Association for Public Policy Analysis & Management

Preservation of existing affordable housing is not a new idea. However, most prior and current research and programs have focused on preserving subsidized upon subsidy expiration. While this is undoubtedly important, the vast majority of housing units considered affordable are not covered by any form of subsidies. These naturally-occurring affordable housing (NOAH) units house the majority of low-income Americans, but many units are at risk of becoming unaffordable due to market forces.

Recent city plans and programs have identified NOAH preservation as a key pillar of maintaining affordability (e.g., Minneapolis 2040, City of Oakland Site Acquisition, Rehabilitation and Naturally Occurring Affordable Housing (NOAH) Preservation Program). These programs generally provide incentives for property owners or seek non-profit partners who acquire and or operate the properties to maintain property affordability. While these programs are attractive in theory, implementation becomes challenging given that NOAH tend to occur in smaller multifamily properties with a diverse and often localized ownership base, often obscured by multiple legal entities representing the same owner. Unlike subsidized housing, databased of NOAH owners are not readily available. In addition, local partners may often be more willing to work with city and county preservation programs, but this requires an even more detailed understanding of ownership structures. Altogether, scaled coordination of incentives or purchasing becomes embroiled in understanding who owns what and where, and who is a willing partner in preservation program.

This paper seeks to understand the scope of preservation challenges and recommend an open source data-driven solution to such coordination challenges. We develop a framework to obtain, clean, and categorize multifamily ownership data that saves time and provides transparency. We then provide case studies from three jurisdictions that have implemented this approach and have reduced coordination time and cost in their preservation activity.

This paper provides a clear improvement on previous attempts to clean and aggregate multifamily ownership data at a large scale, by using publicly available OpenRefine software. OpenRefine corrects for textual inaccuracy and aggregates diverse text fields by similarity. We post-process the OpenRefine output to derive owner typologies by size and distance to owned properties. We also use the output over multiple metropolitan areas to assess cross-state concentration of multifamily property ownership. Together, our framework and case studies provide a way forward for policymakers to implement NOAH preservation programs.