Risk Marker or Risk Mechanism? The Effect of Family,Household,and Parental Imprisonment on Children and Adults’ Social Support and Mental Health

Abstract

There is robust evidence of associations between parental imprisonment (PI) and a variety of harms to children, but the consequences of other forms of family imprisonment are largely unknown. Using Household, Income, and Labour Dynamics in Australia (HILDA), a nationally representative Australian data set, this article looked at the direct effects of PI, household member imprisonment (HI), or close family member imprisonment (CFI) on the social support and mental health of nonincarcerated adults and young people. Recent PI, HI, or CFI had no association with social support. Recent CFI did increase men’s risk of poor mental health, but not women’s or young people’s. We consider the implications of these findings in the context of strong negative effects of paternal imprisonment on mothers in the United States.

Keywords

parental imprisonment family imprisonment mental health social support prison

Parental imprisonment (PI) is associated with a variety of long-term harms to children across their life course (Foster & Hagan, 2007, 2015a; R. D. Lee, Fang, & Luo, 2017). Recent research found that PI might also cause severe stress to children’s adult caregivers, increasing risks of maternal depression and marital dissolution (Apel, 2016; Murray, Farrington, & Sekol, 2012; Turney, 2014, 2015). A recent theoretical framework by Foster and Hagan (2015b) proposes that the cumulative effects of PI on caregivers and children may result in children’s lifelong social exclusion. In their model, they emphasize parents’ psychological adjustment and children’s own emotional well-being as focal pathways that mediate these intergenerational effects. However, until now, few existing studies have been able to rigorously investigate any of these theorized psychosocial processes in the context of parental incarceration (Arditti, 2012; Auty, Farrington, & Coid, 2015; Luther, 2015).

Across the range of psychological processes that might be directly affected by PI, mental health, and social support are likely to be of chief importance. First, social support is thought to moderate the mental health consequences of stressful events such as family imprisonment (Arditti, 2005; Capowich, Mazerolle, & Piquero, 2001; Cohen & Wills, 1985; Cullen, 1994; Lösel, Pugh, Markson, Souza, & Lanskey, 2012; Murray et al., 2012). Conversely, mental health may influence people’s perceptions regarding the adequacy of their social support networks and their inclination to seek more support (Velez et al., 2016). In addition, mental health problems in caregivers may subsequently affect children’s ability to maintain supportive social relationships (Weinfield, Sroufe, & Egeland, 2000). An important step in investigating these relationships is to first study the effect of imprisonment on each of these outcomes in individuals who have indirectly experienced incarceration through a family member.

Until now, a global lack of large-scale, comprehensive data sources with information on criminal justice contact and family life has impeded research on the consequences of imprisonment for families (Wakefield, Lee, & Wildeman, 2016). Given that imprisonment disproportionally affects poor and disadvantaged subgroups of the population (Wakefield & Uggen, 2010), it is important to assess the extent to which selection effects may explain negative associations between PI and poor child outcomes into adulthood. Currently, however, only a few studies have been able to distinguish whether adverse outcomes found in children and caregivers are uniquely attributable to the consequences of the parent’s incarceration, rather than to preexisting adverse circumstances (E. I. Johnson & Easterling, 2012; Wildeman, 2014a). The question, therefore, remains whether adverse effects of PI are primarily attributable to preexisting risk (a “risk marker”) or to direct effects caused by imprisonment itself (“a risk mechanism”).

In addition, the few studies that did find robust evidence of any long-term effects of PI on children or their caregivers tend to be situated in the United States (E. I. Johnson & Easterling, 2012, 2013; Wildeman, 2014b). There are good reasons to expect that such results may not be generalizable to children and adults living in countries with lower incarceration rates (Wildeman, 2016). In Australia, where imprisonment is used more rarely, prison populations may be composed of individuals with higher levels of preexisting problems such as addiction or violent behavior. In these circumstances, adults and young people are more likely to benefit from the removal of a problematic family member. The effects of imprisonment in Australia may also be mitigated by more comprehensive welfare support and public services. Finally, Australia has shorter average sentence lengths, which may be associated with lower levels of harm.

A further research gap pertains to the types of affected families included in these studies. Although recent Australian research shows that children affected by the imprisonment of nonparent household members are at least as disadvantaged as prisoners’ own children (K. L. Besemer & Dennison, in press), there is very little empirical evidence regarding the risks associated with the imprisonment of any household members and/or close family other than a parent (Meek, 2008; Meek, Lowe, & McPhillips, 2010; Wildeman & Wakefield, 2014). However, a small number of studies around the world have found evidence that suggests that nonparent family member incarceration may affect children in ways that mirror the consequences of PI. For example, the imprisonment of children’s siblings, fathers, uncles, aunts, and grandparents have all been shown to predict their subsequent delinquent development (Farrington, Barnes, & Lambert, 1996; Farrington, Jolliffe, Loeber, Stouthamer-Loeber, & Kalb, 2001; Slomkowski, Rende, Conger, Simons, & Conger, 2001). In addition, qualitative research observed adverse psychological consequences of sibling imprisonment for other children in the home (Meek, 2008; Meek et al., 2010). Qualitative work by Comfort (2016) has described repeated short prison sentences among close family members as a severe, on-going strain, particularly for people who have to cope with the imprisonments of multiple different relatives. The imprisonment of close family members has also been shown to result in risks to adult cardiovascular health (H. Lee, Wildeman, Wang, Matusko, & Jackson, 2014). Similarly, mothers confronted with the imprisonment of adult sons suffer from increased risks of psychological distress and decreased physical health (Green, Ensminger, Robertson, & Juon, 2006). This research suggest that nonparental family relationships to prisoners, including those who have not previously lived in the household, may nonetheless transmit risk to parents and children. However, it is not clear how such experiences compare to PI, or whether their consequences are equally severe.

If imprisonment not only transmits risks to prisoners’ own children, but also to prisoners’ siblings, grandchildren, or other household members, the group of children and adults potentially affected is much larger than has previously been assumed. Furthermore, as prison populations around the world have risen rapidly over the past 15 years (Coyle, Fair, Jacobson, & Walmsley, 2016), understanding the extent to which prisons directly transmit long-term harms to family members is vital. If family imprisonment is indeed a “risk mechanism” rather than just a “risk marker” (Murray & Farrington, 2005) of disadvantageous life outcomes, this has profound importance for correctional policy and institutional reform.

In this article, we begin to address these fundamental research gaps. Using both random and fixed effects models, we examine whether any of the three forms of family imprisonment is a “risk marker” for (i.e., associated with) a lack of social support or poor mental health (Research Question 1). We also test whether any of the three forms of family imprisonment are a potential “risk mechanism,” (i.e., whether an association remains after controlling for the time-stable, individual characteristics of people affected) for a lack of social support and poor mental health (Research Question 2). For each of these effects, we look at whether they apply equally to men, women, and young people (aged 15-20 years).

Background

The Australian Context

In December 2016, the Australian imprisonment rate reached a new peak of 208 prisoners per 100,000 adults (Australian Bureau of Statistics [ABS], 2016), compared with a global average of only 144 per 100,000 (Coyle et al., 2016). The U.S. imprisonment rate continued to be more than twice as high, at 458 prisoners per 100,000 adults in 2015 (Carson, 2016). As data regarding the parental status of prisoners are not routinely collected in Australia (Dennison, Stewart, & Freiberg, 2013), there is little information regarding the increasing numbers of children likely to be affected. Extrapolating from a 2001 New South Wales estimate that 4.3% of children experienced PI in their lifetime (Quilty, Levy, Howard, Barratt, & Butler, 2004), PI could affect at least 230,000 children currently living in Australia¹. And, though there are no local or national estimates for children affected by the imprisonment of family or household members other than parents, recent research suggests that this number could be considerably higher still (K. L. Besemer & Dennison, in press).

