A study of 1,500 Detroit women puts half the region's evictions outside the court system

The Georgia study is the one to read for design. Researchers linked every birth record from 2000 to 2016 to court records from the same years, then compared mothers evicted during the second or third trimester with mothers evicted during the first, and found the later group more likely to have low birth weight and premature babies ^[5]. The comparison group had also been evicted. That removes much of what usually confounds this question, since the women who end up in eviction court differ from the women who do not in income, health and neighborhood.

It also sets the study's ceiling. The court record is what makes the linkage possible, so a removal that never produced a filing cannot enter the sample.

A second study went the other way. It used medical records for children and surveys of caregivers in four U.S. cities between 2011 and 2019, covering court-ordered and informal evictions, and found evicted children at higher risk of intellectual, motor and behavioral problems, of hospitalization after an emergency room visit, and of worse health overall ^[6]. Its comparison group is children with no eviction history, so that contrast also carries whatever else separates the two sets of families.

Doubling the 2025 filing count gives roughly 40,700 removals ^[20]. The two numbers are not the same unit: 20,370 is cases filed ^[1], and the phys.org article does not say how many ended with a judge signing an order. Under Michigan law that signature is what completes an eviction ^[9].

The by-half estimate comes out of a specific population. The SECURE Study covers nearly 1,500 Black women aged 18 to 46 in metro Detroit, about 800 of whom were evicted at some point, through the courts or outside them ^[3]. Its lead researcher, who describes the project as the first and most comprehensive study of court-ordered and illegal evictions among Black women in the United States ^[15], writes that Black women and children are the groups most affected, at the intersection of structural racism and sexism ^[17]. Reading the ratio across all Detroit renters extends it beyond the group measured.

Formal protections are strongest for tenants with a written lease, the researcher writes ^[16]. Twenty-six percent of the women who were pregnant or postpartum when they were illegally evicted had no lease, and 21% of all illegally evicted interview participants were on month-to-month tenancies ^[7][8]. Michigan requires a landlord to give 30 days notice for a non-renewal, and bars non-renewal in retaliation for a repair request ^[10].

The qualitative half of the work is 55 in-depth interviews with women who reported illegal evictions ^[11]. Nearly a third of them said they had also been evicted as children ^[12]. One mother, herself evicted as a child, said she rented a house that needed repairs and agreed with the landlord to withhold rent for three months until the work was done ^[13]. "The guy, you know, threatened me, and he wouldn't fix the house," the mother said ^[14].