The Ties behind Eviction: Using Social Network Analysis to Explore High-Filing Landlord-Attorney Connections

Background & Purpose

Eviction is known to exacerbate socioeconomic, health, and racial inequities in addition to perpetuating chronic homelessness. Decades of research demonstrate that experiencing an eviction increases likelihood of exposure to major public health concerns such as HIV and drug overdose. Nationwide, families with school-age children and Black women are most likely to experience eviction.

Although eviction moratoria were temporarily implemented in response to COVID-19, these protections have since been lifted and eviction rates now exceed pre-pandemic averages. Little is known about the social network of high-filing landlords and the attorneys they partner with to navigate the legal process; a deeper understanding could inform policy initiatives for preventing homelessness. This study explores characteristics of these networks as well as differences between in state and out-of-state owner networks.

Methods

Design: This study employs a social network analysis of cross-sectional data compiled from public court and property records from 2023 in St. Louis City and County.

Sample: The sample was comprised of the one hundred landlords who owned the greatest number of taxable parcels in St. Louis in 2023 to allow for targeted analysis of large enterprises. Additionally, the sample included all attorneys who facilitated an eviction in 2023 (n=276).

Procedures: An undirected two-mode network consisted of owner and attorney nodes tied by the presence of a filing initiated by the owner-attorney dyad and weighted by number of filings. Single-mode undirected networks of owners and attorneys were generated wherein unweighted ties indicated the presence of an eviction filing conducted with the same owner/attorney.

Analysis: Descriptive social network analyses evaluated isolates, components, and density to show incidence and magnitude of owner-attorney collaboration. Degree and betweenness were measured to assess for key agents and communication channels within the network. The Louvain algorithm was applied to analyze single-mode networks for the presence of subcommunities, allowing further insight into latent collaboration patterns.

Results

Of the top 100 owners, 43 filed an eviction with 29 attorneys in 2023. The two-mode network had a low density (D=0.001) and was comprised of 291 isolates, four small components, and a single main component with 85 nodes (owners n=56, attorneys n=29). Louvain analysis of the main components of the single-mode networks of owners (total n=53; in-state n=19, out-of-state n=34) and attorneys (n=26), detected four communities in each (owner Q=0.2, attorney Q=0.25).

Conclusions & Implications

This study provides initial insight into the composition and collaboration patterns of networks responsible for eviction filings in St. Louis. The low density of all three networks indicates that both owners and attorneys tend to file evictions with a select number of partners. Of the forty-three owners who filed evictions against St. Louis residents in 2023, the majority were located outside of Missouri - suggesting that out of state enterprises may be powerful drivers of local eviction rates. Ongoing research by the team incorporates additional owner characteristics and Exponential Random Graph Modeling to test homophily between in state and out-of-state owners in order to further explore opportunities to prevent and reduce homelessness in St. Louis.