Although steep rises in the American imprisonment rate have triggered research targeted at the consequences of incarceration for family life (Wildeman, Wakefield, & Turney, 2013), the Australian body of family imprisonment research remains small (Dennison et al., 2013). So far, Australian findings have corresponded with international research in finding strong associations between parental and family imprisonment and social exclusion (K. L. Besemer & Dennison, in press), though the extent to which this represents a causal relationship has not yet been tested.

Qualitative research shows that in Australia, as in other countries, imprisonment is often a distressing experience for prisoners’ children and their other family members (Dennison & Besemer, in press), suggesting possible adverse mental health consequences. However, empirical findings on the effects of PI on children have been mixed. Using a population-based record linkage study, two recent Australian studies found associations between parental offending and adverse developmental outcomes in early childhood, including aggression, social competence, language and cognitive skills, and physical health and well-being at age 5 years (Laurens et al., 2017; Tzoumakis et al., 2017). On the contrary, an Australian longitudinal study found that, after controlling for preexisting risk, having a mother whose current partner had ever been incarcerated was not significantly associated with problem behaviors in 14-year olds (Kinner, Alati, Najman, & Williams, 2007).

International Findings

As current research on the effects of imprisonment on family relationships has almost exclusively focused on prisoners’ children, little is known about the way imprisonment may influence other close family members. A recent theoretical model by Foster and Hagan (2015b) suggests a range of risk mechanisms through which PI may affect on children, both through effects on their parents as well as through individual-level effects. It is likely that many of these effects also operate when children are affected by imprisonment through other close family connections. Here, we focus on two of these proposed effects of PI: low social support and poor mental health.

Social Support

Social support, defined as the provision of emotional and instrumental assistance through informal social networks, is thought to be one of the fundamental protective influences mitigating the harmful consequences of stress in adverse situations (Arditti, 2005; Capowich et al., 2001; Cohen & Wills, 1985; Cullen, 1994; Lösel et al., 2012; Murray et al., 2012). As the imprisonment of a family member is likely to be a source of stress in the lives of most people affected, social support is not only a key outcome in itself, but is also likely to be a key determinant of other psychosocial outcomes.

Although social support has rarely been measured directly in the context of family imprisonment, it has sometimes been assumed that disadvantages that characterize the lives of prisoners’ relatives reflect deficits in the extent and quality of children and adults’ social support networks (e.g., Murray, 2007). However, the relationship between social disadvantage and social support is complex. There has been some evidence that certain poor and disadvantaged populations may have equal or even greater access to social support than other households, both through strong neighborhood-based networks in some deprived areas, and through stronger family connections (Bailey, Besemer, Bramley, & Livingston, 2015; Matthews & Besemer, 2015; Pinkster, 2007). Consequently, it is not clear whether family members of prisoners, while disadvantaged in other ways, are more likely to have prior problems with social support.

Qualitative studies suggest that imprisonment may affect upon families’ social relationships as a result of family members’ own shame and anxiety as well as because of hostile or disapproving social responses from others (Braman, 2004; Condry, 2013; Phillips & Gates, 2011). These findings are confirmed in research using data from the Fragile Families and Child Wellbeing Study (FFCWS), a longitudinal survey of parents and children in 20 large U.S. cities. Turney, Schnittker, and Wildeman (2012) show that mothers who had experienced recent paternal incarceration were less likely to feel they had friends or family members who might provide small or large amount of financial support to them, even after controlling for their previous social support. Imprisonment is also thought to cause long-term impairments to the social and human capital of prisoners’ families and communities (Hagan & Dinovitzer, 1999; Rose & Clear, 1998, 2003).

In addition, some of the effects of imprisonment found in other studies are likely to affect indirectly on both the emotional and instrumental support of family members. For example, if the prisoner provided a substantial part of his or her household’s income, loss of this income could result in children and adult’s forced disengagement from social activities and events they can no longer afford to participate in, resulting in affected social contacts (K. L. Besemer & Dennison, 2018; Hagan & Dinovitzer, 1999; Murray & Farrington, 2008; Schwartz-Soicher, Geller, & Garfinkel, 2011). Children of prisoners are more likely to experience multiple house moves, as well as changes of school and caregiver (Dallaire, 2007; Jorgensen, Hernandez, & Warren, 1986; Murray, 2005, 2007; Rose & Clear, 2003), which may disrupt existing friendships and impede the subsequent development of strong peer relationships (K. L. Besemer & Dennison, 2018; Bocknek, Sanderson, & Britner, 2009; Oishi & Schimmack, 2010). Although the combination of these circumstances suggests that parental incarceration may affect on both children and adults’ social networks, social support has rarely been investigated in this context (Luther, 2015).

Mental Health

Although many qualitative studies suggest that imprisonment diminishes the psychological well-being of prisoners’ family members and spouses (e.g., Braman, 2004; Comfort, 2008, 2016; Condry, 2013; Dennison & Besemer, in press; Meek, 2008), few studies have tested this effect empirically. A number of cross-sectional studies have shown that family imprisonment is associated with substantially worse physical and mental health (K. L. Besemer & Dennison, in press; Gaston, 2016; Wildeman, Lee, & Comfort, 2013). However, it is not clear whether these problems preceded the imprisonment, or whether they primarily reflect the overall disadvantageous circumstances affecting this population. In a recent U.S. study, adult children of imprisoned parents were found to be predisposed toward greater mental health problems; however, the authors were not able to correct for preexisting risks (Gaston, 2016). The only direct effect of imprisonment on mental health was found in a study using the U.S. “Fragile Families” data, with paternal incarceration associated with elevated levels of life dissatisfaction in mothers, as well as a heavily increased risk of a major depressive episode (Wildeman, Schnittker, & Turney, 2012). There are few other studies regarding the mental health consequences of any other adults confronted with the imprisonment of a close relative. However, a systematic review of international studies testing the association between paternal imprisonment and various measures of mental health of children found only a very weak, inconsistent association between paternal imprisonment and children’s mental health (Murray et al., 2012).

In relation to our two research questions, we formulated the following hypotheses:

Hypothesis 1A (H1A): People who experience PI, household member imprisonment (HI), or close family member imprisonment (CFI) have a greater overall likelihood of low social support.

Hypothesis 1B (H1B): People who experience PI, HI, or CFI have a greater overall likelihood of low mental health.

Hypothesis 2A (H2A): People have a greater likelihood of low social support directly after experiencing PI, HI, or CFI.

Hypothesis 2B (H2B): People have a greater likelihood of low mental health directly after experiencing PI, HI, or CFI.

Method

Sample

Our sample comes from unit record data in the Housing, Income, and Labour Dynamics in Australia (HILDA) longitudinal survey; an annual nationally representative panel study containing up to 15 years of detailed socioeconomic information about approximately 28,000 individuals at the time of our analysis. In HILDA, data are collected about each household member, although interviews are only conducted with people aged 15 years and above. In addition to the main interview, which is normally conducted in person, interviewees were also asked to complete a self-report questionnaire that included more sensitive questions. This questionnaire covers a range of items including mental and physical health status (the Short Form-36 health survey, SF-36), social support, and life events (Summerfield et al., 2016; Watson & Wooden, 2012). The main dependent variables examined in this study were available in all annual waves of HILDA (2001-2015), although information about life events, including family imprisonment, was only collected from Wave 2 onward.

