Maternal Separations During the Reentry Years for 100 Infants Raised in a Prison Nursery

Abstract

Prison nurseries prevent maternal separations related to incarceration for the small subset of children whose pregnant mothers are incarcerated in states with such programs. For a cohort of 100 children accepted by corrections into one prison nursery, subsequent separation patterns are analyzed. The largest numbers are caused by corrections’ removal of infants from the nursery and infants reaching a one-year age limit. Criminal recidivism and substance abuse relapse threaten continued mothering during reentry. Focused and coordinated services are needed during prison stay and reentry years to sustain mothering for women and children accepted into prison nursery programs.

Prison nurseries provide a housing space inside a secure criminal justice corrections setting where incarcerated women can co-reside with their newborn infants for various periods of time during which the mother completes an imposed judicial sentence while being the primary caregiver for her infant. Such arrangements have existed throughout the world and to a lesser extent in the United States (US; Cook & Davies, 1999; Weintraub, 1987). They have varied widely in terms of underlying philosophy, political will to support them, and resources available to mother and child (Goshin & Byrne, 2009; Kauffman, 2001; Radosh, 1988; Robertson, 2008: Shepard & Zemans, 1950). In the US, prison nursery programs have existed at one time or another within local, state, and federal jurisdictions with an erratic history of sustainability (Byrne, 2010; Craig, 2009). Such programs in this country have long been assumed by advocates to both create and sustain secure infant attachments, to establish a foundation that protects future child development, and to prevent the adult participants’ criminal recidivism. Opponents have claimed they interfere with prison security, interrupt child development and impose eventual separations. These assumptions have only recently been tested in the first longitudinal study of maternal and child outcomes for a cohort of prison nursery co-residents in a New York State prison system (Byrne, n.d.). This paper extends previously reported attachment findings from this study (Borelli, Goshin, Joestl, Clark, & Byrne, 2010; Byrne, Goshin & Joestl, 2010) by exploring the patterns of union and separation experienced by a cohort of former prison co-residing mothers and infants who gave informed consent to participate in longitudinal studies that began in prison and have continued from one through eight reentry years.

MATERNAL INCARCERATION AND EFFECTS OF SEPARATION ON CHILDREN

Forced separation and children of incarcerated parents

The bulk of literature on children of incarcerated women has been limited to anecdotal evidence of the harm caused by forced separation, and disseminated with the intent to support advocacy for recognition of these children’s needs. While advocacy work is critical for this vulnerable group of dependents, many aspects of child development in this population remain unexamined or poorly understood in the absence of data-based studies. There is also a gap in what is known about caregiving trajectories in these families both before and after incarceration. The limited body of research suffers from methodological limitations, including small convenience samples, no comparison groups, and no assessment of ecological risks such as substance use, known to be associated with both incarceration and child development. (Murray, Farrington, Sekol, & Olsen, 2009). Families have rarely been followed during the years in which former offenders reenter the community, creating the untested assumptions that post-release life will occur as planned, intervention effects seen during incarceration will continue, and problems experienced by children during the period of incarceration will subside.

Published research has focused on samples of school-aged children and adolescents or included children at widely varying ages in a single study. Maternal incarceration during school-age and adolescence is associated with maladaptive social and emotional development, academic failure, and criminal justice system involvement (Cho, 2009; Huebner, & Gustafson, 2007; Murray, Farrington, Sekol, & Olsen, 2009). Children of incarcerated women, when compared to those of incarcerated men, are more likely to have witnessed criminal activity and their parent’s arrest – experiences which are associated with anxious/depressed behavior problems and post-traumatic stress symptoms (Dallaire, & Wilson, 2010; Phillips, & Zhao, 2010). Children’s living arrangements are also much more likely to be disrupted when their mother is incarcerated as opposed to their father (Glaze, & Maruschak, 2008). Moreover, periods of separation from mother and alternative caregiving may precede the arrest by one to three years (LaLonde & George, 2002).

Though compelling in terms of the need to protect children, results of studies to date obscure potential effects of maternal incarceration specific to developmental stages. Little information has been provided about infants and toddlers, arguably the most vulnerable group within this population. An estimated 6% of the children of incarcerated mothers are below 2 years of age (Glaze, & Maruschak, 2008). Approximately 4–10% of incarcerated women are pregnant upon intake into a correctional facility (Greenfeld, & Snell, 1999; Glaze, & Maruschak, 2008; Pollock, 2002).

Disproportionately high rates of insecure attachment either to the incarcerated mother or the alternative caregiver in the community seen in preschoolers and younger school age children suggest the need for interventions to promote stable, sensitive caregiving in this younger group (Poehlmann, 2005). Insecure attachment is a stronger predictor of future suboptimal development for children growing up in risky environments than for children living in low-risk conditions (Belsky, & Pasco Fearon, 2002). Nurturing the caregiver-child relationship may be particularly critical with vulnerable and marginalized children.

Commonly available contact methods used with older children of women involved in the criminal justice system, such as brief visitation, letters, and phone calls, are not developmentally sufficient for infants and toddlers. The need to tailor contact methods to the developmental level of children has not been fully explored by departments of correction (Poehlmann, Dallaire, Loper, & Shear, 2010). Intermittent and fragmented methods are not likely to support the creation or maintenance of a secure attachment relationship. Didactic parenting programs, which are provided in many correctional facilities, may also be insufficient in this regard, as children are rarely included (Loper & Tuerk, 2006). Transportation and geographic isolation of women’s prisons create barriers to family visitation and to the implementation of parenting programs with a visitation component (Pollock, 2002).

