Jailed Parents and Their Young Children: Residential Instability, Homelessness, and Behavior Problems

Abstract

This study examined family disruption in the form of jailed parents’ housing instability in the year leading up to their most recent incarceration, including periods of homelessness with and without their children, and links between parental housing instability and children’s behavior problems. Using the Family Stress Proximal Process Model to understand the links between stressors related to family disruption and child outcomes, the study analyzed data from interviews and surveys with 165 jailed fathers and mothers with young children (age 2–6 years) regarding jailed parents’ reports of housing instability during the 12 months prior to their incarceration and child behavior problems. Analyses showed that housing instability, homelessness, and recidivism in jailed parents were relatively common, with a significant proportion of the disruptions occurring with young children, although many disruptions involved parental absence from children. Results indicated that the more months that parents lived with their children prior to incarceration in jail during the past year, the less housing instability the parents experienced. Additionally, multiple regression analyses revealed that more housing instability experienced by parents in the year leading up to their incarceration in jail were associated with elevations in children’s internalizing and externalizing behavior problems. These results have implications for future research that explores family disruption as a mechanism in understanding recidivism and homelessness among adults and risk for child behavior problems in families affected by parental incarceration.

Introduction

During the past several decades, parental incarceration and family housing instability have significantly increased in the United States (Bassuk et al., 2014; Glaze & Maruschak, 2008). In the prior 40 years, incarceration rates have quadrupled (Travis, Western, & Redburn, 2014) and most incarcerated individuals are parents (Glaze & Maruschak, 2008). From 2007 to 2010, there was a 20% rise in homeless families and children (U.S. Department of Housing and Urban Development, 2011) and between 2012 and 2013, there was an additional 8% increase (Bassuk et al., 2014). Current estimates indicate that on any given night in 2016, approximately 550,000 people were homeless, and 35% were families with children (U.S. Department of Housing and Urban Development, 2016). In addition, more than 5 million children (or 7% of all U.S. children) have experienced a coresident parent leaving to go to jail or prison (Murphey & Cooper, 2015). Although many incarcerated parents experience homelessness and housing instability (Glaze & Maruschak, 2008), and their children are at risk for experiencing homelessness (Wildeman, 2014), few studies have examined the co-occurring experiences of parental incarceration and homelessness (or housing instability, which is strongly linked to homelessness; e.g., Bassuk, Rubin, & Lauriat, 1986) in relation to children’s adjustment.

Arditti’s (2016) Family Stress Proximal Process (FSPP) model is utilized to help explore associations among parental housing instability, parental incarceration, and child behavior problems (Figure 1). Grounded in Bronfenbrenner and Ceci’s (1994) bioecological theory, the FSPP model situates itself in the idea that proximal family processes drive child development, and that parental incarceration imposes additional economic, social, and psychological stress on already vulnerable families and children (Arditti, 2016). Because hardship associated with parental incarceration can undermine family relationships and child adjustment over time, and because of high rates of recidivism, parental incarceration is conceptualized as an ongoing stressor rather than a single event (Arditti, 2016). The FSPP model identifies causes of family disruption when a parent is incarcerated, including parental absence from the family because of incarceration, as well as individual and family residential mobility or homelessness that can lead to and stem from parental incarceration. Instability resulting from incarceration can lead to increases in both internal and external stressors and emotional and psychological distress in parents (e.g., parenting stress) and in children, which can lead to an array of emotions and internalizing (withdrawal, sadness) and externalizing (acting out, aggression) child behavior problems (Arditti, 2016; Arditti, 2003).

Figure 1.

Family Stress Proximal Process Model Linking Parental Incarceration and Child Adjustment

Overall, the FSPP model highlights the importance of specific proximal family processes and how alterations in day-to-day family life shape the developmental trajectories of children and other family members when parents are incarcerated, suggesting that parental incarceration contributes to parenting stress and ambiguous loss, which are thought to influence child adjustment (Figure 1). Scholars already have established that housing instability and risk of homelessness, as well as food insecurity and lower rates of education, are examples of disadvantage that are often apparent in the lives of incarcerated parents, even prior to incarceration (Cox & Wallace, 2013; Foster & Hagan, 2009; Haskins, 2014; Johnson, 2009; Wildeman & Wakefield, 2014). Inclusion of these contextual variables in the FSPP model may help account for how accumulated risk, which has been associated with increased rates of both first time and repeated incarcerations (Western & Petit, 2010), relates to child adjustment. A growing body of literature suggests that economic and social experiences of children before one of their parents is incarcerated are, on average, already challenging (i.e., economic difficulties, lower father engagement, violence) (Murphey & Cooper, 2015; Phillips et al., 2006; Wildeman & Wakefield, 2014). Recognizing this, Arditti (2016) argues that even in instances of instability prior to incarceration, the detainment of a parent does more harm than good for children because it imposes even more economic and social problems. While parental incarceration is a significant cause of parental absence, parental housing instability and homelessness do not always signify parental absence, as they can also occur with children in tow. In this sense, residential mobility (whether the child experiences it personally or not) can be seen as a disadvantage that may lead to feelings of stress and ambiguous loss which are further complicated by parental incarceration, thus impacting family well-being and child adjustment. Highlighted in the FSPP, incarcerated parents often experience a great deal of stress, as a result of the incarceration, pre-existing vulnerabilities, and stigma directed toward individuals and families because of the incarceration (Arditti, 2016); this distress has been linked to child behavior problems in families not experiencing incarceration (Crnic et al., 2005), and these processes are likely influential for children of the incarcerated as well.

Consistent with the FSPP model, research has established that family disruptions in the forms of housing instability and parental incarceration are related in several ways. In some cases, housing instability may precede a parent’s incarceration, with impoverished and unsafe neighborhoods or limited employment opportunities contributing to a parent’s involvement in the criminal justice system (Tasca et. al. 2011). For others, family instability, housing insecurity, and even homelessness are results of parental incarceration, further contributing to family financial strain and shifting family roles (Wildeman, 2013). As a consequence of these factors, families with incarcerated parents are at high risk to experience residential instability; indeed, incarceration contributes to a decrease in stable employment, access to social services, and other aspects of individual and family functioning that allow for secure and sustained housing (Geller & Curits, 2011). This finding is in line with prior research documenting that individuals who have contact with the criminal justice system report more residential mobility than other vulnerable groups (Harding & Winship, 2016). Geller and Curtis (2011) found that 22% of homeless men also reported a history of incarceration which proved larger than prior research which found that nearly one-quarter of the shelter population of a sample in New York City reported a history of incarceration in the previous two years (Metraux & Culhane, 2006). High rates of residential mobility are frequently reported before incarceration, as well. Approximately 6% of individuals in state prisons experienced periods of homelessness (James & Glaze, 2006) and about 9% of incarcerated parents in state prisons reported homelessness within the year before their arrest (Glaze & Maruschak, 2008). Moreover, common characteristics of residential instability highlight the internal and external stressors delineated in the FSPP model. Aspects of housing instability such as the experience of evictions, overcrowded conditions, or difficulty paying rent are related to family stress and disruption (Desmond, 2016) and may result in adverse child outcomes as well. In contrast, stable housing offers families key support structures regarding safety and security, and is a positive factor that may aid in children’s adjustment.

