The objective of this demonstration evaluation was to rigorously examine the Domestic Violence Housing First model, which provides housing-inclusive advocacy and flexible funding to help survivors achieve safe and stable housing. Over 400 people who survived DV and were homeless or unstably housed participated in a quasi-experimental, longitudinal evaluation study that followed them over two years after they sought services from one of five participating DV agencies. Careful attention was paid during recruitment efforts to ensure that all eligible survivors were invited to participate in the study. Participants were interviewed (in English or Spanish) shortly after seeking services (baseline), and at 6-, 12-, 18- and 24-month follow-up. Data were also collected from advocates and agency records.

The three pillars of the Domestic Violence Housing First model:

1. Housing-inclusive advocacy: A critical component of the model is that advocates focus on addressing needs identified by survivors rather than on needs pre-determined by the agencies. Advocates are also mobile, meeting survivors where it is safe and convenient for them, and advocacy continues for as long as survivors need support.

2. Flexible financial assistance: Many survivors need not only advocacy to obtain safe and stable housing, but also temporary financial assistance to support themselves and their families. Funds are therefore targeted to support survivors so they can rebuild their lives, including covering transportation, school supplies, uniforms and permits required for employment, as well as time-limited and flexible rental assistance.

3. Community engagement: Advocates proactively engage people in the community who can help support the safety, stability, and well-being of survivors. Advocates engage with health care professionals, law enforcement and the legal systems, educators and school administrators, religious and spiritual leaders, and others. The evaluation design allowed us to examine the first two pillars of the model: housing-inclusive advocacy and flexible funding (all of the agencies engaged actively with their communities). While all of the participating agencies reported using the DVHF model, they acknowledged that due to limited resources (e.g., staff turnover, limited funds) it was often the case that survivors received “what was available at the time.” Similar to programs around the country, they may or may not be able to meet all of survivors’ needs. This quasi- experimental design allowed us to examine the model in “real world” conditions.

Retention across all study participants was high across all time points, from 92 percent at the 6-month follow-up to 89 percent at the 24-month follow-up. Thirty participants did not receive any services or funding within the first six months from the recruiting agency, and were removed from analyses. Sixty-four percent of the remaining sample had received DVHF, and 36% had received services as usual (SAU).