Perceptions of Partner Decarceration among Survivors of Intimate Partner Violence during the COVID-19 Pandemic.

The COVID-19 pandemic prompted the early release of thousands of incarcerated individuals, including those with histories of intimate partner violence (IPV) perpetration. Survivor advocates stress the importance of adequate supports for decarcerated individuals during re-entry, and notification and supports for their partners or ex-partners if there is a history of IPV. This survey assessed IPV survivors' expectations of and experiences with decarceration in the state of Michigan. Findings highlight that out of 42 survivors with recently decarcerated (ex-)partners, 64.3% reported helpful behavior on the part of their released partner. By contrast, out of 72 survivors with still-incarcerated (ex-)partners, the same percentage - 64.3% - expected harmful behavior from their partner if released. Decarceration efforts may distinguish between individuals who are likely to harm versus help (ex-)partners upon release. Nonetheless, survivors reported several unmet needs, indicating the need for better re-integration services for decarcerated individuals and their families.

Introduction

The COVID-19 pandemic is associated with an increase in intimate partner violence (IPV) across the globe [1]. IPV victimization results in consequences such as physical injury, emotional distress, increased HIV risk, and significant economic burdens on both survivors and the broader society, making it a significant public health risk [2-4].

Stay-at-home orders forced many couples to remain at home together, increasing stress and the window of time during which abuse can occur in violent relationships [5]. Survivors of IPV have voiced hesitation about seeking services due to fear of contracting COVID-19 [6].

The COVID-19 pandemic also led to the early release of thousands of incarcerated individuals, including but not limited to those with histories of IPV perpetration [7]. Even before the pandemic, US prisons and jails were frequently overcrowded and posed an increased risk of infection outbreaks [8, 9]. The impending public health threat of the pandemic, which led to over 33,000 inmate infections and 159 inmate deaths in Michigan alone as of May 2022 [10], propelled additional decarceration efforts to minimize adverse impacts on incarcerated individuals, particularly people of color, older people, and those with comorbidities [11, 12].

Advocates have highlighted the importance of considering IPV survivors’ safety in these decarceration efforts. In May 2020, the Massachusetts Coalition Against Sexual Assault and Domestic Violence called for survivors to receive advance notice of the release of individuals who caused them harm and for the provision of robust advocacy and safety planning for survivors [13]. The concern for survivor safety is supported by research that points to societal reentry following incarceration as a time of heightened IPV risk and homicide, particularly for women re-engaging with released male partners and those with a history of IPV [14, 15]. Advocates have also highlighted the need for re-entry services for newly released individuals [13]. Recently incarcerated individuals are more likely to face numerous socioeconomic disadvantages, such as unemployment, financial instability, substance use, and prior IPV perpetration and victimization, which are all tied to increased IPV perpetration risk [16-18]. Supporting released individuals in overcoming these challenges therefore has the potential to aid both their successful rehabilitation and the wellbeing of survivors.

Despite the urgency of decarceration and the importance of robust support for newly-released individuals and IPV survivors, data on survivors’ experiences with released partners during the pandemic are limited. Research conducted before the pandemic demonstrates the importance of attention to survivors’ unique stated needs and experiences when designing services for them [19]. To our knowledge, no published research has quantitatively evaluated the experiences of IPV survivors with incarcerated or recently released partners during the COVID-19 pandemic. Survivors’ feedback on what measures would help them feel safe upon partner release is critical for structuring future policy and social services.

Methods

Study Design and Recruitment

Study participants were recruited via Qualtrics Panels, using quota sampling methods to achieve a demographically matched sample of the State of Michigan for race/ethnicity (15% Black, 6% Asian, 7% multiracial or other race/ethnicity), region (15% rural-residing determined based on residence ZIP code), and age (34% aged 35–54, and 21% over the age of 55) [20]. Eligible participants were at least 18 years of age; women, nonbinary, and/or transgender; resided in the state of Michigan; had been in a romantic and/or sexual relationship within the past year. The research team worked in consultation with the Michigan Domestic and Sexual Violence Prevention and Treatment Board (MDSVPTB) to design the survey questions and safety protections. A total of 1169 complete survey responses were collected between June 26 and August 11, 2020; details available in Peitzmeier et al. [20]. This research was exempt from review by the University of Michigan Institutional Review Board.

Survey Measures

Standard demographic assessments were used. Intimate partner violence included physical, sexual, psychological, and technology-facilitated abuse, which were measured with items adapted from validated screening tools; details are available in Peitzmeier et al. [20]. Participants were also screened for “COVID-related IPV” based on reports from domestic violence hotlines of individuals intentionally placing their partners at risk for COVID as a form of abuse [6]. Individuals who screened positive for any lifetime IPV were asked questions about incarcerated partners and were included in the analytic sample (N = 549).