Dependent Variables

We examine two key outcome measures: social support and mental health. Social support is measured through the respondent’s own perceived adequacy of their social support. It is a 10-item measure covering three key domains of social support—emotional support, practical support, and loneliness (Milner, Krnjacki, Butterworth, & LaMontagne, 2016). As in Honey, Emerson, and Llewellyn (2011), the social support scale was compiled by taking the average of these items. Items included: “people don’t come to visit me as often as I would like”; “I often need help from other people but can’t get it”; and “I seem to have a lot of friends.” Each item is rated on a Likert-type scale ranging from 1 strongly disagree to 7 strongly agree. Negatively worded items were reverse-scored. The distribution of social support was heavily skewed. Only 9.00% of all scores were below 4, whereas 69.25% of all measured social support scores ranged from 5 to 7. Natural log transformations did not reduce skewness. We, therefore, converted the social support scale into a binary measure, where weak social support is identified as a score within the lowest quartile of scores of the HILDA population as a whole.

Mental health was measured using the five-item Mental Health Inventory (MHI-5), a subscale from the SF-36 general health measure. This instrument is composed of five items: How much of the time during the last month have you: “been a very nervous person”; “felt calm and peaceful?”; “felt downhearted and blue?”; “been a happy person?” and “felt so down in the dumps that nothing could cheer you up?” The MHI-5 is widely used as a validated screening instrument for depressive symptoms, mood disorders, and other mental health problems in the general population (Beusterien, Steinwald, & Ware, 1996; Rumpf, Meyer, Hapke, & John, 2001; Silveira et al., 2005). In our analysis, we use a continuous transformation of the MHI-5 score, recoded on a 0 to 100 range with a higher score meaning better mental health. The distribution of mental health was heavily skewed, 50.00% of all mental health scores in the data set were between 80 and 100. Natural log transformations did not reduce skewness. We, therefore, dichotomized this variable by defining poor mental health as a score within the lowest quartile of scores of the HILDA population as a whole. Although dichotomizing outcome variables necessarily results in a loss of power, categorization of our dependent variables was unavoidable. Left continuous, the nonnormal distributions of both outcome variables would have otherwise violated the assumptions of parametric statistical tests such as random and fixed effects models (Altman & Royston, 2006; Streiner, 2002).

Independent Variables

We looked at the influence of three key types of events on our outcome variables: (a) the imprisonment of a “close family member” (CFI), (b) the imprisonment of a person who is currently living in the respondent’s household (HI), and (c) the imprisonment of a young person’s mother or father (PI). The experience of CFI was measured in a self-completion questionnaire, in which respondents were asked to indicate whether they had experienced one of a series of events in the past year. One of these life events was whether they had a “close family member detained in a jail/correctional facility.” We also use information from the life events section to create indicator variables denoting HI. Individuals were included as having experienced HI if a current cohabitant had indicated that they had been “detained in a jail/correctional facility” in the past year. HI and CFI are not as closely overlapping as one might assume. Of people who have a current household member who went to prison, only 24.7% indicated that a close family member went to prison. And, for people who indicated that a close family member went to prison, only 5.0% were living with a household member who went to prison in the past year.² PI was derived by identifying young people for whom any biological or social parent had indicated that they had been detained. This variable could only be included for young people up to the age of 20 years, as parental information was largely unavailable for older adults. As many young people lived with the imprisoned parent after release there was considerable overlap between the group of people experiencing PI and HI. Of the 264 young people who had a parent go to prison, 177 also experienced the imprisonment of a household member (67.0%).

Other covariates included as controls in our model were age, marital status, number of children in the household, household income, general health, and remoteness area. Age was defined as the respondent’s age on June 30th, in the year the survey wave commenced. Marital status was recorded at the time of the interview. For young people, “separated” and “widowed” were recorded as “single” because there were almost no cases in these categories. The number of children in the household denoted people aged 0 and 17 years living in the household. If the respondent was aged 15 years to 17 years, this number also included the respondent themselves. Household income was calculated as the Organisation for Economic Co-operation and Development (OECD)-equivalized income and converted as a percentage of median income in the survey year to correct for inflation and household size. Remoteness area denoted the respondent’s residential location, classified within the Australian Standard Geographical Classification System (i.e., major cities, inner regional, outer regional, remote, very remote). Remoteness area was scored 0 to 5, with higher scores indicating greater rurality.

Analytic Approach

We examine the effects of CFI, HI, and PI on children and adults’ social support and mental health using random and fixed effects analysis in STATA 14.2. Both random and fixed effect models take into consideration the clustering of observations within persons and have the capacity to handle unbalanced panel designs (D. R. Johnson, 1995). However, in the random effects model, we simultaneously compare individuals with each other, as well as examine differences within individuals over time. As these models do not remove potential bias caused by unobserved differences between people, we can only use them to identify whether family imprisonment acts as a risk marker. In the fixed effects models, we compare scores associated with waves in which family imprisonment took place with the same individual’s mean score during all other observations. As these models use only within-person comparisons, we are able to remove bias associated with unmeasured time-constant differences between people in the population. In addition, we control for relevant time-varying covariates by adding these into the analysis as control variables. Consequently, while the random effects models show whether family, household, or PI are associated with different mental health and social support, fixed effects models are less sensitive to differences caused by preexisting characteristics of people affected. Comparison of the random and fixed effects models thus provides an indication of whether family imprisonment is only a risk marker, or whether it could be a risk mechanism for these outcomes.

Results

Sample Description

The analytical sample comprised 28,355 people with an average of 6.8 observations overall, of whom 4.8% had at least one observation just after the imprisonment of a close family member. This overall sample was subdivided into key demographic groups: men (n = 13,938), women (n = 14,417), and young people aged 15 years to 20 years (n = 3,769). Due to missing mental health and social support information for some individuals, samples for the random effects analyses were slightly smaller (N = 27,905 overall). In fixed effects models, cases were excluded if there was no variation in the dependent variable across observations, resulting in samples of 11,547 for social support and 11,924 for mental health. More information about the overall sample can be found in Table 1.

Table 1:

Characteristics of the Analytic Sample at First Observed Wave

Sample characteristic	Whole sample (N = 28,355)	Men (n = 13,938)	Women (n = 14,417)	People aged 15-20 years (n = 3,769)
Will experience CFI (% of sample)	4.8	3.9	5.7	5.8
Will experience HI (% of sample)	1.7	1.4	2.0	2.3
Will experience PI (% of sample)	—	—	—	0.6
Age (M) in years	29.9	29.4	30.5	17.7
Marital status (% of sample)
In a relationship	63.4	64.5	62.4	21.9
Separated	7.6	6.1	8.8	—
Widowed	4.0	1.5	6.4	—
Single	25.0	27.9	22.4	78.1
Number of children in household (M)	2.3	2.3	2.3	2.4
Employment status (% of sample)
Employed (PT/FT)	63.2	70.0	57.0	54.7
Unemployed	5.6	6.6	4.7	15.2
Not in labor force	31.2	23.4	38.3	30.0
General health (M)	69.6	69.9	69.3	71.0
Mental health (M)	73.2	74.5	72.0	70.9
Remoteness area (0-4; M)	0.5	0.5	0.5	0.5

Note. Sample characteristics are provided for respondents’ first observed wave only, as responses on these variables may change across survey waves. CFI = close family member imprisonment; HI = household member imprisonment; PI = parental imprisonment; PT = part time; FT = full time.

Social Support

Our first hypothesis (Hypothesis 1A) was that people who experience PI, HI, or CFI have a greater overall likelihood of low social support. Hypothesis 2A was that people have a greater likelihood of low social support directly after experiencing PI, HI, or CFI. Results for both Hypotheses 1A and 2A can be seen in Tables 2 and 3. Random effects models show that without control variables, CFI and HI are indeed associated with a significantly higher likelihood of having inadequate social support. The odds of experiencing low social support are more than 25% higher for people affected by CFI and 37% higher for those who experience HI. After controlling for standard covariates, there was still an association between family, or household, imprisonment, and social support.