Parental arrest and incarceration are commonly seen as stressors for the entire family unit and causes of family disruption (Turney & Wildeman, 2017). Some studies have found that parental arrest, whether or not it leads to incarceration, can initiate a negative cascade of events for children and families, regardless of the outcome of the arrest (Shlafer et al., 2013). Similarly, arrest of a parent, even if the parent is only absent from the family briefly, may have negative economic consequences for families because of the need to pay bail or fines (Grogger, 1995) and if family funds are limited, an individual may need to remain in jail without being convicted of any crime because of the inability to post bail, thus lengthening the separation between parents and their children (Wagner & Rabuy, 2017). In other cases, arrest can lead to conviction and jail or prison time, resulting in extended absences from children and continuing family disruption. Similar financial uncertainty is often accompanied by frequent residential locations, thus mobility and housing instability are more common in low-income families (Merrick et al., 2018). For example, from 2005 to 2010, residents below the poverty line had a 52.5% moving rate, compared to people at or above 150% of poverty with a moving rate of only 31.6% (Ihrke & Faber, 2012). Children in these hosueholds experience a range of negative emotions, including confusion, anger, and sadness, and they may begin to see the instability as a sort of norm (Kirkman et al., 2010). Residential mobility also predicts increased behavior problems such as aggression and acting out (Wildeman, 2010; Wildeman & Wakefield, 2014; Fowler et al., 2014) and seems to have a greater negative effect on child outcomes when children experience additional co-occurring risks, such as poverty (Ziol-Guest & McKenna, 2014). For these children, housing instability may occur for more negative reasons, such as job loss, deteriorating housing, strained household relationships, and foreclosure or eviction (Cohen & Wardrip, 2011). Evidence suggests that the negative implications of mobility may be greater at certain points of development. Findings from this work suggests that children who move frequently (i.e., three or more times) between birth and 5 years of age demonstrate problems with self-regulation (Schmitt et al., 2015) and early academic achievement (Schmitt & Lipscomb, 2016). Instability in the home environment causes disruptions for families that may be upsetting to young children, which may cause them to act out (Evans & Wachs, 2010). Young children have little to no say in these relocations, yet experience the consequences that are associated with them.

These difficulties may continue as children grow older. For example, on average, preschoolers exhibit higher levels of aggression (Geller et al., 2012; Wildeman, 2010) and school-age children and young adults exhibit more externalized and delinquent behavior in the aftermath of parental incarceration compared to their peers (Geller et al., 2012; Haskins, 2015; Porter & King, 2015). There is less understanding of why this occurs, however. It is theorized that children’s behavioral responses to parental incarceration can be shaped by disruptions occurring within their families, as well as their school and community environments (Dwyer Emory, 2018), although fewer studies have examined these relations empirically. Research exploring links between parental incarceration and child behavior problems have identified three consistent changes within the family that may contribute to the association: parental disengagement from the family, material hardship, and caregiver stress. Child reports of closeness with their father mediated the association between paternal incarceration and adolescent delinquent behavior (Porter & King, 2015). Additionally, families typically face direct and indirect costs of incarceration (which often contribute to already poor economic situations) (Geller, Garfinkel, & Western, 2011), such material hardship is associated with behavior problems in children (Zilanawala & Pilkauskas, 2012). Lastly, incarceration is associated with anxiety, stress, and depression for the non-incarcerated parent (Wildeman et. al., 2012) and distress in caregivers is a predictor of behavior problems in children (Turney, 2012). However, it should be highlighted that there is significant variability in children’s well-being when parents are incarcerated, and some children function quite well despite experiencing such risk (Poehlmann-Tynan & Arditti, 2017).

As previously stated, not all parental residential instability takes place with children. Although findings looking at the subgroup of the parents whose residential mobility takes place without children is scant, evidence from the non-custodial parenting literature can be applied to understand the effects of non-residential parental moves on children. Paternal criminal behavior has been linked to lower levels of involvement with children and to an increased risk of living apart from them altogether (Jaffee, et al., 2001); non-residential parenting, in turn, consists of lower levels of support compared to parents who live in the child’s home (Bastaits et al., 2012). Families in which one parent frequently changes residence receive less financial support (Braver et al., 2003), and children of divorced (non-incarcerated) parents who move frequently also show more signs of hostility in their interpersonal relations, suffer more distress, perceive their parents less favorably, show more ambiguity in continued communication and contact, and rate their personal and emotional adjustment less favorably (Braver et al., 2003). These findings show a trend towards negative outcomes for children whose parents frequently relocate even when the child’s housing is considered to be relatively stable. However, none of these studies focused on children with incarcerated parents.

Although previous research has found that family residential instability, material hardship, homelessness, and children’s behavior problems increase following a recent parental incarceration, studies have not examined the period just prior to a parent’s incarceration in any detail (Geller et al., 2009; Wildeman, 2014). Within the context of the FSPP, the current study highlights family factors associated with child adjustment by examining parental residential instability, homelessness, and absence as significant causes of family disruption. By focusing specifically on the disadvantage of housing instability, our study contributes significantly to the literature because of its specific emphasis on mobility profiles of children and families in the year prior to incarceration. In recognizing this, analyses explore a direct pathway from residential mobility to child outcomes, while also investigating the indirect pathways to child adjustment through parenting stress. To do so, the current study examined residential instability during the 12 months prior to incarceration, including periods of parental absence, and children’s behavior problems during the incarceration to address the following questions: (1a) For jailed parents with young children, how stable or unstable was the parent’s housing in the year leading up to their current incarceration in jail? (1b) Did residential instability involve prior incarcerations? (1c) Did residential instability involve periods of homelessness? (1d) Did residential instability or homelessness include children or involve parental absence? (2a) What are the predictors of residential instability in jailed parents with young children in the year prior to incarceration, accounting for pre-incarceration use of public assistance and parental education, age, gender and race? (2b) Do jailed parents who identify as their child’s primary caregiver experience less housing instability than those who indicate that they are not the child’s primary caregiver? (3a) Is jailed parent residential instability and homelessness related to children’s internalizing and externalizing behavior problems? (3b) Does parenting stress partially mediate this pathway?

