resource. New Jersey agencies were responsible for all successful placements (n = 13) and for 12 of the 15 referrals. There were 13 “no placement made” dispositions in total, and they occurred across all three regions: five in New Jersey, two in the Bay Area, and six in North Texas.
Disposition of Placement Requests.
Findings: User Feedback Survey
Agencies providing services to domestic violence survivors exclusively or in combination with sexual violence services accounted for nearly 62% of all respondents to the survey. However, nine agencies (22%) provide services to human trafficking survivors either exclusively or in combination with other crimes such as domestic violence or sexual assault. Of the three responses marked other, one agency served refugees, one agency was described as a behavioral health organization, and one response was incomplete.
Overall Satisfaction
Respondents who used the Safe Shelter Collaborative were overwhelmingly satisfied (4.3 on a 5-point scale). Furthermore when asked whether they would recommend this feature to other agencies, 100% responded with a “yes.” With regard to improving the tool, the majority of responses (54%) called for an expansion of the number of agencies participating in the Safe Shelter Collaborative.
Freeing Up Valuable Advocate Time
In concurrence with the administrative data, survey users overwhelming believed that use of the Safe Shelter Collaborative saves valuable advocate time in locating housing. Respondents were asked two questions: “How long did it take to find housing before the Safe Shelter Collaborative?” and “How long did it take to find housing after they began using the Safe Shelter Collaborative?” Sixteen respondents answered both questions. Of these, 75% (n = 12) reported significant reductions in the amount of time it takes to find a placement. Only four responses indicated no change in time. No respondents reported that using the Safe Shelter Collaborative added to the amount of time for the process. The most common reduction (44%) was hours to minutes (n = 7). Three responses indicated a change from days to hours. Some responses indicated that the Safe Shelter Collaborative cut process time by 50-75%, such as “1 hour to 15 minutes,” or “2 hours to 1 hour.”
Service Expansion
There were 33 responses to the question of organizational service expansion related to human trafficking. Agencies have expanded their services to human trafficking survivors and expanded their networks for working with other agencies. The most common types of expansions included receiving additional training, joining established human trafficking community task forces and/or collaborations, and adding screening for human trafficking survivors to agency procedures or protocols. A total of 27 agencies expanded their services to be more inclusive of human trafficking survivors. Of those 27, 18 agencies made more than one change. Table 4 tracks the number of agencies who expanded their services by type of service expansion.
Type of Service Expansion.
In addition, the survey asked respondents whether participating in the Safe Shelter Collaborative introduced them to working with agencies that they had not worked with before. There were 39 responses to this question with just over half (n = 20) stating their agency was now working with new agencies. New Jersey agencies had the greatest number of expansions (n = 14).
Overall Impact
Three themes emerged from an open-ended question regarding the overall impact of using the Safe Shelter Collaborative on the agency’s ability to help survivors of human trafficking. The first theme identified the transformation of the process of finding emergency beds into a more efficient and time saving process (39%, n = 9). A second theme (26%, n = 6) addressed the issue of supporting the collaboration of agencies to find shelter. The third most often cited theme related to the expansion of skills and knowledge about human trafficking (26%, n = 6).
The second open-ended question asked for specific examples about finding emergency housing for a survivor as a result of participating in the project. The following quotes best reflect the possibilities that use of the tool provides: An agency reached out to us and we housed an human trafficking (HT) survivor during her wait for her Violence Against Women Act (VAWA) Visa.
The first survivor I placed using this tool was a male, foreign national, minor survivor of labor trafficking on the LGBTQI spectrum. This would have taken days to find successful placement in the past. As a result of the tool, we had secured a safe, appropriate, and trauma-informed bed for him in less than 10 minutes. A transgender client was able to do one intake and find shelter that was unheard of prior. We would have to call tons of shelters and ask if they took transgendered clients—With the training and the ease of the app our clients are not judged and accepted more openly.