Table 2:

Inadequate Social Support and Family Imprisonment

	Random effects model				Fixed effects model
	No controls		With controls		No controls		With controls
Variable	OR	SE	OR	SE	OR	SE	OR	SE
Close family imprisonment	1.255***	0.081	1.169*	0.076	1.027	0.069	1.009	0.069
Age			1.002*	0.001			.989***	0.002
Marital status (reference group = married)
Separated			1.800***	0.073			1.499***	0.075
Widowed			1.221**	0.073			1.358***	0.108
Single			1.327***	0.043			1.361***	0.061
Number of children in household			1.009***	0.002			1.005*	0.003
Employment status (reference = employed)
Unemployed			1.328***	0.054			1.070	0.047
Not in labor force			0.989	0.024			.925 **	0.026
Household income			0.998***	0.000			0.999***	0.000
General health			0.965***	0.001			0.974***	0.001
Remoteness area			1.027	0.018			0.970	0.026
Observations	190,659		190,659		106,442		106,442
Individuals	27,905		27,905		11,547		11,547

Note. OR = odds ratio.

p < .05. **p < .01. ***p < .001.

Table 3:

Inadequate Social Support and Household Imprisonment

	Random effects model				Fixed effects model
	No controls		With controls		No controls		With controls
Variable	OR	SE	OR	SE	OR	SE	OR	SE
Household imprisonment	1.375*	0.185	1.342*	0.180	1.020	0.144	1.049	0.149
Age			1.002*	0.001			0.990***	0.002
Marital status (reference group = married)
Separated			1.800***	0.073			1.499***	0.075
Widowed			1.223**	0.073			1.359***	0.108
Single			1.328***	0.043			1.362***	0.061
Number of children in household			1.009***	0.002			1.005*	0.003
Employment status (reference = employed)
Unemployed			1.329***	0.054			1.070	0.047
Not in labor force			0.989	0.024			0.925**	0.026
Household income			0.998***	0.000			0.999***	0.000
General health			0.965***	0.001			0.974***	0.001
Remoteness area			1.027	0.018			0.970	0.026
Observations	190,659		190,659		106,442		106,442
Individuals	27,905		27,905		11,547		11,547

Note. OR = odds ratio.

p < .05. **p < .01. ***p < .001.

In Table 4, we present the results for PI fixed and random effects models for 15- to 20-year olds, along with the same age groups for CFI and HI. All odds ratios (ORs) in Table 4 come from separate regression models. Where models include controls, these are the same as in Tables 2 and 4. As can be seen in Table 4, there was no significant association between PI and social support in any of the models. These results were also consistent for males and females, once controls were included. Consequently, we can reject Hypothesis 1A for young people. Fixed effects models show that people of any age and gender do not, on average, experience lower social support in years they experience CFI or HI. Therefore, with regard to social support, we also reject our hypothesis that people have a greater likelihood of low social support directly after experiencing PI, HI, or CFI (Hypothesis 2A). On the whole, these results do not provide any evidence that people experience any difference in social support in the specific year that their relatives are imprisoned as compared with other years.

Table 4:

Inadequate Social Support by Type of Family Imprisonment, Gender and Age Group

	Random effects model				Fixed effects model
	No controls		With controls		No controls		With controls
Variable	OR	SE	OR	SE	OR	SE	OR	SE
All aged 15-20 years
Close family member incarceration	1.553*	0.274	1.283	0.224	0.864	0.185	0.852	0.189
Household member incarceration	1.268	0.360	1.530	0.476	0.801	0.298	0.955	0.361
Parental incarceration	2.073	1.105	1.727	0.962	1.092	0.589	1.297	0.704
Women
Close family member incarceration	1.270**	0.106	1.213*	0.101	0.998	0.087	0.997	0.088
Household member incarceration	1.205	0.215	1.189	0.212	0.898	0.164	0.937	0.173
Men
Close family member incarceration	1.261*	0.129	1.142	0.117	1.075	0.114	1.029	0.111
Household member incarceration	1.658*	0.341	1.583*	0.323	1.233	0.273	1.215	0.271

Note. Controls are the same as in Table 3. Coefficients of CFI, HI, and PI are from separate regression models. OR = odds ratio; CFI = close family member imprisonment; HI = household member imprisonment; PI = parental imprisonment.

p < .05. **p< .01. ***p < .001.

Mental Health

We hypothesized that people who experience PI, HI, or CFI have a greater overall likelihood of low mental health (Hypothesis 1B) and that people have a greater likelihood of low mental health directly after experiencing PI, HI, or CFI (Hypothesis 2B). In random effects models with and without control variables, CFI is strongly associated with having mental health scores in the lowest quartile (Table 5). Without controls, people affected by CFI were 54% more likely to have poor mental health. The fixed effects models in Table 5 also show a strong direct relationship between family imprisonment and mental health. In years in which people report the imprisonment of a close family member, people have a significantly higher likelihood of poor mental health.

Table 5:

Low Mental Health and Close Family Imprisonment

	Random effects model				Fixed effects model
	No controls		With controls		No controls		With controls
Variable	OR	SE	OR	SE	OR	SE	OR	SE
Close family imprisonment	1.540***	0.098	1.385***	0.090	1.196**	0.079	1.178*	0.083
Age			0.973***	0.001			0.979***	0.002
Marital status (reference group = married)
Separated			1.792***	0.071			1.543***	0.079
Widowed			1.557***	0.093			1.807***	0.157
Single			1.143***	0.034			1.150**	0.047
Number of children in household			1.000	0.002			1.001	0.003
Employment status (reference = employed)
Unemployed			1.492***	0.060			1.242***	0.054
Not in labor force			1.138***	0.027			1.017	0.028
Household income			0.999***	0.000			1.000***	0.000
General health			0.936***	0.001			0.947***	0.001
Remoteness area			0.940***	0.015			0.959	0.026
Observations	190,659		190,659		106,442		106,442
Individuals	27,905		27,905		11,547		11,547

Note. OR = odds ratio.

p < .05. **p < .01. ***p < .001.

HI is also associated with poor mental health in the random effects model (Table 6). However, in the fixed effects model, we found no significant direct effect of HI on mental health. We can, therefore, conclude that the association between HI and poor mental health that was found in the fixed effect model is a consequence of unmeasured time-invariant differences between sample members. This finding is particularly striking when comparing the effect of HI with the effect of some of the controls. Becoming widowed is associated with an 81% higher chance of poor mental health, whereas being separated increases the risk of low mental health by 54%. Even unemployment, a relatively common life event, has a much stronger average effect on mental health than HI.

Table 6:

Low Mental Health and Household Imprisonment

	Random effects model				Fixed effects model
	No controls		With controls		No controls		With controls
Variable	OR	SE	OR	SE	OR	SE	OR	SE
Household imprisonment	1.644***	0.215	1.434**	0.191	1.204	0.163	1.197	0.170
Age			0.973***	0.001			0.979***	0.002
Marital status (reference group = married)
Separated			1.794***	0.071			1.543***	0.079
Widowed			1.560***	0.093			1.810***	0.157
Single			1.144***	0.034			1.152***	0.047
Number of children in household			1.000	0.002			1.001	0.003
Employment status (reference = employed)
Unemployed			1.495***	0.060			1.243***	0.054
Not in labor force			1.138***	0.027			1.018	0.028
Household income			0.999***	0.000			1.000***	0.000
General health			0.936***	0.001			0.947***	0.001
Remoteness area			0.940***	0.015			0.959	0.026
Observations	190,659		190,659		106,442		106,442
Individuals	27,905		27,905		11,924		11,924

Note. OR = odds ratio.

p < .05. **p < .01. ***p < .001.