The primary outcome of interest was participants’ experiences of partner or ex-partner incarceration and/or release during the COVID-19 pandemic. Participants were first asked whether they had a partner or ex-partner who was currently incarcerated. If they answered yes, they were then asked whether this partner made threats toward them while incarcerated; how likely this person would be to contribute positively to the participant and/or their child(ren) if released; how likely this person would be to attempt to harass, hurt, or kill the participant or harm their child(ren) if released; and what the participant would need to feel safe if this person were released.

Participants were also asked whether they had a partner or ex-partner who had been released from incarceration due to COVID-19. If they answered yes, they were asked whether they had been informed in advance of this person’s release; whether this person had contributed positively to the participants’ and/or their child(ren)’s wellbeing since release; and whether this person had attempted to harass, harm, or kill the participant or their child(ren) since release.

Data Analysis

Descriptive statistics were used to ascertain the prevalence of experiences related to partner or ex-partner incarceration and/or release during COVID-19 across various demographic groups (see Table 1). Chi-squared and Fisher’s exact tests were used to compare differences in several key variables by demographic groups.

Table 1

Overall sample description of incarceration-related survey items (N = 549)

	Total
Has a currently incarcerated (ex-)partner (N = 549)	72 (13.4%)
I expect currently incarcerated (ex-)partner to be harmful if released (N = 70)	45 (64.3%)
I expect currently incarcerated (ex-)partner to be helpful if released (N = 70)	23 (31.9%)
(Ex-)partner was released due to COVID-19 (N = 549)	42 (7.7%)
If (ex-)partner released, I was informed of their release beforehand (N = 42)	24 (57.1%)
If (ex-)partner released due to COVID-19, they were harmful post-release (N = 42)	13 (31.0%)
If (ex-)partner released due to COVID-19, they were helpful post-release (N = 42)	27 (64.3%)

Results

Currently Incarcerated Partners

Out of 549 survivors, 72 (13.4%) reported having at least one partner or ex-partner who was currently incarcerated (Table 1). Participants who were less educated, transgender, younger, LGBQ + , and who had children in the home were more likely to report having currently incarcerated partners. Currently incarcerated partners were imprisoned on charges of breaking and entering (38.9%), domestic violence (36.1%), drug-related charges (30.6%), or other (23.6%), including murder or weapons charges (5.6%). Forty-two percent of participants with incarcerated partners had been threatened by their partner(s) from jail or prison. Those who were less educated, disabled, transgender, and higher-income were more likely to report receiving threats.

Forty-six percent of participants expected their incarcerated partner to cause only harm to them or their children if released; 14.3% expected them to only provide help; 18.6% expected a mix of helpful and harmful behavior; and 21.4% expected neither helpful nor harmful behavior. Participants who were disabled or transgender were particularly likely to expect released partner harm. Most of the 45 participants reporting that their partner would be likely to harm them if decarcerated said services and support would help them feel safe if their partner were decarcerated, including protective orders (60%), support from family and friends (24.4%), legal help (22.2%), and help relocating (15.6%) (Table 2).

Table 2

Demographic description of respondents with currently incarcerated partner(s) (N = 549)

	Total (N = 549)	Has currently incarcerated (ex-)partner (N = 549)	Incarcerated (ex-)partner threatened me while incarcerated (N = 72)	I expect incarcerated (ex-)partner to be… if released (N = 70)
		Yes	Yes	Harmful (N = 70)	Helpful (N = 72)
Total		72 (13.4%)	30 (41.7%)	45 (64.3%)	23 (31.9%)
Education
Did not complete H.S. (n, (%))	27 (4.9%)	9 (33.3%)*	7 (77.8%)*	8 (88.9%)	2 (22.2%)
H.S. diploma/GED (n, (%))	115 (20.9%)	17 (14.8%)*	10 (58.8%)*	12 (75.0%)	3 (17.6%)
Some college (n, (%))	190 (34.6%)	25 (13.2%)*	6 (24.0%)*	11 (45.8%)	7 (28.0%)
College graduate (n, (%))	217 (39.5%)	21 (9.7%)*	7 (33.3%)*	14 (66.7%)	11 (52.4%)
Disability (n, (%))	161 (29.3%)	33 (20.5%)	22 (66.7%)*	28 (87.5%)*	8 (24.4%)
Gender
Cisgender woman (n, (%))	534 (97.3%)	64 (12.0%)*	22 (34.4%)*	37 (59.7%)*	22 (34.4%)
Trans/nonbinary (n, (%))	15 (2.7%)	8 (53.3%)*	8 (100.0%)*	8 (100.0%)*	1 (12.5%)
Age
18–24 (n, (%))	132 (24.0%)	19 (14.4%)*	10 (52.6%)	12 (63.2%)	11 (57.9%)*
25–44 (n, (%))	251 (45.7%)	48 (19.1%)*	18 (37.5%)	30 (63.8%)	10 (20.8%)*
45 + (n, (%))	166 (30.2%)	5 (3.0%)*	2 (40.0%)	3 (75.0%)	2 (40.0%)*
Urbanicity
Rural (n, (%))	102 (18.6%)	12 (11.8%)	3 (25.0%)	7 (58.3%)	3 (25.0%)
Suburban (n, (%))	53 (9.7%)	4 (7.5%)	1 (25.0%)	1 (25.0%)	0 (0.0%)
Urban (n, (%))	394 (71.8%)	56 (14.2%)	26 (46.4%)	37 (68.5%)	20 (35.7%)
Annual income
< $40,000 (n, (%))	235 (42.8%)	39 (16.6%)	9 (23.1%)*	21 (56.8%)	11 (28.2%)
$40,000-$79,999 (n, (%))	190 (34.6%)	19 (10.0%)	11 (57.9%)*	14 (73.7%)	6 (31.6%)
> $80,000 (n, (%))	100 (18.2%)	11 (11.0%)	9 (81.8%)*	9 (81.8%)	3 (27.3%)
LGBQ + (n, (%))	110 (20.1%)	24 (21.8%)*	13 (54.2%)	17 (73.9%)	10 (41.7%)
Person of color (n, (%))	196 (35.8%)	44 (12.5%)	19 (43.2%)	15 (57.7%)	13 (46.4%)*
Children under 18 in home (n, (%))	262 (47.7%)	54 (20.6%)*	25 (46.3%)	8 (47.1%)	17 (31.5%)