Discussion and Implications for Practice and Future Research
The questions for the pilot evaluation were as follows: (a) Has access to potential emergency beds increased? and (b) Has the Safe Shelter Collaborative tool transformed the process of finding emergency shelter? The Safe Shelter Collaborative set out originally to expand access to available shelter beds for human trafficking survivors. However, in recognition that many available beds were located in domestic violence agencies, which from time to time also exceeded capacity, the project was expanded to include domestic violence survivors as well. Although no baseline data exist for the point of comparison, the number of agencies trained and certified for use of the Safe Shelter Collaborative tool appears to increase the opportunities for use of available beds. Participation in the project creates awareness of human trafficking and helps to identify survivors among agencies that had not previously served human trafficking survivors. The networks of agencies that participated in the Safe Shelter Collaborative assisted each other in maximizing the utilization of their already existing unused beds for trafficking survivors.
The process of finding emergency shelter has also been transformed by participating in the Safe Shelter Collaborative. Prior to using the Safe Shelter Collaborative, agencies needing to find emergency housing for abuse survivors relied on a “telephone model” whereby an advocate makes telephone calls from one agency to another until available space is identified. This process could take several hours until a successful referral is made. Other than the data collected for the Safe Shelter Collaborative User Survey, no other data currently exist that allow for comparisons in success rates between the telephone model and the Safe Shelter Collaborative. The administrative data and the user feedback survey concur that use of the Safe Shelter Collaborative greatly reduces the time necessary for finding a suitable placement.
Domestic violence workers have expressed an interest in using safe technology that enhances their work efficiency and protects the confidentiality of their clients (Murray, Pow, Chow, Nemati, & White, 2015). The Safe Shelter Collaborative technology design, with the incorporation of the CEAT designed by service providers, meets those standards.
Two primary issues challenged administrative data analysis. One challenge related to the absence of an evaluation perspective before the launch of the project. Such a perspective could have assisted in identification of the data to be collected, data definitions, data analysis planning, and periodic monitoring of the data to review responses and response rates. Identifying possible challenges to data collection from the beginning of a project is critically important, especially for a project such as this with potential national implications. Second, the high percentage of missing data further impedes the generalizability of the data and relegates the conclusions to a “shows promise,” but additional evaluations must be undertaken to see whether positive effects remain.
In addition to these issues, there was also a high number of “unclear” “don’t know” responses with regard to demographic data that require consideration for changes to data collection. For example, there was a high percentage of “unclear” responses regarding immigration status. Although this item was included as a way to identify a survivor’s potential legal issues, this question is uncomfortable for survivors worried about deportation and imprisonment and equally uncomfortable for agency personnel who do not want to jeopardize developing trust with survivors or their safety. Another less threatening way to ask this question should be explored. And evaluators should see data missing about immigration as a protective shield for survivors and not as a nuisance.
The disposition of placement requests also needs special attention. Despite 73 requests for housing, there were only 45 instances where the originating agency returned to the Safe Shelter Collaborative site to update the disposition of the request. Although a higher percentage of completed disposition data are desirable, it is important to place these missing data within the context of a busy emergency housing agency. Considering the crisis-oriented work conducted by residential programs for abuse survivors, it is understandable that an advocate may be too busy and/or may ultimately forget to input disposition information. In early 2017, to encourage agencies to add disposition data, Caravan Studios added a reminder alert. The alert is triggered after a specific amount of time has elapsed from the initial request for housing, and disposition data have not been entered. Besides the missing data, the reported data also bear examination. There were as many successful placements as unmet requests (n = 13), and there were more instances of referrals made (n = 15). First, there appears to be no plan to follow the survivor beyond a referral until a successful placement is made. Caravan Studios should explore the feasibility of following survivors until successful placements are made. Second, it is possible that at the time of the initial request for placement, emergency bed resources at Safe Shelter Collaborative member agencies were also experiencing high demands and therefore no placement was possible. The number of unfulfilled placements also concurs with research demonstrating shortages in existing housing resources for both human trafficking and domestic violence survivors (National Human Trafficking Resource Center 2012; National Network to End Domestic Violence, 2017).