The direct effect of close family imprisonment on mental health was not significant across all demographics (Table 7). We found no significant effect of CFI on either the mental health of women or young people. Furthermore, there was no association between young people’s experiences of PI, HI, or CFI and their mental health outcomes in that year. Indeed, the direct effect of CFI on mental health was only significant for men. Therefore, we found partial support for Hypothesis 1B and no support for Hypothesis 2B with the exception of men who experienced CFI.

Table 7:

Low Mental Health by Type of Family Imprisonment, Gender and Age Group

	Random effects model				Fixed effects model
	No controls		With controls		No controls		With controls
Variable	OR	SE	OR	SE	OR	SE	OR	SE
Young people aged 15-20 years
Close family member incarceration	2.216***	0.346	1.836***	0.285	1.290	0.240	1.350	0.269
Household member incarceration	1.269	0.360	0.948	0.271	0.780	0.243	0.728	0.245
Parental incarceration	1.214	0.623	0.877	0.471	0.869	0.462	1.060	0.632
Women
Close family member incarceration	1.435***	0.114	1.327**	0.108	1.128	0.093	1.121	0.098
Household member incarceration	1.574**	.265	1.363	0.235	1.118	0.194	1.088	0.200
Men
Close family member incarceration	1.713***	0.180	1.457***	0.157	1.326*	0.147	1.280*	0.151
Household member incarceration	1.755**	0.365	1.534*	0.321	1.350	0.291	1.378	0.307

Note. Controls are the same as in Table 6. Coefficients of CFI, HI, and PI are from separate regression models. OR = odds ratio; CFI = close family member imprisonment; HI = household member imprisonment; PI = parental imprisonment.

p < .05. **p < .01. ***p < .001.

In our previous analyses, we found that there was an effect of close family imprisonment on men, but not on women or young people. For our final analysis, we used an interaction term in our fixed effects models to identify whether social support changes the relationship between men’s experience of close family imprisonment and their subsequent mental health. In a model without controls, social support was found to have a strong direct effect on mental health (OR = 2.588, SE = 0.377, p < .001). However, after controlling for this direct effect, we found no significant interaction between family imprisonment and social support (OR = 1.365, SE = 0.236, p = .072). In a model with controls, we also found that social support did not moderate the relationship between close family imprisonment and men’s mental health (OR = 1.270, SE = 0.230, p = .189). We conclude that low social support increases the risk of mental health problems in general, but does not moderate the relationship between close family imprisonment and men’s mental health.

Discussion

The aim of this study was to examine the association between the various types of family imprisonment (CFI, HI, and PI) and both social support and mental health. After introducing controls, we found no effect of any kind on social support, not even a baseline correlation. And, although we found an overall association between CFI and mental health, the only significant direct effect on mental health was for adult men. We found no effects of either PI or HI on mental health. We will discuss each of these findings in turn.

Our findings do not support theoretical assumptions (e.g., Arditti, Lambert-Shute, & Joest, 2003) that view family imprisonment as a determinant of impaired social support of either adults or young people. We did, however, find that family members of prisoners had slightly lower social support than other people with comparable demographic characteristics. This is not to say that the imprisonment of parents, household members, or close family members has no social consequences. It is possible that adults or young people experience negative or stigmatizing responses from more distant friends and acquaintances, but are nonetheless able to draw on the same levels of social support from their closer friend and family relationships. It may also be that although people perceived themselves to have the same access to social support as on previous survey years, this support varied in terms of its nature or effectiveness. These nuances should be investigated in future work. However, our study found no evidence that on average parental, household or close family imprisonment is a “mechanism” of weak social support.

Although there was no average direct effect of CFI, HI, or PI on social support, such average effects may obscure positive or negative effects on certain subgroups within this population. The strong emotional responses to family imprisonment observed in qualitative research with prisoners’ relatives in Australia as well as other countries (e.g., Braman, 2004; Comfort, 2008, 2016; Condry, 2013; Dennison & Besemer, in press; Meek, 2008) may be indicative of such subgroup effects. It is possible that such studies capture families affected by longer and more serious prison sentences, and therefore, finds much greater negative psychological consequences in family members affected. Further research should explore such potential diversity in outcomes, and identify the characteristics of people who may be at greater risk in the event of family imprisonment.

For mental health, we found an overall negative baseline association. That is, overall, people who experience the imprisonment of family members have poorer mental health. Although not unexpected, this is an important reminder of the overall vulnerability of this population. The direct effect of imprisonment on mental health is interesting because it was found to only be present in men. As the majority of research on adult family member effects of imprisonment has focused on female caregivers (e.g., Green et al., 2006; Turney, 2014; Wildeman, Lee, & Comfort, 2013; Wildeman et al., 2012), very little is known about the way men experience such events. Without further research, any further hypotheses regarding the reasons for this gender division can only be speculative. However, although we have no specific information about respondents’ relationship to the “close family member” who is imprisoned, we do have this information for any current household members. Of all women over 14 years of age who are affected by the imprisonment of a current household member, most are confronted by the imprisonment of a spouse (36.2%). Men are far less likely to be affected by spousal imprisonment (15.75% of all HI). Women are slightly more likely than men to be confronted with the imprisonment of their child or step child (25.84% of HI experienced by women, 22.71% of HI for men). It is possible that women, more than men, may have previously been exposed to high levels of psychological distress because of their family members’ offending. Some may have also been a victim of the family members’ abuse. Women may, therefore, be more likely than men to experience family imprisonment positively (Johnston, 2006). Alternatively, women may describe a wider range of family relationships as “close family,” and thus be less emotionally affected by these imprisonments than men. As can be seen in Table 1, a far greater proportion of women experienced household or close family imprisonment overall.

A further possibility is that there may be gendered differences in the way social support is used effectively to mitigate the negative effects of CFI and HI on adult men and women. The social support measure used in this study reflects perceived access to social support, but not the extent to which social support is actually used. There may be gendered differences in the use of social support that are not adequately captured by access to support as a control variable. Further research should investigate the social contexts and mechanisms that might explain the gendered nature of this association.

Limitations

Although this study uses a very large and representative longitudinal data set of young people and adults affected by imprisonment, a number of caveats are warranted. First, even in a large national population sample, we only observe a relatively small number of parental, household, and close family imprisonments. Consequently, the small sample size may have reduced the power of these statistical tests, so that smaller effects on social support or mental health may not have been identified. In addition, though the HILDA data set as a whole is a representative sample of the Australian population, these households may not constitute a representative subset of all families affected by imprisonment.

Second, while fixed effects models are useful for controlling for all time-invariant unobserved differences between families who experience imprisonment and those who do not, they can only control for time-varying covariates if those are included in the model. As in all statistical models, this leaves open the possibility that important control variables may have been omitted. For example, it was not possible to distinguish between the effects of maternal and paternal imprisonment on young people. These two experiences may have different effects on the lives of young people, which were not captured when the two groups were combined. We were also unable to control for disruptions and changes in people’s lives that could have happened at the same time as the incarceration of a family member. Such events may have included traumatic experiences, which could have influenced the results. For young people, disruptions associated with imprisonment may include changes in primary caregivers. Other important unmeasured variations may have resulted from the extent to which people felt subjectively close to the family member in prison, or what kind of relationship people had to the family member imprisoned (other than for parents.) Similarly, our data do not contain information about the duration of parental, household, or close family imprisonment, or how often this experience had already happened in that person’s life prior to the first measurement. Due to small sample sizes, we were also unable to consider differential effects by Indigenous status and ethnicity. All these factors could influence the severity of impact. Aggregating across these diverse experiences may, therefore, hide considerable variation in results. In addition, fixed effects models can only measure change in participants if we have information about them over more than one time point. This meant that cases had to be excluded if there was no variation in the dependent variable across observations. For the fixed effects models of social support, 16,358 cases (and 84,217 observations) were dropped because of a lack of variation. For the fixed effects models on mental health, 15,981 cases (81,035 observations) were dropped. On average, cases without variation had slightly lower mental health and social support. Also, we only were able to include young people aged 15 years and above. It is possible that PI, HI, and CFI effects may be greater in the lives of young children.

