The Challenge

In January 2019, over 5,000 people were experiencing homelessness in Chicago. The majority of these individuals (4,030) were in shelters, where unfortunately the route back into stable housing was slow and often unsuccessful. The average length of stay in shelter was 111 days and only 39% of household exits were to permanent destinations such as their own rental unit or moving in with family permanently. For individuals and families, this extended stress and trauma from experiencing long-term homelessness could exacerbate many related challenges, including mental or physical health issues, substance use problems, or unemployment.

Long shelter stays were a problem for individuals and families experiencing the trauma of extended homelessness, and for the Chicago shelter system, which needed to manage its capacity to serve those in crisis. Current consensus for best supporting individuals experiencing homelessness focuses on taking a housing first approach, where homelessness (including shelter stays) should be rare, brief, and non-reoccurring.While the City of Chicago Department of Family and Support Services (DFSS) had little control over the availability of permanent housing, the department did fund approximately 49 shelter facilities (run by approximately 30 provider agencies) that differed in their ability to quickly and successfully rehouse individuals. When DFSS began working with the GPL, the department asked for help in figuring out how to improve rehousing supports for individuals experiencing homelessness while they are in shelter, while avoiding incentives for shelters to become long-term service delivery hubs.

DFSS and the GPL uncovered three key challenges:

• Quickly and stably rehousing clients was not the clear primary goal of the shelter system: Shelters didn’t have clarity on their priority goals and were working towards too many performance measures, contributing to shelters delivering a broad range of services and achieving low performance. For example, performance measures might have included the share of clients who: received a needs assessment, were connected to supportive services, participated in youth leadership development, accepted assistance to address needs, increased their cash benefits or income, were engaged in employment readiness programs, or had exited to stable housing. In addition, these performance measures were expressed without any indication of their relative priority level. DFSS was not confident that all of these targets were realistic or that it was measuring the right intermediate steps to assess progress towards high-level goals. With limited resources, it was critical for shelters to have a clear understanding of what activities and priority outcomes they should focus on. However, DFSS engagement and communication with shelters was primarily focused on ensuring compliance with funding requirements and maintaining the overall system capacity (i.e. total number of beds), with little attention remaining to tackle barriers to quickly and stably rehousing clients. DFSS recognized the need to reorient its shelters’ activities and communicate proactively to shelters to both clarify and work towards the primary goal of supporting families and individuals to quickly achieve safe and stable housing.

• Shelter activities were not consistently aligned with the needs of the specific populations served: Many of the city’s shelters housed clients with complex needs (ranging from domestic violence survivors, to families with young children, to veterans with medical vulnerabilities), but shelters varied greatly in the nature and intensity of support they provided. Even for shelters serving the same target population, there was significant variation across setting type (e.g. beds vs. cots, congregate vs. private rooms), type and intensity of services and programs offered, case management approaches (e.g. staff-to-client ratios, meeting frequency), goals (e.g. safe place to sleep vs. move to permanent housing), and resource allocation. While variation is not necessarily problematic, a client could potentially experience a drastically different level of support depending on which shelter they walked in to, regardless of their particular needs. In addition, a wide variety of service delivery practices across shelters made it very challenging for DFSS to interpret shelter performance and support the sharing of best practices.

• Shelters were spread thin, often relying on a broad span of supports delivered in-house to meet their clients’ needs: Over time, individual shelters had developed a broad range of solutions as client needs arose, leaving them with resources spread thin across many services and underutilizing external programs that could provide additional supports more efficiently and effectively. This shelter-based service delivery risked further institutionalizing shelters as a primary access point for supportive services, potentially contributing to extended stays. Shelters needed to shift from the direct provision of a broad, diverse set of services, to a focus on a targeted set of re-housing supports alongside a strong referral function for additional needed supports.

The Project

With GPL support, DFSS communicated an updated vision for shelter goals and critical activities across program models through a new procurement, supported shelters to improve on core internal rehousing activities, and identified priority referrals to support families experiencing homelessness in achieving re-housing goals.