Method

Participants

Data for this study were collected from 165 jailed parents in three midwestern counties in both rural and urban settings. Of the 165 jailed parents who participated in the study, 140 (84.8%) identified as men and 25 (15.2%) identified as women (Table 1). 44.8% percent of jailed parents identified themselves as Black, 33.3% White, 7.3% Latino and 14.6% multiple or other races. Incarcerated parents ranged in age from 18 to 49 years, with a mean of 29 (SD=5.83). Target children ranged from 2–6 years of age, with an average of 4 years (SD=1.31); children in that age range were selected because they are among the most common group of children affected by parental incarceration (Glaze & Maruschak, 2008; Murphey & Cooper, 2015) and it is an important time for family influences on children’s development. Of the children, 89 (53.9%) were boys and 73 (44.2%) were girls. The most commonly reported level of education for jailed parents was high school graduation or the equivalent (n=60, 36.4%), with some parents (n=3, 1.8%) reporting less than a 7th grade education and others (n=3, 1.8%) being college graduates. More than half of parents (56.4%, n=93) were employed prior to the current incarceration and 44.8% (n=74) received public assistance, with family income averaging just over $15,000 (SD= $18,533). In addition, 69.7% (n=115) of incarcerated parents reported living with their child prior to incarceration, with 47.9% (n=79) describing themselves as the child’s primary caregiver. Demographic information is displayed in Table 1. Jailed parents were incarcerated for drug-related charges (15%), probation violations (21%), battery/violence (13%), nonpayment of child support (15%), domestic dispute/domestic violence (17%), DUI or DWI (11%), and other crimes (theft, property damage) (8%). Of the 165 jailed parents who participated in the study, 3 (1.8%) had missing housing data and an additional 3 (1.8%) had missing child behavior checklist data. Prior to conducting regression analyses, however, we used a multiple imputation procedure (see Results section for a description).

Table 1.

Participant Demographic Information

Variable	N	%	Range	Mean ± S.D.
Jailed Parent Age	162	--	18–49	29.08 ± 5.83
Jailed Parent Gender
Father	140	84.8%	--	--
Mother	25	15.2%	--	--
Jailed Parent Race
African American	74	44.8%	--	--
Caucasian	55	33.3%	--	--
Latino	12	7.3%	--	--
Native American	3	1.8%	--	--
Other or Multiple Races	18	10.9%	--	--
Jailed Parent Education Level
Junior High School	3	1.8%	--	--
Partial High School	38	23%	--	--
High School Graduate	60	36.4%	--	--
Partial College	58	35.2%	--	--
College Graduate	3	1.8%	--	--
Pre-Incarceration Employment	93	56.4%	--	--
Receiving Public Assistance	74	44.8%	--	--
Income	162	--	$0 - $115,000	$15,377 ± $18,533
Primary Caregiver prior to Incarceration	79	47.9%	--	--
Child Age	162	--	2–6	4.05 ± 1.31
Child Gender
Boy	89	54.9%	--	--
Girl	73	45.1%	--	--

Notes: Not all cells add up to 165 because of missing data

Procedure

Recruitment efforts began with the jailed parent. Weekly, administrative staff at jails in three midwestern counties that represented diverse urban and rural populations, provided either the names of newly sentenced parents who had children between 2 and 6 years of age or access to a database with this information. Identified inmates then participated in a brief initial screening with a trained researcher to determine if they met research criteria indicating that they: (1) were at least 18 years old, (2) had a child who lived with kin within the county in which the inmate was serving time (or an adjacent county), (3) had retained legal rights to the child and had not committed a crime against the child, (4) had cared for the child at least part of the time prior to incarceration, (5) could understand and read English, and (6) had already been sentenced to serve jail time or were accused of committing a misdemeanor crime that would result in jail (rather than prison) time. If the incarcerated individual had more than one child in the age range, one child was randomly selected for participation in the study (termed “target child”). Those who met the inclusion criteria were invited to participate in the study, and those who agreed signed informed consent forms and participated in an interview, a vocabulary assessment, and self-administered questionnaires. The study was approved by the Institutional Review Board from our university and an NIH Certificate of Confidentiality was used.

Three jail systems participated in this research, all of which were run by county sheriff’s departments. The sheriff’s departments were in charge of both law enforcement and the jails in their counties. The first jail is located in a large urban community that experiences significant racial disparities in arrest and incarceration rates. The facility has an 823-bed capacity, with an average daily population of 788 inmates (21% women and 79% men). The second jail site is located in a rural county. The jail has a 458-bed capacity and in 2009, the jail had a daily count of 277 inmates (90% men, 10% women), although the daily count decreased during the study period. The third jail is located in an urban community and holds a mix of individuals from urban and rural locations. In 2010, the average daily population of the county’s three jails was 704.

Each jail added questions to their intake forms inquiring if the individual had children and the children’s ages. Incarcerated individuals who had children in the specified age range were approached. If they met inclusion criteria, the jailed parent was presented with a consent form, which was read aloud. Consented inmates were interviewed by a researcher in a private area within the cell block, with security staff nearby. We asked jailed parents about demographics, children’s living arrangements prior to and following incarceration, children’s experience of incarcerated-related events, and previous and current contact with children and the children’s caregivers. We were unable to compensate jailed parents for their study participation.

Measures

Housing instability and transitions.

The Residential Time-Line Follow-Back (RTLFB) Inventory was used to index jailed parents’ housing instability, transitions, and homelessness. Following Tsembris et al. (2007), the RTLFB was adapted to aid jailed parents in recalling their housing transitions and periods of homelessness from autobiographical memory. The RTLFB Inventory utilizes the timeline follow back interviewing method (that was originally developed to retrospectively assess substance abuse), including a calendar and memory cues, to elicit a retrospective housing history. The interviewer led the incarcerated individual through the RTLFB measure, asking parents to indicate their housing status each week of the 12 months leading up to their incarceration, highlighting transitions made from one form of residence to the next. At each transition, the incarcerated individual provided information about the living establishment, living arrangements, whether or not their children were with them, and the reason for leaving each living situation during the last 12 months. Incarcerated individuals were prompted by the interviewer to give greater detail in regard to their recent housing history. Responses were coded by living establishments (house/apartment, incarceration, homeless, or recovery center), living arrangement (alone, with family, with roommates, with partner, with children), and the reason for leaving.

Parental absence.