Finally, our null findings may be a consequence of timing. In this research, mental health, and social support were captured at 1 year intervals, meaning that the first point of measurement after imprisonment varied anywhere between 0 and 12 months after the parental, family, or household imprisonment event. If psychological or social effects have an onset much longer after the event, they may not be captured in these models. Research on psychological trauma has shown that mood and anxiety disorders may develop many years after exposure to psychological stress (Heim, Newport, Mletzko, Miller, & Nemeroff, 2008; Lupien, McEwen, Gunnar, & Heim, 2009). To examine such long-term consequences, we would require a data set spanning a much longer time frame.

Finally, we would like to stress that these are just two of the many potential adverse outcomes that may occur as a consequence of family incarceration in Australia and in other countries. Further research should investigate whether the imprisonment of close family affects other adverse outcomes for adults and young people, such as academic attainment, physical health, and delinquent or criminal behavior.

Conclusion

We found that young people and adults affected by household and close family imprisonment were significantly more likely to suffer from mental health problems than people without such experiences. Direct mental health effects of family imprisonment were found for men only. Although social support generally reduced the risk of mental health problems in men, social support did not moderate the relationship between CFI and mental health problems. There was also no evidence that CFI or HI was directly related to mental health problems in women or young people. PI was not found to have any effect on young people’s mental health. Furthermore, family or household members of prisoners were not at greater risk of poor social support, nor did we find any direct connection between CFI, HI, or PI and social support. These results are different from those in American studies that found strong, direct negative effects of paternal imprisonment on maternal mental health (Wildeman et al., 2012) and maternal social support (Turney et al., 2012). As previously discussed in our article, the Australian welfare and corrections context differs considerably from that in the United States. These differences should serve as a reminder that results from the United States may not be directly generalizable across very different national contexts. Although currently uncommon in the family imprisonment literature (but see S. Besemer, van der Geest, Murray, Bijleveld, & Farrington, 2011; Murray, Bijleveld, Farrington, & Loeber, 2014), cross-national studies could play an important role in further exploring such differences, and testing the robustness of current empirical findings across different cultural and policy contexts.

It is also noteworthy that we found no direct effects of close family, household, or even PI on young people aged 15 years to 20 years. As previously stated in this article, it is possible that such effects manifest much later in children and young people’s lives (see Gaston, 2016), or only affect younger age groups. However, it is also possible that many of the serious problems identified in current literature are at least partially attributable to factors specific to the American prison context, and to the considerable preexisting disadvantage affected children suffer from.

Nonetheless, this study did find that adults and children affected by imprisonment have very high levels of mental illness, and that for men, these problems become worse after close family imprisonment. Even if, for women and young people, mental health is not directly influenced by imprisonment of household or family members, their overall vulnerability to mental health problems is in itself a strong argument for greater support.

Footnotes

Authors’ Note:

This article uses unit record data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey. The HILDA Project is funded by the Australian Government Department of Social Services (DSS) and managed by the Melbourne Institute of Applied Economic and Social Research. The findings and views reported in this article, however, are those of the authors and should not be attributed to either DSS or the Melbourne Institute. Professor Susan Dennison was supported by an Australian Research Council Future Fellowship (FT0991557). Views expressed in this article are those of the authors and do not necessarily represent those of the Australian Research Council. Where quoted or used, they should be clearly attributed to the authors.

Notes

ORCID iD

Kirsten L. Besemer

Kirsten L. Besemer is a research fellow at Griffith University in Brisbane, Australia. Her research examines through what mechanisms incarceration may affect on prisoners’ families and communities.

Steve G. A. van de Weijer is a postdoctoral researcher at the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), Amsterdam, the Netherlands. His research focuses on life-course criminology, intergenerational transmission of crime, and genetic influences on criminal behavior.

Susan M. Dennison is a professor and deputy head of school (research) in the School of Criminology and Criminal Justice at Griffith University as well as deputy director of the Griffith Criminology Institute. She is a former Australian Research Council Future Fellow, examining the impact of parental incarceration on the family system and the mechanisms that mediate developmental outcomes for incarcerated parents, their children, and caregivers.

References

Altman

D. G.

Royston

(2006). The cost of dichotomising continuous variables. British Medical Journal, 332, Article 1080. doi:10.1136/bmj.332.7549.1080

Apel

(2016). The effects of jail and prison confinement on cohabitation and marriage. The ANNALS of the American Academy of Political and Social Science, 665, 103-126. doi:10.1177/0002716216629360

Arditti

J. A.

(2005). Families and incarceration: An ecological approach. Families in Society: The Journal of Contemporary Social Services, 86, 251-260. doi:10.1606/1044-3894.2460

Arditti

J. A.

(2012). Parental incarceration and the family: Psychological and social effects of imprisonment on children, parents, and caregivers. New York: New York University Press.

Arditti

J. A.

Lambert-Shute

Joest

(2003). Saturday morning at the jail: Implications of incarceration for families and children. Family Relations, 52, 195-204. doi:10.1111/j.1741-3729.2003.00195.x

Australian Bureau of Statistics. (2016). Prisoners in Australia, 2016 (4517.0). Available from http://www.abs.gov.au

Auty

K. M.

Farrington

D. P.

Coid

J. W.

(2015). The intergenerational transmission of criminal offending: Exploring gender-specific mechanisms. British Journal of Criminology, 57, 215-237. doi:10.1093/bjc/azv115

Bailey

Besemer

Bramley

Livingston

(2015). How neighbourhood social mix shapes access to resources from social networks and from services. Housing Studies, 30, 295-314. doi:10.1080/02673037.2014.1000834

Besemer

K. L.

Dennison

S. M.

(2018). Social exclusion in families affected by paternal imprisonment. Australian & New Zealand Journal of Criminology, 51, 221-238. doi:10.1177/0004865817701530

10.

Besemer

K. L.

Dennison

S. M.

(in press). Intergenerational social exclusion in prisoners’ families. In Hutton

Moran

(Eds.), Handbook on prison and the family. Basingstoke, UK: Palgrave.

11.

Besemer

van der Geest

Murray

Bijleveld

C. C. J. H.

Farrington

D. P.

(2011). The relationship between parental imprisonment and offspring offending in England and the Netherlands. British Journal of Criminology, 51, 413-437. doi:10.1093/bjc/azq072

12.

Beusterien

K. M.

Steinwald

Ware

J. E.

Jr. (1996). Usefulness of the SF-36 health survey in measuring health outcomes in the depressed elderly. Journal of Geriatric Psychiatry and Neurology, 9, 13-21. doi:10.1177/089198879600900103

13.

Bocknek

E. L.

Sanderson

Britner

P. A.

(2009). Ambiguous loss and posttraumatic stress in school-age children of prisoners. Journal of Child and Family Studies, 18, 323-333. doi:10.1007/s10826-008-9233-y

14.

Braman

(2004). Doing time on the outside: Incarceration and family life in urban America. Ann Arbor: University of Michigan Press.

15.

Capowich

G. E.

Mazerolle

Piquero

(2001). General strain theory, situational anger, and social networks: An assessment of conditioning influences. Journal of Criminal Justice, 29, 445-461. doi:10.1016/S0047-2352(01)00101-5

16.

Carson

E. A.

(2016). Prisoners in 2015. Washington, DC: Bureau of Justice Statistics. Retrieved from https://www.bjs.gov/index.cfm?ty=pbdetail&iid=5869

17.

Cohen

Wills

T. A.

(1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310-357. doi:10.1037/0033-2909.98.2.310

18.