As part of this, project partners:

1. Executed a procurement for shelter services that clarified rehousing as the primary goal, in order to create a unified spectrum of shelter program models

The DFSS Homeless Services Division team set out to redesign over 70% of the city’s overall  shelter system, elevating an explicit focus on rehousing clients quickly. To do so, DFSS executed a results-driven request for proposals (RFP) for its shelter portfolio, which communicated a shared, high-level goal for the overall shelter system to “provide a safe accessible place to stay for those experiencing homelessness and move them towards and into appropriate and stable permanent housing.” Key outcome metrics to assess progress towards this goal included the percentage of households exiting to permanent housing, and the average length of stay in shelter. These outcome metrics were elevated over the numerous performance targets characteristic of previous contracts, which were shared instead as potential supplemental measures of key activities, to be refined together over the course of the contracts. This helped lay the groundwork for ongoing collaboration between shelters and DFSS in order to sharpen their shared understanding around which shelter activities are most critical to achieving rehousing outcomes.

The procurement brought four previously separate shelter program models under one “Shelter” RFP, reframing the shelter service array as a spectrum of services that focused on meaningful differences in levels and types of services. Agencies, armed with this increased clarity around intended service intensity, core activities, and associated performance metrics, were asked to self-identify along the spectrum in their applications. This spectrum was defined by 3 levels of service intensity, with sub-models tailored to specific population

• Emergency shelters that focus on meeting the immediate needs of clients experiencing homelessness and providing connections to the broader homeless services system;
• Interim shelters that provide a variety of services and/or linkages dedicated to rehousing clients as quickly as possible; and
• Specialized interim shelters that can provide a set of specialized and/or more intensive services that address the needs of particular sub-populations (e.g. domestic violence, youth, medically vulnerable clients) in support of housing placement.

2. Collaboratively developed and tested strategies with shelters to improve rehousing outcomes

DFSS launched an active contract management strategy, bringing together 25 shelter providers monthly to develop and implement ideas for improving rehousing outcomes.

DFSS and shelters collaboratively identified the most promising levers for improving key outcomes within shelter service delivery:

For emergency shelters serving single adults:

• Increasing the share of chronically homeless and veteran clients with a complete Coordinated Entry System (CES) assessment, a prerequisite for accessing most housing resources. Veteran clients and individuals experiencing chronic homelessness were often the most likely to be matched to permanent housing based on the Continuum of Care’s current prioritization of individuals on the waitlist for housing resources. However, clients couldn’t be placed in permanent housing without a completed CES assessment, and many shelters had low assessment completion rates. One reason for this was the inaccessibility of assessments: assessments were primarily available only during standard working hours, assessments required clients to spend a substantial amount of time answering detailed, and often sensitive, questions about themselves, and it was not always clear to clients why the assessment was important to complete. Shelters identified ways to improve assessment completion, focusing initially on making assessments more accessible to clients. Strategies to overcome scheduling difficulties, as clients often work during the day, included scheduling an assessor to be on site in the evenings when clients are available or assessing some clients by phone. To increase the portion of clients who showed up for assessment appointments, providers began providing reminders the day before an appointment and emphasized the importance of the CES assessment as the primary opportunity for permanent housing during intake conversations. To increase client willingness to complete the assessment, providers allowed clients to enter answers to sensitive questions themselves, and some provided coffee and tea refreshments during appointments. As a result of these strategies, and in combination with other efforts by DFSS and the Continuum of Care, the percentage of chronically homeless and veteran clients assessed for Coordinated Entry increased from an average of 61% to an average of 69%.
• Supporting clients in navigating alternative housing programs if they do not match to housing resources through CES. 91% of entering clients had a monthly income below $1,000, making market-rate housing unsustainable, but many may have been eligible for other housing opportunities (e.g. programs for seniors or individuals with a disability). However, alternative housing programs were fractured, and often difficult to apply for. DFSS and the shelters identified a list of promising practices to support clients in navigating alternative housing resources, made up of solutions they generated collaboratively and lessons from shelters that had already demonstrated success. Initial practice changes tested included creating resources for shelter staff to support clients through housing navigation, such as a list of housing resources for which clients could be eligible and steps for navigating those options, and a script to mitigate client concerns when deciding between limited housing options. Some shelter providers also began to explore separating housing navigation staff resources from general case management, because it required a specific set of expertise and skills.
• Adjusting data collection practices to capture more robust data around exit destinations, a priority outcome metric to determine whether individuals exiting the shelter have been connected to stable housing. Many shelter providers did not have robust data on exit destinations for their clients, so they didn’t know whether a client had moved on to stable housing or if they were continuing to experience homelessness elsewhere. Without knowing whether individuals exited to stable housing, it was difficult for shelters to know if their efforts to support clients in achieving more stable housing were making a difference. To address this, initial practice changes tested included creating a suggestion box and comment cards, where clients who did not complete a formal exit interview could share their contact information, in addition to their feedback and interest in follow-up services. For returning clients, providers began asking at intake where they went the last time they exited shelter to update their exit destination from their previous stay accordingly. Providers also began to leverage moments when clients returned to shelter after exiting (e.g. for mail, to access additional services like furnishing for a new home) to ask them to fill out a form with their new address. As a result of these efforts, the percentage of exit destinations captured by shelters increased from an average of 27% to an average of 43%.