As it is important to differentiate between parental absence and housing instability, interviews with the incarcerated individuals included detailed questions about parent-child separations that occurred in the past year. We also asked more generally about any prior periods of parental absence that occurred for any reason, even for reasons other than incarceration, homelessness, or housing transitions. In addition, we collected data on the length of parent-child separations. However, data regarding parent-child communication during absences prior to the current incarceration were not collected.

Demographic and child and family characteristics.

Interviews were conducted with incarcerated parents in the jail. Jailed parents were asked about demographics, relationship status, family life, previous criminal activity, substance abuse, and previous trauma. They also answered questions pertaining to their children, which included their demographic information, current living situation, details of parent-child contact, and concerns about the children. Jailed parents were asked if they would consider themselves the primary caregiver of the target child; if the parent said “I don’t know,” it was coded as “no.” Responses were either coded into quantitative variables or left as qualitative data depending on the nature of the question. Parental race was binary coded (white, non-white).

Parenting stress.

Jailed parents reported on their parenting stress using a revised edition of the Parenting Stress Index (PSI-M) tailored for incarcerated parents (Loper et al., 2009; Tuerk & Loper, 2006). The PSI was established under the theoretical model of the determinants of dysfunctional parenting (Abidin, 1995). The goal of the index was to establish the relationship between the total stress that a parent experiences and certain child characteristics. The revised edition used in this study was adapted and reproduced with special permission of the publisher. In this revised edition, parents were asked to rate either strongly agree, agree, not sure, disagree, or strongly agree on topics such as: “I often have the feeling that I cannot handle things very well;” and “I feel capable and on top of things when I am caring for my child.” The responses were reverse coded as needed and combined into a summative score used in the analyses.

Child behavior problems.

Jailed parents completed either the Preschool Child Behavior Checklist (CBCL; Achenbach & Rescorla, 2000) for target children between 2 and 5 years or the School-Age Child Behavior Checklist (CBCL; Achenbach & Rescorla, 2001) for children who were 6 years of age. The CBCL is a widely-used standardized behavior rating that is completed by an adult with whom the child lives, and it has high internal consistency (Cronbach’s αs range from .78 to .97). The CBCL scales are appropriate for children ages 18 months to 18 years. The preschool form lists 99 problem behaviors, whereas the school-age form lists 113 problem behaviors. Jailed parents rated each behavior on a scale of 0 (not true), 1 (somewhat or sometimes true), or 2 (very true or often true) in reference to the last 2 months. Responses are then summed and converted into T-scores based on children’s age and gender to obtain scores for Internalizing and Externalizing Problem Scales. Although raw scores are generally preferred when analyzing CBCL data, this is not possible when multiple forms (based on children’s age) are used. Thus, T-scores were used in analyses. High scores on the internalizing and externalizing subscales indicate more problematic behaviors; T-scores of 63 and above are considered in the clinical range and scores between 60 and 62 indicate subclinical range. In the present study, means were relatively high (internalizing problems T-score M = 51.60, SD = 10.38 and externalizing problems T-score M = 50.28, SD = 10.86), resulting in 22.64% of children (n=36) scoring in the subclinical or clinical range for internalizing scores and 18.24% (n=29) with subclinical or clinical range externalizing scores. (These results are broken down by number of moves and length of time that the incarcerated parent spent with the child in the year prior in Table 3.)

Table 3.

Clinical Child Behavior Problems by Number of Parental Moves Versus Total Months that the Incarcerated Parent Was Absent From the Child in the Previous Year (N=159)

Child Behavior Problems and Number of Parental Moves (Including Current Incarceration)
	1 move (stable, then jail)	2 moves (modest stable, then jail)	3 moves (unstable, then jail)	4 or more moves (highly unstable, then jail)
Internalizing Problems
Non-clinical range	33 (72%)	35 (95%)	27 (79%)	28 (67%)
Borderline or Clinical Range	13 (28%)	2 (5%)	7 (21%)	14 (33%)
Externalizing Problems
Non-clinical range	36 (78%)	35 (95%)	29 (85%)	30 (71%)
Borderline or Clinical Range	10 (22%)	2 (5%)	5 (15%)	12 (28%)
Total Behavior Problems
Non-clinical range	35 (76%)	35 (95%)	29 (85%)	29 (69%)
Borderline or Clinical Range	11 (24%)	2 (5%)	5 (15%)	13 (30%)
Total n	46	37	34	42
Child Behavior Problems and Total Months Parent was Absent From the Child in Previous Year (Could be Continuously or Intermittently Absent)
	0–2 months	3–5 months	6–10 months	11–12 months
Internalizing Problems
Non-clinical range	45 (78%)	5 (50%)	9 (83%)	64 (80%)
Borderline or Clinical Range	13 (28%)	5 (50%)	2 (17%)	16 (20%)
Externalizing Problems
Non-clinical range	48 (83%)	7 (70%)	10 (83%)	65 (79%)
Borderline or Clinical Range	10 (17%)	3 (30%)	1 (17%)	15 (20%)
Total Behavior Problems
Non-clinical range	46 (79%)	7 (70%)	9 (83%)	66 (82%)
Borderline or Clinical Range	12 (20%)	3 (30%)	2 (17%)	14 (18%)
Total n	58	10	11	80

Note: Three behavior checklists were missing, and thus the total number of children included in this table is 159.

Results

Of the 162 jailed parents who had complete housing timeline data, most (n=115; 71.5%) reported one or more changes in residence, not including their recent stay in jail. Jailed parents reported moving one to nine times in the year prior to the current incarceration, with a mean of 2.7 moves. About half of all the jailed parents interviewed (n=77, 47.5%) reported at least two moves within the last year in addition to their jail stay, and about in 1 in 4 (n=43, 26.5%) jailed parents reported 4 or more moves for the year, not including the current jail stay.

In our descriptive analyses, we differentiated between reasons for housing instability and frequency and length of parental absence (see Table 2). About 1 in 3 jailed parents (n=53, 32.1%) reported that they had at least one other incarceration within the past 12 months, and an additional 13 stayed in a halfway house or residential treatment center. By their very nature, these two transitions did not involve children and thus constituted parental absence. For the 53 parents who reported an additional incarceration in the previous year, 30% (n=16) indicated that it lasted for about a month, but for nearly a quarter of parents (n=13, 24.5%) the incarceration lasted for more than 6 months. Looking beyond the previous year, nearly 90% (n=148) of jailed parents in the study reported a history of incarceration in jail or prison prior to their current jail incarceration.

Table 2.