H.S., high school; LGBQ + , lesbian, gay, bisexual, queer/questioning. Asterisk denotes p < 0.05

Recently Decarcerated Partners

Eight percent (n = 42) of the sample reported having at least one partner or ex-partner who was decarcerated due to COVID-19, with higher rates among participants who were disabled, transgender/nonbinary, those aged 25–44, LGBQ + , less educated, and those with children living in the home. Of these, 57.1% of participants were informed of their partner’s release beforehand, with disabled, transgender, and higher-income individuals more likely to report being informed. Released partners had been jailed on charges of breaking and entering (52.5%), domestic violence (35.7%), drug-related (23.8%), or other (9.5%) offenses (Table 3).

Table 3

Demographic description of respondents with decarcerated partner(s)

	Total(N = 549)	(Ex-)partner was released due to COVID-19 (N = 549)	I was informed of (ex-)partner’s release beforehand (N = 42)	(Ex-)partner was… post-release(N = 42)
		Yes	Yes	Harmful	Helpful
Total		42 (7.7%)	24 (57.1%)	13 (31.0%)	27 (64.3%)
Education
Did not complete H.S. (n, (%))	27 (4.9%)	7 (25.9%)*	6 (85.7%)	2 (28.6%)	5 (71.4%)
H.S. diploma/GED (n, (%))	115 (20.9%)	15 (13.0%)*	10 (66.7%)	6 (40.0%)	8(53.3%)
Some college (n, (%))	190 (34.6%)	14 (7.4%)*	5 (35.7%)	4 (28.6%)	8 (57.1%)
College graduate (n, (%))	217 (39.5%)	6 (2.8%)*	3 (50.0%)	1 (16.7%)	6 (100.0%)
Disability (n, (%))	161 (29.3%)	21 (13.0%)*	18 (85.7%)*	6 (28.6%)	15 (71.4%)
Gender
Cisgender woman (n, (%))	534 (97.3%)	34 (6.4%)*	16 (47.1%)*	11 (32.4%)	20 (58.8%)
Trans/nonbinary (n, (%))	15 (2.7%)	8 (53.3%)*	8 (100.0%)*	2 (25.0%)	7 (87.5%)
Age
18–24 (n, (%))	132 (24.0%)	10 (7.6%)*	6 (60.0%)	4 (40.0%)*	7 (70.0%)
25–44 (n, (%))	251 (45.7%)	29 (11.6%)*	16 (55.2%)	6 (20.7%)*	20 (69.0%)
45 + (n, (%))	166 (30.2%)	3 (1.8%)*	2 (66.7%)	3 (100.0%)*	0 (0.0%)
Urbanicity
Rural (n, (%))	102 (18.6%)	4 (3.9%)	1 (25.0%)	2 (50.0%)	3 (75.0%)
Suburban (n, (%))	53 (9.7%)	4 (7.5%)	2 (50.0%)	0 (0.0%)	4 (100.0%)
Urban (n, (%))	394 (71.8%)	34 (8.6%)	21 (61.8%)	11 (32.4%)	20 (58.8%)
Annual income
< $40,000 (n, (%))	235 (42.8%)	21 (8.9%)	7 (33.3%)*	9 (42.9%)	10 (47.6%)*
$40,000-$79,999 (n, (%))	190 (34.6%)	14 (7.4%)	10 (71.4%)*	3(21.4%)	10 (71.4%)*
> $80,000 (n, (%))	100 (18.2%)	6 (6.0%)	6 (100.0%)*	0 (0.0%)	6 (100.0%)*
LGBQ + (n, (%))	110 (20.1%)	15 (13.6%)*	9 (60.0%)	8 (53.3%)*	7 (46.7%)
Person of color (n, (%))	196 (35.8%)	28 (8.0%)	7 (50.0%)	5 (35.7%)	9 (64.3%)
Children under 18 in the home	262 (47.7%)	31 (11.8%)*	18 (58.1%)	10 (32.3%)	19 (61.3%)