Comfort

(2008). Doing time together: Love and family in the shadow of the prison. Chicago, IL: The University of Chicago Press.

19.

Comfort

(2016). “A twenty-hour-a-day job”: The impact of frequent low-level criminal justice involvement on family life. The ANNALS of the American Academy of Political and Social Science, 665, 63-79. doi:10.1177/0002716215625038

20.

Condry

(2013). Families shamed: The consequences of crime for relatives of serious offenders. Oxford, UK: Routledge.

21.

Coyle

Fair

Jacobson

Walmsley

(2016). Imprisonment worldwide: The current situation and an alternative future. Bristol, UK: Policy Press.

22.

Cullen

F. T.

(1994). Social support as an organizing concept for criminology: Presidential address to the academy of criminal justice sciences. Justice Quarterly, 11, 527-559. doi:10.1080/07418829400092421

23.

Dallaire

D. H.

(2007). Children with incarcerated mothers: Developmental outcomes, special challenges and recommendations. Journal of Applied Developmental Psychology, 28, 15-24. doi:10.1016/j.appdev.2006.10.003

24.

Dennison

S. M.

Besemer

K. L.

(in press). Missing and missing out: Social exclusion in children with an incarcerated parent. In Condry

Scharff Smith

(Eds.), Prisons, punishment and the family: Towards a new sociology of punishment. Oxford, UK: Oxford University Press.

25.

Dennison

S. M.

Stewart

Freiberg

(2013). A prevalence study of children with imprisoned fathers: Annual and lifetime estimates. Australian Journal of Social Issues, 48, 339-362. doi:10.1002/j.1839-4655.2013.tb00286.x

26.

Farrington

D. P.

Barnes

G. C.

Lambert

(1996). The concentration of offending in families. Legal and Criminological Psychology, 1, 47-63. doi:10.1111/j.2044-8333.1996.tb00306.x

27.

Farrington

D. P.

Jolliffe

Loeber

Stouthamer-Loeber

Kalb

L. M.

(2001). The concentration of offenders in families, and family criminality in the prediction of boys’ delinquency. Journal of Adolescence, 24, 579-596. doi:10.1006/jado.2001.0424

28.

Foster

Hagan

(2007). Incarceration and intergenerational social exclusion. Social Problems, 54, 399-433. doi:10.1525/sp.2007.54.4.399

29.

Foster

Hagan

(2015a). Maternal and paternal imprisonment and children’s social exclusion in young adulthood. Journal of Criminal Law & Criminology, 105, 387-430.

30.

Foster

Hagan

(2015b). Punishment regimes and the multilevel effects of parental incarceration: Intergenerational, intersectional, and interinstitutional models of social inequality and systemic exclusion. Annual Review of Sociology, 41, 135-158. doi:10.1146/annurev-soc-073014-112437

31.

Gaston

(2016). The long-term effects of parental incarceration does parental incarceration in childhood or adolescence predict depressive symptoms in adulthood? Criminal Justice and Behavior, 43, 1056-1075. doi:10.1177/0093854816628905

32.

Green

K. M.

Ensminger

M. E.

Robertson

J. A.

Juon

H. S.

(2006). Impact of adult sons’ incarceration on African American mothers’ psychological distress. Journal of Marriage and Family, 68, 430-441. doi:10.1111/j.1741-3737.2006.00262.x

33.

Hagan

Dinovitzer

(1999). Collateral consequences of imprisonment for children, communities, and prisoners. Crime and Justice, 26, 121-162.

34.

Heim

Newport

D. J.

Mletzko

Miller

A. H.

Nemeroff

C. B.

(2008). The link between childhood trauma and depression: Insights from HPA axis studies in humans. Psychoneuroendocrinology, 33, 693-710. doi:10.1016/j.psyneuen.2008.03.008

35.

Honey

Emerson

Llewellyn

(2011). The mental health of young people with disabilities: Impact of social conditions. Social Psychiatry and Psychiatric Epidemiology, 46, 1-10. doi:10.1007/s00127-009-0161-y

36.

Johnson

D. R.

(1995). Alternative methods for the quantitative analysis of panel data in family research: Pooled time-series models. Journal of Marriage and the Family, 57, 1065-1077. doi:10.2307/353423

37.

Johnson

E. I.

Easterling

(2012). Understanding unique effects of parental incarceration on children: Challenges, progress, and recommendations. Journal of Marriage and Family, 74, 342-356. doi:10.1111/j.1741-3737.2012.00957.x

38.

Johnson

E. I.

Easterling

(2013). Misidentified or misunderstood? A reply to Wildeman, Wakefield, and Turney (2013). Journal of Marriage and Family, 75, 259-262. doi:10.1111/j.1741-3737.2012.01026.x

39.

Johnston

(2006). The wrong road: Efforts to understand the effects of parental crime and incarceration. Criminology & Public Policy, 5, 703-719. doi:10.1111/j.1745-9133.2006.00414.x

40.

Jorgensen

J. D.

Hernandez

S. H.

Warren

R. C.

(1986). Addressing the social needs of families of prisoners: A tool for inmate rehabilitation. Federal Probation, 50, 47-52.

41.

Kinner

S. A.

Alati

Najman

J. M.

Williams

G. M.

(2007). Do paternal arrest and imprisonment lead to child behaviour problems and substance use? A longitudinal analysis. Journal of Child Psychology and Psychiatry, 48, 1148-1156. doi:10.1111/j.1469-7610.2007.01785.x

42.

Laurens

K. R.

Tzoumakis

Kariuki

Green

M. J.

Hamde

Harris

. . . Dean

(2017). Pervasive influence of maternal and paternal criminal offending on early childhood development: A population data linkage study. Psychological Medicine, 47, 889-901. doi:10.1017/S0033291716003007

43.

Lee

Wildeman

Wang

E. A.

Matusko

Jackson

J. S.

(2014). A heavy burden: The cardiovascular health consequences of having a family member incarcerated. American Journal of Public Health, 104(3), 421-427. doi:10.2105/AJPH.2013.301504

44.

Lee

R. D.

Fang

Luo

(2017). Parental incarceration and social exclusion: Long-term implications for the health and well-being of vulnerable children in the United States. In Bishop

J. A.

Rodriguez

J. G.

(Eds.), Inequality after the 20th century: Papers from the sixth ECINEQ meeting (pp. 215-234). Bingley, UK: Emerald. doi:10.1108/S1049-258520160000024009

45.

Lösel

Pugh

Markson

Souza

Lanskey

(2012). Risk and protective factors in the resettlement of imprisoned fathers with their families. Milton, UK: Ormiston Children and Families Trust. Retrieved from http://www.crim.cam.ac.uk/research/fathers_in_prison/final_report.pdf

46.

Lupien

S. J.

McEwen

B. S.

Gunnar

M. R.

Heim

(2009). Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nature Reviews Neuroscience, 10, 434-445. doi:10.1038/nrn2639

47.

Luther

(2015). Examining social support among adult children of incarcerated parents. Family Relations, 64, 505-518. doi:10.1111/fare.12134

48.

Matthews

Besemer

(2015). Social networks, social capital and poverty: Panacea or placebo? Journal of Poverty and Social Justice, 23, 189-201. doi:10.1332/175982715X14448122286274

49.

Meek

(2008). Experiences of younger siblings of young men in prison. Children & Society, 22, 265-277. doi:10.1111/j.1099-0860.2007.00108.x

50.

Meek

Lowe

McPhillips

(2010). The impact of a custodial sentence on the siblings of young offenders: Matching service to needs. Prison Service Journal, 190, 26-31.

51.

Milner

Krnjacki

Butterworth

LaMontagne

A. D.

(2016). The role of social support in protecting mental health when employed and unemployed: A longitudinal fixed-effects analysis using 12 annual waves of the HILDA cohort. Social Science & Medicine, 153, 20-26. doi:10.1016/j.socscimed.2016.01.050

52.