Jailed Parents’ Experiences of Housing Instability and Parental Absence in the Past Year (N=162)

a. Overall Parental Housing Instability
Frequency of Jailed Parents’ Moves With or Without Children, Including the Current Incarceration (n=162)
Reason for Move	1 move (stable, then jail)	2 moves (modest stable, then jail)	3 moves (unstable, then jail)	4 or more moves (highly unstable, then jail)
Prior Incarceration	16	9	4	24
Recovery or Halfway House	8	4	1	0
Homelessness	5	1	5	11
Apartment or Home	18	24	24	8
Total	47	38	34	43
b. Housing Instability and Parental Absence
Frequency of Jailed Parents’ Moves Without Children, Not Including the Current Incarceratio n (n=63)
Reason for Move	0 moves (stable, then jail)	1 move (modest stable, then jail)	2 moves (unstable, then jail)	3 or more moves (highly unstable, then jail)
Prior Incarceration	16	9	4	24
Recovery or Halfway House	8	4	1	0
Homelessness	0	1	4	2
Apartment or Home	0	4	3	7
Total	24	18	12	33
c. Housing Stability With Children
Jailed Parents with Stable Housing, Living With Children (n=23)
d. Housing Stability and Parental Absence
Jailed Parents with Stable Housing, Living Apart from Children (n=76)

In contrast to prior incarceration or residential treatment, parental homelessness or moves to an apartment or home may or may not have involved children. Of the 115 jailed parents who experienced housing transitions, 52 (45.2%) reported that the moves included their children (and thus did not involve parental absence). In other words, the parent-child relationship remained stable, but the child moved with the parent, which could also be challenging. Of the 115 jailed parents who reported at least one move in the 12 months prior to their jail stay, 22 (19.1%) reported experiencing periods of homelessness. Of those experiencing homelessness, parents reported a range of 1 to 12 months of homelessness, with a mean of 3.5 months (SD=3.2). Of the 22 jailed parents who reported that they were homeless at least one of the 12 months prior to their current incarceration, 7 (32%) indicated that their child (or children) were homeless with them. The affected children experienced between one and 10 months of homelessness prior to the parent’s incarceration. Parents reported that sometimes the family lived in a car, engaged in couch surfing with friends, stayed in a shelter (although this was not common as many shelters do not take children), or spent time on the street. On average, affected children experienced 3.1 months of homelessness (SD = 3.2).

On a relatively positive note, 23 (14.2%) jailed parents reported continuously coresiding with the target child in the year prior to the current incarceration, and thus the children did not experience residential instability or parental absence prior to the current jail stay. On a less positive note, however, an additional 35 (21.6%) jailed parents reported living with the child fairly consistently over the previous 12 months, with absences ranging from one to two months. Yet most (n=104, 64.2%) incarcerated parents reported two or more months during the past year when they were absent from the child, although this does not imply lack of contact or communication. Sixty-three jailed parents reported housing instability and parental absence in the past year, whereas 76 jailed parents reported housing stability and parental absence in the past year. In looking back further than 12 months prior to the current incarceration, two-thirds of jailed parents (n=108, 65.5%) reported a previous separation from the child. Lengths of these separations ranged from as little as 4 days up to 5 years, with a median length of 3-month separations. Reasons for the separations varied, with 66% occurring as a result of previous incarceration(s) and 13% occurring as a result of conflict with or separation from the child’s caregiver. Table 3 provides a breakdown of clinical levels of children’s behavior problems according to the number of parental moves and total number of months that parents were absent from the child’s residence during the prior year. Total months living apart from the child, however, masks the number of transitions that children experienced regarding living with their parents, as the months absent could occur continuously or intermittently.

A multiple linear regression analysis was then conducted to examine predictors of the number of moves parents experienced in the year prior to their incarceration in jail. Prior to conducting regression analyses, a multiple imputation procedure was used to address missingness, which appeared to occur at random (Raghunathan, Lepkowski, van Hoewyk, & Solenberger, 2001; Van Buuren, 2007). The procedure involved generating five datasets in which missing values were randomly produced conditional upon other variables in the analysis. Subsequent analyses were applied to all five datasets, with aggregated results reported. (Findings were similar in the original and aggregated data sets). Predictors entered in the first step of the regression model were jailed parent characteristics (age, education, pre-incarceration use of public assistance, race, gender). In the second step, variables entered were months that the jailed parent had spent living with the child in the past 12 months, whether or not the parent experienced previous separations from the child, whether or not the jailed parents had served as the child’s primary caregiver, and if there had been previous incarcerations in the year. Results are shown in Table 4a. These findings indicated that parents who had experienced an additional incarceration in the past year experienced more moves overall. In addition, Black parents reported fewer moves than White parents. Parental gender was not significant in the analysis.

Table 4a.

Multiple Regression Results Predicting Number of Parental Moves (N=165)

		Unstandardized Coefficients		Standardized Coefficients
Model	Variable	B	SE B	Beta	t	p
Step 1	JPa Age	−0.005	0.022	−0.020	−0.249	0.803
	JP Education	0.003	0.155	0.001	0.017	0.987
	JP Public Assistance	0.231	0.276	0.071	0.839	0.403
	JP Race	−1.184	0.282	−0.340*	−4.195	0.000
	JP Gender	0.303	0.420	0.063	0.721	0.472
Step 2	JP Age	−0.015	0.018	−00.055	−0.845	0.400
	JP Education	0.016	0.128	.008	0.122	0.903
	JP Public Assistance	0.151	0.225	0.046	0.668	0.505
	JP Race	−0.838	0.231	−0.241*	−3.632	0.000
	JP Gender	0.107	0.348	0.022	0.306	0.760
	Months Lived with Child	−0.035	0.022	−0.117	−1.631	0.105
	Previous P-Cb Separations	0.147	0.234	0.043	0.628	0.531
	Child’s Primary Caregiver	−0.364	0.240	−0.113	−1.519	0.131
	One or more Additional Incarceration(s)	1.689	0.233	0.493*	7.262	0.000

Notes:

^aJailed Parent.

^bParent-Child.

^*p<.05

A binary logistic regression analysis was conducted to examine predictors of any parental homelessness in the year prior to incarceration in jail. Predictors entered in the first step of the model were the same as above (jailed parent age, education, pre-incarceration use of public assistance, race, gender). In the second step, variables entered were again the same as above (months that the jailed parent lived with the child in the past 12 months, previous parent-child separations, whether or not the jailed parents had served as the child’s primary caregiver, and if there had been previous incarcerations in the year). In the first step, jailed parent age and race were significant and jailed parent education was a trend, with older parents, White parents, and parents with less education more likely to be homeless. In the second step, parents who had not served as primary caregivers for their children more likely to be homeless, as were parents who had experienced an additional incarceration in the past year, as displayed in Table 4b. Parental gender was not significant in the analysis.

Table 4b.

