Improving how Connecticut supports families with young children experiencing homelessness

Summary

The challenge

In 2019, Connecticut launched a multi-agency effort to better understand and support vulnerable populations experiencing homelessness. The Governor’s Task Force on Housing and Supports for Vulnerable Populations brought together state agencies, community organizations, and service providers to coordinate a response.

This task force will take a data-driven approach to seek a better understanding of this population, and to pilot better ways to make sure we can quickly meet all of a household’s critical needs, stabilizing them more quickly, and achieving better outcomes for these Connecticut residents as well as for our publicly funded systems.

Lisa Tepper Bates, Co-chair, Governor's Task Force

As part of this effort, we focused on families with children from birth to age five — a group with distinct and often overlooked needs. Many of these families didn’t appear in official homelessness data at all. When families doubled up with relatives or friends rather than entering the shelter system, they fell outside the definitions that triggered formal tracking. That meant the state couldn’t see the full scope of the problem, let alone prioritize resources toward it.

The challenges went beyond visibility. Early childhood needs — especially childcare — created additional barriers to employment and stability that older populations didn’t face. And the agencies responsible for supporting these families lacked shared tools and data systems, so coordination happened informally and inconsistently. The central tool for prioritizing families for housing — the family by-name list — was difficult to use and limited in what it could capture, leaving decision-makers without the structured data they needed to act.

The solution

We started with discovery research — interviewing state agencies, community partners, and families themselves to understand how people actually moved through existing systems and where breakdowns occurred. We mapped how data flowed across programs and found consistent gaps in visibility, coordination, and usability that prevented effective prioritization.

One tool kept surfacing in those conversations: the family by-name list, which Coordinated Access Networks relied on to prioritize families for housing. But its design limited its effectiveness. The list couldn’t capture the nuanced information providers needed to make good decisions — things like a family’s childcare situation, employment barriers, or how long they’d been unstably housed. Improving this one tool could improve decision-making across the entire system.

Rather than simply adding fields to the existing list, we rethought what data the system needed to collect and how it should be organized to support prioritization decisions. We developed a working SQL-based prototype to demonstrate how improved data models could drive better outcomes — showing providers and decision-makers what a more useful tool would look like in practice.

A related problem was that agencies had been interpreting key terms differently. What counted as “homeless” or “unstably housed” varied depending on who was collecting the data, which made cross-agency coordination unreliable. We created a shared data dictionary that established common definitions so that data meant the same thing regardless of which agency collected it.

Together, the research, prototype, and data framework gave Connecticut’s Coordinated Access Networks a concrete foundation for identifying and prioritizing families in need — along with design and technical recommendations to guide future system improvements.

The results

• Produced a comprehensive research foundation that clarified how families with young children experience homelessness and interact with state systems, surfacing gaps invisible in existing data
• Built a working prototype of an improved family by-name list demonstrating how better data structures could improve prioritization and coordination across Coordinated Access Networks
• Established a shared data framework through a data dictionary that aligned multiple agencies on common definitions and data usage
• Delivered actionable design and technical recommendations to guide future system improvements for identifying and supporting vulnerable families