Poster #144 - Displacement Without Separation: Eviction Reform and Domestic Violence Recidivism and Revictimization in Wisconsin

• In Event: Friday Poster Reception

Abstract

Domestic violence and housing instability are deeply intertwined. In Milwaukee and Dane counties, Wisconsin, 14.9% of eviction filings co-occur with a domestic violence (abbreviated as "DV") case at the same address within one year, and women are the named defendants in 57.5% of DV-linked filings — meaning the survivor is often the tenant of record. When a landlord evicts such a household, the survivor and abuser often remain together out of financial necessity: the man's criminal record and the woman's new eviction record each make it harder to secure a separate lease, pushing the pair to relocate as a unit. Whether faster eviction protects survivors by disrupting abusive relationships or destabilizes households in ways that fuel further violence remains an open question.

This paper estimates the effect of eviction reform on three outcomes for DV-affected households: abuser DV criminal recidivism, survivor DV re-victimization (a new DV case in which the co-residing survivor appears as victim), and household dissolution. I exploit Wisconsin Act 76 (effective March 1, 2014), which compressed eviction timelines, authorized corporate property managers to file on behalf of owners, and added DV-specific protections including hotline-notice requirements and crime-victim eviction safeguards.I link three Wisconsin Circuit Court Access Program (CCAP) extracts for 2010–2017 — 61,455 eviction filings, 31,585 DV criminal cases, and 8,814 DV restraining orders — to InfoUSA, an annual commercial consumer panel covering roughly 900,000 Milwaukee and Dane County individuals, organized as households and addresses, per year. Although 95.9% of post-2008 DV petitioner names in CCAP are court-sealed, abuser names and case addresses remain visible. I therefore identify DV-exposed adults indirectly: the abuser's name and case address are matched to InfoUSA to recover the abuser's anonymized household identifier, and every other adult in that household is treated as a DV-exposed co-resident. Each co-resident receives a stable individual identifier that supports longitudinal tracking without ever accessing sealed names.I estimate a propensity-score-matched difference-in-differences across the March 2014 cutoff on 4,507 DV-exposed adults (1,351 treated, 3,156 matched controls). Matched controls inherit treated judgment dates to remove calendar-time confounds, and outcomes are measured cumulatively at one- to five-year horizons.

At three years post-judgment, Act 76 increased abuser DV criminal recidivism by 8.1 percentage points (p = 0.008, a 27% increase over the control mean of 29.9%), increased survivor DV re-victimization by 5.9 percentage points (p = 0.031), and reduced household dissolution by 4.4 percentage points (p = 0.030). A non-DV criminal placebo yields negative, significant coefficients, ruling out a generalized criminogenic channel. Pre-trend tests pass for re-victimization (F = 1.05, p = 0.369) and dissolution (F = 1.51, p = 0.210). The mechanism is displacement without separation: 89% of evicted DV households in the matched sample relocate as an intact unit, so abuser reoffending translates directly into renewed victimization of co-residing survivors.These findings carry direct implications for criminal-justice policy. Procedural reforms that speed or expand eviction — even when paired with DV-specific protections — can worsen criminal outcomes for survivors when households relocate intact.

Author

• Xiaoou Zhan, University of Wisconsin-Madison
  Presenting Author