H.S., high school; LGBQ + , lesbian, gay, bisexual, queer/questioning. Asterisk denotes p < 0.05

Participants largely reported that their released partners were helpful to them or their children (64.3%), with those who were higher-income reporting partner helpfulness at higher rates. However, just under one-third (31.0%) reported that released partners were harmful, including 3 individuals who said their partner had attempted to harm them or their children or that their partner had attempted to kill them since release. Harmful behaviors were more likely to be reported by LGBQ + individuals and individuals who were 18–25 or over 45.

Discussion

Roughly 1 in 10 survivors in the state of Michigan has a currently incarcerated partner and 1 in 12 survivors had a partner who was decarcerated in the first months of the COVID-19 pandemic. Encouragingly, participants with released partners were twice as likely to report that their released partner had contributed positively to their or their children’s well-being (64.3%) as they were to report harmful behavior (31.0%). In contrast, participants with partners who were still incarcerated were twice as likely to expect that their partner would be harmful (64.3%) as they were to be helpful (31.9%) if released. This suggests that decarceration efforts in the early months of the pandemic may have reasonably distinguished between individuals who were more versus less likely to cause harm to survivors. Notably, 35.7% of released partners and 36.1% of those still incarcerated had been convicted of domestic violence crimes. Overall, many decarcerated individuals were able to contribute positively to the well-being of their partners or children. Survivor and/or re-entry services may have also been helpful in preventing harm.

Decarceration efforts were not perfect: it should be noted that 3 (7.1%) participants with recently released partners reported that their partner had tried to harm them or their children or attempted to kill them since being released. This warrants further investigation into survivors’ specific situations, as well as re-entry programming to help mitigate post-release factors associated with IPV perpetration, with extra attention to populations likely to experience more partner harm and/or less partner help. We note that demographic trends in who is likely to be helped or harmed by partner decarceration must be further investigated, including the ways in which systemic oppression impacts individuals’ ability to help their partners, ex-partners, or children post-incarceration, and should not be used to make decisions regarding individuals’ readiness for release from incarceration without rigorous future research. We also note that 1 in 7 individuals who still had incarcerated partners expected those partners to contribute only positively to their and their families’ well-being if released, indicating that a substantial proportion of strong candidates for decarceration were missed by decarceration efforts at the time of the survey.

Concerningly, only just over half of participants with decarcerated partners were alerted about their partner’s release beforehand. This lack of warning contravenes best practices, may needlessly endanger individuals, and may especially impact marginalized survivors with less access to legal or other resources [13]. This calls for improvements to systems of notifying individuals of their partners’ upcoming release, including clearly and accessibly informing survivors of how to apply to be notified of said release.

Participants with incarcerated partners reported several unmet needs that would help them feel safe in case of their partner’s release, in particular protective orders. Due to reduced domestic violence services during the pandemic, these needs may be particularly difficult to meet. Protective orders may be particularly, but not exclusively, helpful to survivors if obtained prior to their partner’s release, which hinges in part on survivors’ upfront knowledge of upcoming partner release.

This study has several limitations. Released partners’ behavior was evaluated at a single time point early in the pandemic, so subsequent harmful and/or helpful behavior may emerge that was not captured by this study. Participants’ expectations of partner behavior upon release may not always be accurate, making it difficult to compare currently and formerly incarcerated partners. We found similar rates of domestic violence convictions between incarcerated and released partners. However, it is unclear to what extent domestic violence convictions were considered in decarceration decisions. This is an important consideration for future decarceration efforts, which may also be complicated by the fact that domestic violence often goes unreported, and that many arrested domestic violence perpetrators do not end up being charged with domestic violence. Additionally, it is not certain whether the incarcerated or decarcerated partners participants described in this study were the same partners who had previously engaged in IPV against them. Future research can build upon these findings by employing targeted recruitment methods to obtain larger and longitudinal samples, collect more information about incarcerated partners, and consider collecting qualitative data from survivors. Future studies focusing on identifying services and interventions that would support survivor safety in the context of decarceration, particularly for marginalized populations with limited access to resources, are urgently needed.