For interim shelters serving families:

• Supporting clients in finding and matching to affordable units. Finding affordable units was identified as a critical component of supporting families to achieve stable housing. DFSS and providers identified this as an important lever, and future efforts will focus on improving a set of practices to help match families to affordable housing units.
• Increasing income through employment. DFSS and shelter providers identified lack of income from employment as the most prevalent barrier preventing families from exiting to permanent housing. Initial practice changes that were tested to increase income through employment included: increasing connections to potential job opportunities by bringing in employers with active job leads for a regular cadence of on-site presentations, updating an onsite job board, providing transportation to job interviews, bringing in former clients to share success stories with employment, and connecting clients to nearby childcare programs. As part of their efforts to connect families to employment opportunities, shelters made it a priority to strengthen referrals to workforce programs (see the call-out box below).

The Innovation

With support from the GPL, DFSS has redesigned the way that Chicago’s shelter system supports individuals and families experiencing homelessness. As a result of these efforts, project partners have:

1. Redistributed approximately $19 million in shelter funding for over 3,300 shelter beds to explicitly focus on quickly and stably rehousing clients  

The new results-driven procurement for shelter services has redesigned over 70% of the city’s overall shelter system (approximately $19 million in shelter funding) to explicitly focus on rehousing outcomes. The new contracts have clarified the goals and core activities of the spectrum of shelter programs and defined rehousing as the primary goal.

These adjustments to the RFP led to applications that were much more responsive to the activities and goals of different program models (i.e. emergency shelters, interim shelters, and specialized interim shelters). For example, several interim shelters for single adults were previously defined as such solely because they were 24-hour programs. The new service spectrum, articulated in the procurement, allowed these shelters to “drop-down” into the emergency category, which better reflected their low-intensity service models. Now, these shelters are working towards a core set of activities that is most appropriate for the population they are serving and are able to connect with similar shelters to troubleshoot barriers and identify best practices.

2. Identified and improved service delivery activities that help to rehouse clients, while equipping DFSS with better data to drive decision-making

At the project’s outset, DFSS and shelter providers did not have clarity over which (of the many) service delivery activities contributed most to quickly and stably rehousing shelter clients. Through continuous, collaborative data-driven meetings, DFSS and shelters identified the most promising levers for improving key outcomes within shelter services and worked to improve them. As a result, they have increased the proportion of chronically homeless and veteran clients in emergency shelters that have completed a Coordinated Entry System assessment to be matched with housing resources (from 61% to 69%). Shelter providers have also increased the proportion of known exit destinations for adults leaving emergency shelter (from 27% to 43%), making it easier to follow-up with clients to determine if they are stably housed. 

3. Internalized capacity and tools for DFSS to continually update the role of shelters in helping make homelessness rare, brief, and non-reoccurring  

Project partners dedicated a new role on the Homeless Services team to driving performance-focused efforts across the homeless services portfolio, and additional staff capacity will remain dedicated to supporting active contract management meetings within that portfolio. This structure will generate insight and action to improve shelter service delivery, continuing to test whether improved shelter service delivery can have a meaningful impact on high-level outcome goals (including increasing exits to permanent housing and decreasing lengths of stay in shelter). 

These learnings will also be reflected in updates to Chicago’s broader Continuum of Care program model definitions (reflecting essential services and outcome goals), as well as in the next DFSS procurement cycle, which will include a re-organization of program models based on learnings about unique needs of different populations, updated hypotheses about priority activities for in-shelter services, and a more nuanced understanding of the resources required to execute these priority activities.