Logistic Regreession Results Predicting Parental Homelessness in Year Prior to Current Jail Incarceration (N= 165)

	Wald
Predictors, Step 2	B	S.E.	(df=1)	Sig	Exp(B)
JPa Age	0.094	0.045	4.396*	0.036	1.099
JP Education	−0.615	0.317	3.763+	0.052	0.541
JP Public Assistance	−0.320	0.591	0.293	0.588	0.726
JP Race	1.600	0.618	6.694*	0.010	4.953
JP Gender	−0.223	0.757	0.087	0.768	0.800
Months Lived with Child	0.031	0.056	0.315	0.575	1.032
Previous P-Cb Separations	−0.184	0.644	0.081	0.776	0.832
Child’s Primary Caregiver	1.519	0.629	5.830*	0.016	4.566
One or more Additional Incarceration(s)	−1.350	0.586	5.304*	0.021	0.259

Notes:

^aJailed Parent.

^bParent-Child.

^*p<.05,

⁺p<.10

To examine if parental housing instability and homelessness related to young children’s behavior problems, two multiple linear regression analyses were conducted, one for internalizing and one for externalizing problems. In each analysis, the first step included jailed parent jailed parent age, education, pre-incarceration use of public assistance, race, gender, months that the jailed parent lived with the child in the past 12 months, previous parent-child separations, whether or not the jailed parents had served as the child’s primary caregiver, and if the jailed parent had been incarcerated previously in the year. The second step included the total amount of housing instability the parent experienced in the past year and whether or not the parent experienced homelessness in the past year. Results indicated that for the externalizing problems but not the internalizing problems model, more parental housing instability significantly predicted higher levels of child behavior problems, as displayed in Table 5. However, control variables, including parent gender, and parental experiences of homelessness in the prior year did not predict child behavior problems.

Table 5.

Jailed Parent Housing Instability, Homelessness, and Parental Absence as Predictors of Child Behavior Problems (N=165)

Externalizing Problems
	Unstandardized Coefficients		Standardized Coefficients
Predictors, Step 2	B	SE B	Beta	t	p
JPa Age	−0.082	0.164	−0.045	−0.498	0.619
JP Education	−1.320	1.161	−0.103	−1.137	0.258
JP Public Assistance	1.063	2.061	0.049	0.516	0.607
JP Race	0.883	2.039	0.041	0.433	0.666
JP Gender	−3.405	3.158	−0.107	−1.078	0.283
Months Lived with Child	0.128	0.200	0.064	0.641	0.523
Previous P-Cb Separations	−0.966	2.126	−0.042	−0.454	0.650
Child’s Primary Caregiver	0.203	2.249	0.009	0.090	0.928
Number of JP Housing Transitions	1.664	0.735	0.250	2.264*	0.025
JP Homelessness	−4.110	3.436	−0.123	−1.196	0.234
Internalizing Problems
	Unstandardized Coefficients		Standardized Coefficients
Predictors, Step 2	B	SE B	Beta	t	p
JPa Age	−0.039	0.153	−0.023	−0.255	0.799
JP Education	−0.866	1.079	−0.073	−0.803	0.423
JP Public Assistance	−2.312	1.915	−0.116	−1.207	0.229
JP Race	0.129	1.894	0.006	0.068	0.946
JP Gender	2.104	2.934	0.072	0.717	0.475
Months Lived with Child	0.064	0.186	0.035	0.346	0.730
Previous P-Cb Separations	−.0471	1.976	−0.022	−0.238	0.812
Child’s Primary Caregiver	0.048	2.090	0.002	0.023	0.982
Number of JP Housing Transitions	1.063	0.683	0.174	1.556	0.122
JP Homelessness	−3.79	3.193	−0.124	−1.190	0.236

Notes:

^aJailed Parent.

^bParent-Child.

^*p<.05

To examine if parenting stress accounted for or partially mediated the relation between parental housing instability and young children’s behavior problems, two additional multiple linear regression analyses were conducted, one for internalizing and one for externalizing problems. As in the analyses above, the first step included jailed parent characteristics (age, gender, education, pre-incarceration use of public assistance, and race), whether or not the parent had served as the child’s primary caregiver, months that the parent had spent living with the child in the past year, and whether or not the parent experienced previous separations from the child. The second step included the total amount of housing instability the parent experienced in the past year and whether or not the parent experienced homelessness in the past year, and the third step included the parenting stress score. Results indicated that the parenting stress measure did not predict internalizing or externalizing child behavior problems, as displayed in Table 6. Again, control variables, including parent gender, and parental experiences of homelessness in the prior year did not predict child behavior problems, but housing instability predicted child externalizing problems.

Table 6.

Jailed Parent Housing Instability, Homelessness, Parental Absence, and Parenting Stress as Predictors of Child Behavior Problems (N=165)

Externalizing Problems
	Unstandardized Coefficients		Standardized Coefficients
Predictors, Step 3	B	SE B	Beta	t	p
JPa Age	−0.096	0.165	−0.053	−0.579	0.564
JP Education	−1.260	1.165	−0.098	−1.081	0.282
JP Public Assistance	1.145	2.066	0.053	0.554	0.580
JP Race	0.633	2.066	0.029	0.306	0.760
JP Gender	−3.035	3.196	−0.096	−0.950	0.344
Months Lived with Child	0.133	0.200	0.067	0.665	0.507
Previous P-Cb Separations	−1.150	2.142	−0.050	−0.537	0.592
Child’s Primary Caregiver	0.572	2.299	0.026	0.249	0.804
Number of JP Housing Transitions	1.616	0.738	0.243	2.189*	0.030
JP Homelessness	−3.977	3.445	−0.119	−1.154	0.251
JP Parenting Stress	−0.057	0.071	−0.074	−0.800	0.425
Internalizing Problems
	Unstandardized Coefficients		Standardized Coefficients
Predictors, Step 3	B	SE B	Beta	t	p
JPa Age	−0.055	0.154	−0.033	−0.357	0.722
JP Education	−0.798	1.081	−0.068	−0.738	0.462
JP Public Assistance	−2.218	1.918	−0.111	−1.157	0.250
JP Race	−0.155	1.917	−0.008	−0.081	0.936
JP Gender	2.525	2.966	0.087	0.852	0.396
Months Lived with Child	0.070	0.186	0.038	0.377	0.707
Previous P-Cb Separations	−0.680	1.988	−0.032	−0.342	0.733
Child’s Primary Caregiver	0.468	2.133	0.023	0.219	0.827
Number of JP Housing Transitions	1.008	0.685	0.165	1.472	0.144
JP Homelessness	−3.648	3.197	−0.119	−1.141	0.256
JP Parenting Stress	−0.064	0.066	−0.092	−0.980	0.329

Notes:

^aJailed Parent.