Murray

(2005). The effects of imprisonment on families and children of prisoners. In Liebling

Maruna

(Eds.), The effects of imprisonment (pp. 442-492). Portland, OR: Willan Publishing.

53.

Murray

(2007). The cycle of punishment: Social exclusion of prisoners and their children. Criminology & Criminal Justice, 7, 55-81. doi:10.1177/1748895807072476

54.

Murray

Bijleveld

C. C. J. H.

Farrington

D. P.

Loeber

(2014). Effects of parental incarceration on children: Cross-national comparative studies. Washington, DC: American Psychological Association.

55.

Murray

Farrington

D. P.

(2005). Parental imprisonment: Effects on boys’ antisocial behaviour and delinquency through the life-course. Journal of Child Psychology and Psychiatry, 46, 1269-1278.

56.

Murray

Farrington

D. P.

(2008). Parental imprisonment: Long-lasting effects on boys’ internalizing problems through the life course. Development and Psychopathology, 20, 273-290. doi:10.1017/S0954579408000138

57.

Murray

Farrington

D. P.

Sekol

(2012). Children’s antisocial behavior, mental health, drug use, and educational performance after parental incarceration: A systematic review and meta-analysis. Psychological Bulletin, 138, 175-210. doi:10.1037/a0026407

58.

Oishi

Schimmack

(2010). Residential mobility, well-being, and mortality. Journal of Personality and Social Psychology, 98, 980-994. doi:10.1037/a0019389

59.

Phillips

S. D.

Gates

(2011). A conceptual framework for understanding the stigmatization of children of incarcerated parents. Journal of Child and Family Studies, 20, 286-294. doi:10.1007/s10826-010-9391-6

60.

Pinkster

F. M.

(2007). Localised social networks, socialisation and social mobility in a low-income neighbourhood in the Netherlands. Urban Studies, 44, 2587-2603. doi:10.1080/00420980701558384

61.

Quilty

Levy

M. H.

Howard

Barratt

Butler

(2004). Children of prisoners: A growing public health problem. Australian and New Zealand Journal of Public Health, 28, 339-343. doi:10.1111/j.1467-842X.2004.tb00441.x

62.

Rose

D. R.

Clear

T. R.

(1998). Incarceration, social capital, and crime: Implications for social disorganization theory. Criminology, 36, 441-479. doi:10.1111/j.1745-9125.1998.tb01255.x

63.

Rose

D. R.

Clear

T. R.

(2003). Incarceration, reentry, and social capital. In Travis

Waul

(Eds.), Prisoners once removed: The impact of incarceration and reentry on children, families, and communities (pp. 189-232). Washington, DC: The Urban Institute Press.

64.

Rumpf

H. J.

Meyer

Hapke

John

(2001). Screening for mental health: Validity of the MHI-5 using DSM-IV Axis I psychiatric disorders as gold standard. Psychiatry Research, 105, 243-253. doi:10.1016/S0165-1781(01)00329-8

65.

Schwartz-Soicher

Geller

Garfinkel

(2011). The effect of paternal incarceration on material hardship. Social Service Review, 85, 447-473. doi:10.1086/661925

66.

Silveira

Taft

Sundh

Waern

Palsson

Steen

(2005). Performance of the SF-36 health survey in screening for depressive and anxiety disorders in an elderly female Swedish population. Quality of Life Research, 14, 1263-1274. doi:10.1007/s11136-004-7753-5

67.

Slomkowski

Rende

Conger

K. J.

Simons

R. L.

Conger

R. D.

(2001). Sisters, brothers, and delinquency: Evaluating social influence during early and middle adolescence. Child Development, 72, 271-283. doi:10.1111/1467-8624.00278/full

68.

Streiner

D. L.

(2002). Breaking up is hard to do: The heartbreak of dichotomizing continuous data. Canadian Journal of Psychiatry, 47, 262-266. doi:10.1177/070674370204700307

69.

Summerfield

Freidin

Hahn

Macalalad

. . . Wooden

(2016). HILDA user manual—Release 15. Melbourne, Australia: Melbourne Institute of Applied Economic and Social Research, University of Melbourne.

70.

Turney

(2014). The consequences of paternal incarceration for maternal neglect and harsh parenting. Social Forces, 92, 1607-1636. doi:10.1093/sf/sot160

71.

Turney

(2015). Liminal men: Incarceration and relationship dissolution. Social Problems, 62, 499-528. doi:10.1093/socpro/spv015

72.

Turney

Schnittker

Wildeman

(2012). Those they leave behind: Paternal incarceration and maternal instrumental support. Journal of Marriage and Family, 74, 1149-1165. doi:10.1111/j.1741-3737.2012.00998.x

73.

Tzoumakis

Dean

Green

M. J.

Zheng

Kariuki

Harris

. . . Laurens

K. R.

(2017). The impact of parental offending on offspring aggression in early childhood: A population-based record linkage study. Social Psychiatry and Psychiatric Epidemiology, 52, 445-455. doi:10.1007/s00127-017-1347-3

74.

Velez

C. E.

Krause

E. D.

McKinnon

Brunwasser

S. M.

Freres

D. R.

Abenavoli

R. M.

Gillham

J. E.

(2016). Social support seeking and early adolescent depression and anxiety symptoms: The moderating role of rumination. The Journal of Early Adolescence, 36, 1118-1143. doi:10.1177/0272431615594460

75.

Wakefield

Lee

Wildeman

(2016). Tough on crime, tough on families? Criminal justice and family life in America. The ANNALS of the American Academy of Political and Social Science, 665, 8-21. doi:10.1177/0002716216637048

76.

Wakefield

Uggen

(2010). Incarceration and stratification. Annual Review of Sociology, 36, 387-406. doi:10.1146/annurev.soc.012809.102551

77.

Watson

Wooden

M. P.

(2012). The HILDA survey: A case study in the design and development of a successful household panel survey. Longitudinal and Life Course Studies, 3, 369-381.

78.

Weinfield

N. S.

Sroufe

L. A.

Egeland

(2000). Attachment from infancy to early adulthood in a high-risk sample: Continuity, discontinuity, and their correlates. Child Development, 71, 695-702.

79.

Wildeman

(2014a). Parental incarceration, child homelessness, and the invisible consequences of mass imprisonment. The ANNALS of the American Academy of Political and Social Science, 651, 74-96. doi:10.1177/0002716213502921

80.

Wildeman

(2014b). Parental incarceration and child wellbeing: An annotated bibliography. The Sills Family Foundation. Retrieved from https://nrccfi.camden.rutgers.edu/2014/10/05/an-annotated-bibliography-parental-incarceration-and-child-wellbeing-by-christopher-wildeman/

81.

Wildeman

(2016). Incarceration and population health in wealthy democracies. Criminology, 54, 360-382. doi:10.1111/1745-9125.12107

82.

Wildeman

Lee

Comfort

(2013). A new vulnerable population? The health of female partners of men recently released from prison. Women’s Health Issues, 23, 335-340. doi:10.1016/j.whi.2013.07.006

83.

Wildeman

Schnittker

Turney

(2012). Despair by association? The mental health of mothers with children by recently incarcerated fathers. American Sociological Review, 77, 216-243. doi:10.1177/0003122411436234

84.

Wildeman

Wakefield

(2014). The long arm of the law: The concentration of incarceration in families in the era of mass incarceration. Journal of Gender Race and Justice, 17, 367-389. doi:10.1177/0004865817701530

85.

Wildeman

Wakefield

Turney

(2013). Misidentifying the effects of parental incarceration? A comment on Johnson and Easterling (2012). Journal of Marriage and Family, 75, 252-258. doi:10.1111/j.1741-3737.2012.01018.x