Despite the rapid response to requests for emergency shelter, the system is only as good as the responses of the geographic network of agencies participating. When an alert is sent, how many of the agencies respond to each request? The “maybe” response encourages advocates to follow up directly with an agency that may have a bed available. A “no” response is also useful as it conveys two ideas: (a) We heard you, and (b) Sorry, we can’t meet this request. The absence of any response leaves a question as to whether the agency saw the request and has no beds available, or has not seen the request at all. The future success of the Safe Shelter Collaborative, therefore, is dependent upon the participation rates of agencies within each geographic network and should be monitored regularly.
There were some notable differences between New Jersey agencies and the North Texas and San Francisco Bay Area agencies that bear consideration. New Jersey agencies were more likely to identify human trafficking survivors and provide disposition of placement request data. They may have had a stronger investment in the project as they were the first of the regional networks to have specialized training in working with human trafficking survivors and were partners in the development of the CEAT. Because their participation periods have been longer than either of the other two networks, they benefit from more experience and training as well. Future evaluation efforts should pursue the factors that contributed to the success of the New Jersey regions and incorporate these concepts into future expansion efforts.
There are other noteworthy findings that might be considered for future research and practice. Contrary to usual and customary practices in which racial and sexual orientation data are collected for referral purposes, the developers of the CEAT consciously decided not to include that data so these demographics were not a hindrance to a successful referral. Interestingly, two respondents to the user feedback survey expressed appreciation for this notable omission.
One of the myths about human trafficking is that it is an international problem with foreign nationals being trafficked into the United States. The reality is that many U.S. citizens are survivors of human trafficking, and may have never left the United States. Only four persons (5%) of the 78 requests for emergency housing were foreign nationals.
With limited time available and engagement of an evaluator at the end of the project period, a post-test was the best available design for the circumstances. Therefore, there was no control or comparison group information to more rigorously evaluate these findings. The online user feedback survey only collected data from a single point in time. Survey completers may also have a strong interest in the topic, and therefore, results could reflect such biases. Additionally, not all respondents chose to respond to all questions, and the reasons for these skipped questions are unknown. This survey also targeted individuals who work at crisis-oriented organizations. Due to the crisis-oriented environment, it is possible that some persons who did not respond to the survey are more frequent users of the tools and did not have time to respond. It was not possible to target the survey to only one person who could represent their agency as several agencies had multiple certified users. Finally, the voices of survivors are not reflected in the program evaluation. Future evaluation strategies should include ways to engage survivors to determine in what ways they were able to benefit directly from receiving a quicker placement, and a possible decrease in the number of times they had to share their stories.
Conclusion
The Safe Shelter Collaborative has been training geographic networks of agencies that provide residential shelter services to use a technological tool to locate emergency safe housing for human trafficking survivors. These efforts appear to expand the number of agencies providing services to human trafficking survivors as well as access to available emergency shelter beds. Many of the certified agencies previously served domestic violence survivors, and this population group has been incorporated into the project as well.
Although this project has only been in operation in a limited number of communities, use of the Safe Shelter Collaborative has been positively embraced by agencies in these areas. Significant time reductions in the process of finding suitable placements were reported, saving valuable time for already busy advocates. Training on human trafficking and participation in the project led many agencies to expand their target population to include human trafficking survivors and join with other agencies in their communities to address this serious problem. As Caravan Studios expands access to different geographical locations, care should be given to address the issues identified by this pilot evaluation.
Footnotes
Acknowledgements
The authors wish to thank the following people for their support of this project: Danika Ali, Matthew Bradley, Gabriela Celeiro, Ashley Garrett, Sarah Jakiel, Meagan Jensen, Sophia Lane, Nicole Levine, Bradley Myles, Audrey Roofeh, Anne Snouck-Hurgronje, Cristal Solorio, Keeli Sorensen, and Ulana Tatunchak.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this project was provided by Partnership for Freedom. Caravan Studios/TechSoup provided funding for the evaluation.