^bParent-Child.

^*p<.05

Discussion

In this study of jailed parents with young children, we examined family disruption prior to and during parental incarceration in the form of parental residential instability, homelessness, and parental absence. We found that recent parental residential instability, including moves with and without children, were common. Moreover, jailed parents’ housing instability, whether or not it involved absence from children, related to parent-reported externalizing behavior problems in young children, and parenting stress did not mediate this association. These findings highlight challenges that children and their families face in the period leading up to an incarceration in jail and provide more detail about how residential instability, recidivism, and homelessness overlap, including relations to child adjustment.

Prior research has found a higher rate of residential instability in families of children with incarcerated parents compared to those without incarcerated parents (Geller, Garfinkel, & Western, 2011), whereas in the present study, we focused on the number and types of moves within a sample of jailed parents with young children. We found that the vast majority of incarcerated individuals saw changes in residence and nearly half saw more than 3 moves in a year. These numbers are high compared to averages found in data collected through the national census. The Migration/Geographic Mobility Data from 2013–2014 found that about one in nine individuals reported a move within a one-year period (U.S. Census Bureau, 2015).

More housing instability in the year before the parent’s current incarceration in jail related to elevations in children’s externalizing behavior problems but not internalizing problems. Similarly, previous research has found that when frequent moves occur within impoverished neighborhoods, children exhibit higher dysregulation when compared to children who live in high-poverty neighborhoods but remain in stable housing (Roy et. al., 2014). Studies using the FFCW sample have found that in addition to increased residential mobility, families with incarcerated parents also face significant economic and family instability (Geller et. al., 2011) and children with incarcerated parents exhibit more behavior problems than their peers, especially externalizing problems (Wildeman & Wakefield, 2014). These findings suggest that incarceration’s destabilizing effects may carry over into the child’s lives, as parents tend to earn less and often live away from their children, diminishing financial support and family ties (Geller et. al., 2009). These family stressors have implications for child functioning as both parental incarceration and residential instability may result in stress and distress expressed as acting out or externalizing behavior problems (Casey et. al., 2015). Children who experience shifts in housing have more difficulty regulating their attention, cognition, and behavior compared to their non-mobile peers, with stressors accumulating as housing transitions increase in frequency (Roy et. al., 2014). These findings suggest a cascade of family processes that can lead to challenges in child adjustment prior to and following the incarceration of a parent. Although parental absence per se (i.e., the inverse of the number of months the parents lived with the child) did not relate to child behavior problems, residential instability related to externalizing but not internalizing problems in this jailed parent sample. It is possible that parents who spend less time with their children may observe and choose to report on the more overtly manifested behavior problems, as opposed to ones that are less apparent and involve child withdrawal or loneliness. Relations among housing moves and externalizing problems may be complex, and additional data from caregivers or teachers may help tease apart how much of the relation is caused by parental perceptions of child behavior when they are frequently absent. In the future, the FSPP model could help researchers unpack some of the family processes impacting a child’s adjustment to parental incarceration. One area that could be explored more in future research is identifying when a parental move with children in tow can provide a secure parent-child relationship that can help children cope with stress, and when such moves create trauma or chaos in children’s lives. Both types of family disruptions may lead to family stress, as conceptualized in the FSPP model, but individual, family, and contextual factors may be protective. Future research should also emphasize the paradox of non-residential parents reporting on behavior concerns of their children, recognizing that parental absence is a possible factor affecting their children but particularly noting the mechanisms by which this assoication opperates.

A strength of the current study is that it focused on jailed parents, as the vast majority of U.S. incarceration occurs at the jail level, with approximately 10.9 million admissions to U.S. jails in 2015, and a daily average of 721,300 people in jails, compared to a total of 1.5 million individuals in state or federal prisons at year-end 2015 (Carson & Anderson, 2016; Minton & Zeng, 2016). Our findings support the idea that many jail incarcerations appear cyclic in nature or involve repeated brief stays in jail, which we term “parental churning” in the criminal justice system (Poehlmann-Tynan et al., 2018). Such churning may cause increased family instability and residential mobility compared to confinement in prison, which in 97% of cases involves sentences of more than one year (Wagner & Rabuy, 2017). Jail churn is particularly high because most people in jails have not been convicted; some have just been arrested while others are too poor to pay bail and must remain behind bars until their hearing or trial. Only a small number of people in jail (187,000 on any given day) have been convicted, generally serving misdemeanor sentences under a year (Wagner & Rabuy, 2017). This type of churning is likely to lead to family disruption and negatively affect children’s development—here, their externalizing behavior problems, which may include aggression, impulsive behavior, and acting out.

Recidivism is a critical issue for those incarcerated in jails and prisons and their families. For example, a 2005 study tracking 400,000 imprisoned individuals in 30 states found that nearly two-thirds of released individuals were rearrested (Durose, Cooper, & Snyder, 2014). Durose et al. also found that more than a third (36.8%) of prisoners who were arrested within 5 years of initial incarceration were arrested within the first 6 months following release, and more than half (56.7%) were arrested in the first year (Durose et al., 2014). In terms of recidivism, incarcerated parents with minor children are not statistically different from other incarcerated adults (Gricius et al., 2016). Additionally, previous research points to no differences in rates of recidivism between incarcerated mothers and fathers (Gricius et al., 2016). Findings in the current study of jailed parents with young children are similar. Nearly all jailed parents reported previous incarcerations, with a third recidivating from a previous incarceration within a year. While the length of incarcerations varied across individuals, nearly a quarter of jailed parents reported a prior incarceration within the year that lasted longer than 6 months. This parental “churning” in and out of corrections facilities is a form of residential instability and parental absence that has implications for children.

Residential mobility plays a key role in recidivism, as living in a more residentially stable household is associated with a lower risk of recidivism (Tillyer & Vose, 2011). Incarcerated individuals who return to neighborhoods with higher concentrations of ex-offenders are more likely to recidivate than individuals returning to neighborhoods with less crime (Stahler et. al., 2013). In addition, Brown and Bloom (2009) identified an increase in the likelihood of recidivism for incarcerated parents related to the strain of caring for a child following the parents’ release (i.e., financial burdens, time constraints, safety concerns). However, incarcerated parents may have a unique mindset regarding recidivism. Children often served as a “hook for change,” meaning that the incarcerated parent realizes the importance of raising children and expresses a desire to be a good parent and change their criminal behavior (Giordano, Cernkovich, & Rudolph, 2002). And as Parke and Clark (2001) state, “if recidivism can be reduced, children will be spared the trauma of repeated separation which, in turn, will improve their psychological adjustment”. The FSPP model views parental incarceration through a similar lens, especially regarding proximal caregiving and child adjustment. In cases of recidivism, the FSPP model emphasizes the repeated grieving over another ambiguous loss which may be further complicated by stigma, lack of social validation, or loneliness.

