Re-envisioning Housing Data: Usability, Integration, and Systems Alignment

Participated in a panel discussion at the 2025 Monarch Housing Associates Conference, “Housing as a Human Right: Driving Progress in Times of Uncertainty,” on how data systems and predictive modeling are being used to address New Jersey’s housing crisis.

Conference Overview

The conference brought together housing advocates, policymakers, service providers, and data experts to discuss strategies for maintaining housing stability during a period of significant federal policy uncertainty and potential funding cuts.

Panel Participants

Tina McGill — Assistant Director, Office of Homelessness Prevention, NJ DCA. Tina leads policy implementation and homelessness prevention strategy, coordinating across state housing programs and community-based organization partnerships.

Marlon Duprey — Community Consultant, Hudson County Advisory Board. Marlon brought the community perspective and lived experience that grounds the data work in reality.

Gavin Rozzi — Director, DHCR Data Center, NJ DCA. I presented on the technical infrastructure behind our housing data systems, including the Homelessness Entry Risk Framework (HERF) and our decision support tools.

Dr. Dan Treglia — Rutgers Center for State Health Policy. Dan’s research focuses on evaluating housing interventions and linking data across multiple government systems, including the Medicaid Section 1115 Demonstration Project.

The Homelessness Entry Risk Framework (HERF)

I presented DHCR’s predictive modeling system, which forecasts homelessness vulnerability at the ZIP code level across New Jersey. HERF integrates real-time economic indicators — eviction filings, rent burden, mortgage distress, employment data — and accounts for structural factors like racial inequity, historical patterns, and policy responsiveness. The system also supports scenario modeling to predict impacts of federal budget cuts, and includes a GPT interface for natural language queries.

Federal Policy Shock Analysis

This was the section that drew the most engagement from the audience. We modeled potential impacts from proposed federal cuts:

A 20% cut to the Housing Choice Voucher program would affect 4,316 households. A 44% cut would hit 9,496 total households — 7,510 able-bodied adults, with the remainder elderly and disabled. Our projections showed 74-81% of terminated households would face homelessness within the first year.

The racial equity dimension was stark: 6 out of 10 households cut in the worst-case scenario would be households of color. We created ZIP code-level risk maps showing this disproportionate impact. For the top 14 communities alone, additive sheltering costs would reach roughly $18M within 90 days.

Systems Integration and Data Linkage

Dr. Treglia presented on the difficulty of linking multiple data systems — HMIS, Medicaid enrollment and claims, managed care organization records, prison health records, and eviction filing data from courts — for comprehensive evaluation of housing interventions.

Medicaid Tenancy and Housing Supports

The panel discussed New Jersey’s use of Medicaid funding for housing services: pre-tenancy supports, move-in assistance, tenancy sustaining services to prevent eviction, and housing modification. Over 40 housing providers have enrolled, with more than 600 successful claims filed.

PowerBI Dashboards and Decision Tools

I demonstrated the interactive data tools we’ve built for DCA leadership — real-time HCV program monitoring across all public housing authorities, risk visualization by ZIP code using HERF, scenario comparison tools for policy analysis, and eviction tracking metrics. Built with React, TypeScript, and modern web technologies.

Policy Implications

The numbers tell a hard story. Losing $440M in federal rent stabilization would cost New Jersey an estimated $1.76-$2.2 billion in downstream crisis response: emergency shelter expansion, healthcare costs for unhoused populations, educational disruption for homeless children, criminal justice interactions, and lost economic productivity.

Given limited resources, the panel identified the highest-impact interventions: scaling CEDD (Comprehensive Eviction Defense & Diversion) with triage at court entry, subsidizing at-risk tenancies through lease-in-place stabilization, deploying rapid stabilization teams via school-linked McKinney-Vento partnerships, incentivizing landlord retention in the affordable housing market, and layering utility assistance to prevent elderly displacement.

We also developed a risk-based framework for allocating scarce resources across three tiers: families with young children, seniors, and disabled adults at the critical level; recently employed workers and young adults aging out at the stabilization level; and chronically homeless populations requiring intensive support.

Looking Forward

The discussion came back to the need for continued investment in data infrastructure — strengthening DCA’s capacity to analyze and respond to housing crises, giving community-based organizations better tools and data access, centralizing administrative processes so CBOs can focus on direct service, and building predictive capabilities that let us get ahead of problems rather than react to them.