These findings also contribute to the literature that links incarceration in prison or jail with homelessness. Previous studies have found that 23–33% of single residents of homeless shelters reported a previous incarceration (Metraux & Culhane, 2006; Geller & Curtis, 2011). Prior research has also found that 6–9% of imprisoned individuals report having been homeless or highly mobile at some point in the year leading up to their most recent incarceration (James & Glaze, 2006, Glaze & Maruschak, 2008). Although recent reports show that the U.S. homelessness rate recently decreased from 18.3 people per 10,000 to 17.7 per 10,000 after decades of increases (Homelessness in America, 2016), for incarcerated individuals as well as families with young children, this does not seem to be the case. Indeed, a history of homelessness is 7.5 to 11.3 times more common among incarcerated individuals than the general population (Greenberg & Rosenheck, 2008). Other studies looking at the co-occurrence of parental incarceration and family homelessness find that nearly one third of homeless children also experience the incarceration of a parent (Casey et. al., 2015). In the current study, of the 115 jailed parents who reported housing transitions in the 12 months prior to their jail stay, nearly 20% reported experiencing periods of homelessness. Of the 22 jailed parents experiencing homelessness, 32% indicated that their children were homeless with them, with an average in these families of 3 months of child homelessness in the year. Many children of incarcerated parents and children experiencing homelessness face adversity. Children experiencing homelessness report lower levels of social support and weaker informal supports that could play a crucial role in diminishing risks among families living in poverty (Wildeman, 2014). Moreover, child maltreatment and problematic health outcomes are key consequences of child homelessness; homeless children tend to be exposed to infectious diseases at higher rates than other children (Wildeman, 2014; Rafferty et al. 2004). The FSPP model views homelessness as another contextual factor that further adds to the stress of parenting pre- and post-incarceration.

In sum, the present study found that housing instability is common in the year prior to a parent’s incarceration in jail, and although it sometimes includes the parents’ young children, more often than not, children of jailed parents repeatedly experience parental absence. Moreover, frequency of jailed parent housing transitions, even controlling for parental absence, is associated externalizing behavior problems in young children with currently jailed parents. Although incarcerated parent gender may be an important consideration regarding children’s risk exposure and effects on children’s overall adjustment, there were no gender effects in the present study. Additional research is needed on children with jailed fathers and mothers in the context of parental housing instability, especially because our study had a small sample of jailed mothers, likely underpowering our analyses for identifying effects of parental gender.

The findings of the present study have implications for future research with children of incarcerated parents and their families. When considering an incarcerated individual’s housing history and separation from children, it is important to gather detailed information over time rather than contrasting “ever incarcerated” parents and their children with “never incarcerated” parents and their children. Designing a longitudinal study of children and families with detailed information about the sequence, length, and timing of parental incarceration, housing instability, parental absence, and parent-child contact in relation to children’s development would allow researchers to follow participants during the parental incarceration and following release, providing a more accurate representation of the transitions made in housing, neighborhoods individuals return to, any instances of homelessness, and incidence and frequency of recidivism. Upon recruitment and initial data collection, a Housing Timeline Follow-Back measure could be used in order to determine residential mobility in individuals in the periods between incarcerations but also when individuals do not recidivate. It would also allow more nuanced tracking of when children live with their parents and when they are separated from them.

Further testing and development of the FSPP model can provide insights into the lived experiences of children with incarcerated parents. The model views family coping with parental incarceration as influenced by internal and external stressors, and our study points to housing instability and incarceration as being closely interwoven, with parental churning in and out of jail identified as a significant problem for children, that may help explain reasons for elevated externalizing behavior problems in children of incarcerated parents, a consistent finding in the literature (e.g., Haskins, 2015). It will be important to measure additional parenting and family stress processes in future studies, combining measures of parental, child, and caregiver perceived stress with physiological measures of stress hormones to further test the model’s predictions. Delineation of examples of internal and external stressors and experiences of ambiguous loss common among families experiencing incarceration would be useful, as would adding assessments of how families talk to children about the parent’s incarceration and other reasons that parents are absent from their children, which may contribute to children’s experience of ambiguous loss.

Limitations

When interpreting our findings, one should keep in mind the study’s limitations, which include challenges related to measurement (reliance on jailed parent self-report data, retrospective reports of housing, no information regarding parent-child contact during past separations), study design (within-group and non-longitudinal), and generalizability (focus on jailed parents, relatively small sample size). Although self-reports provide an important perspective into the lives of incarcerated parents and their families, they also come with methodological concerns such as shared method variance. Issues surrounding image management, introspective reflection, and response bias may emerge as incarcerated individuals report on stressful and vulnerable times within the past year. Retrospective reports of housing, while used in other studies, can be subject to memory issues (especially when substance use is involved) as well as social desirability. In addition, jailed parents may not have been the most accurate reporters of children’s behavior problems since incarceration because they were not living with the children. Measuring stress and coping related to ambiguous loss is challenging as well, and our measures of parenting stress and child behavioral adjustment are unlikely to have captured the construct.

An additional weakness is that the current study focused on assessing retrospectively reported housing timelines of incarcerated individuals for the 12 months leading up to their current stay in jail. While this design provides important insights into residential instability with and without children prior to incarceration, the study did not follow the incarcerated parents once released, which could add key pieces of information including how reunification with children and families proceeds and if they recidivate again shortly thereafter. Moreover, our analyses focusing on predictors of child behavior problems attempted to differentiate between parental residential instability and parental absence, as these are overlapping but distinct issues that can cause stress in children for different reasons. Clearly both are causes of family disruption as discussed in the FSPP model and the overlap and non-overlap should be assessed carefully. Future research also should include assessments of parent-child contact and communication during parental absences, as that is missing from much of the research focusing on children with incarcerated parents. Finally, the present study focused on jailed parents and their families rather than contrasting them with a group who never experienced parental incarceration or who had experienced prior incarceration only (not current incarceration). In this way we were able to focus in detail on issues related to parental incarceration and identify within-group variation; however, we are not able to generalize the findings to children and families who have not experienced incarceration. Previous studies have also focused on urban samples, whereas the present study focused on families from rural and urban locations.

Despite these limitations, the study provided insights into the proximal processes that occur when parents of young children are incarcerated, including linking parental absence from children with more parental residential instability and more parental residential instability with elevations in child externalizing behavior problems.