Prison nurseries as a little known potential intervention

Prison nurseries are rarely used as an intervention to support the mother-child relationship in infants and toddlers of incarcerated women, and yet interest in these programs has grown (Goshin & Byrne, 2009). While their history has been erratic and fraught with barriers to sustainability, with between 1 and 13 states supporting nurseries in any one year since 1901, eight states currently allow incarcerated women to care for their infants in special segregated prison nursery units (Byrne, 2010; Women’s Prison Association, 2009). In early 2010, the U.S. Department of Justice (Office of Justice Programs, 2010) released a call for grant proposals to support the creation, implementation, or expansion of prison nursery programs.

The history of U.S. prison nurseries has been widely described, as have individual state programs and the demographics of women residing in them (Boudouris, 1996; Carlson, 1998, 2001; Gabel & Girard, 1995; Kauffman, 2006; Parke, & Clarke-Stewart, 2003; Pollock, 2002; Shepard, & Zemans, 1950). Descriptions of current U.S. programs suggest enriched, developmentally appropriate environments, segregated from other prison facilities. Programs are staffed by civilians and correction officers. They generally focus on developing the relationship between incarcerated mothers and their infants, promoting child development, and providing the mother with parenting and life skills education (Carlson, 2001, Fearn, & Parker, 2004; Kauffman, 2006; Silverman, 2005). Programs seek to normalize the nursery experience for mothers and children, such as providing first birthday celebrations (Silverman, 2005), although this event can be particularly poignant when it also marks the beginning of statutorily mandated separation at child age of 12 months (Byrne, Goshin & Joestl, 2010).

Measured attachment and development outcomes of prison nursery programs for children in early developmental stages

Results of the first longitudinal study of children who resided in a U.S. prison nursery provide evidence of positive infant, toddler, and post-release preschool outcomes. Children in this group had higher-than-expected rates of secure attachment during infancy and toddlerhood (Byrne, Goshin, & Joestl, 2010). For children in the prison nursery who reached their first birthday (the earliest age at which attachment can be reliably determined), attachment was measured under laboratory conditions inside the prison using the validated and well-established Strange Situation Procedure (SSP; Ainsworth, Blehar, Waters & Wall, 1978). Seventy-five percent of these children were classified secure by blinded certified SSP coders. These results were compared with those of meta-analyzed studies with both low-risk community samples and samples with psychosocial risks similar to those experienced by the mothers residing on the prison nursery (van IJzendoorn, Schuengel, & Bakersman-Kranenburg, 1999). The proportion of infants who achieved secure attachment in the prison sample did not differ from those reported for 15 studies of low-risk community residing infants less than 24 months of age. Moreover, the prison sample contained significantly more secure infants than the proportions reported in seven studies of low socio-economic infants, nine studies of infants with depressed mothers, four studies of infants with drug and/or alcohol abusing mothers, and five studies of infants with maltreating mothers (Byrne, Goshin, & Joestl, 2010).

Children released before a year in the Byrne study and residing in communities within travel distance of the research team were similarly tested with the SSP under laboratory conditions at the University research office after reaching their first birthday. While also achieving secure attachment in proportions comparable to low risk community samples, less of these children were likely to have secure attachment to their mothers than children who remained in the nursery with their mother for a year or more (Byrne, Goshin, & Joestl, 2010). This group was similarly compared to meta-analyzed studies of healthy and high-risk community children (van IJzendoorn, Schuengel, Bakersman-Kranenburg, 1999). The proportion of these early released children who achieved secure attachment also did not differ significantly from the meta-analysis results for 15 samples of healthy, low-risk community children under two years of age and was significantly higher than those reported for the samples of infants raised by maltreating mothers in five studies. They were, however, similar to some of the higher risk samples, including the low socioeconomic samples in seven studies, the infants of depressed mothers in nine studies, and the infants of drug and/or alcohol abusing mothers in four studies (Byrne, Goshin & Joestl, 2010).

This suggests that even infants whose mothers have multiple risk factors threatening the transmission of secure attachment to their infants can, nevertheless, be nurtured toward security in a prison nursery environment with deliberate parenting supports. The success of the longer co-residing infants also suggests that postponing re-exposure to the multiple environmental risk factors in the offending mothers’ families and communities and replacing this with a living arrangement that includes daily supportive parenting resources, even in the restrictive corrections situation, may provide some measure of protection with regard to the process of infant attachment which takes place incrementally over the first years of a child’s life. Findings were especially striking given the low rates of secure attachment in the mothers (Borelli, Goshin, Joestl, Clark, & Byrne, 2010). As measured by the Adult Attachment Interview, two-thirds of mothers in the cohort studied had internalized representations of insecure attachment from their own childhood experiences. This proportion is the inverse for secure and insecure categorizations typically found in healthy, low risk community samples.

Developmental outcomes have rarely been reported for the preschool period for infants raised in a prison nursery. It is known that attachment security is associated with both short- and long-term optimal child development (Sroufe, Egeland, Carlson & Collins, 2005). A rigorous follow-up study of behavioral development for infants enrolled in the nursery study was done during their preschool years. During the preschool period, children in this cohort had lower anxious/depressed behavior problem scores than children from a large national dataset who had been separated from their mother during infancy or toddlerhood because of incarceration (Goshin, 2010). This result remained even after controlling for risks in the child’s environment, such as parenting stress and caregiver substance use.

Separations following prison nursery co-residence

Virtually no reports exist regarding the long-term caregiving trajectories for children who co-resided with their mother in a prison nursery. In all situations continuous contact with the mother, who has been established as the primary caregiver and secure base during infancy, is supportive of continued attachment security. Separations threaten the evolving neurobiological attachment system (Bowlby, 1973; Solomon & George, 1999). It is often assumed that women and their children are released together and that the mother remains the primary caregiver (Bruns, 2006; Fearn, & Parker, 2004; Kauffman, 2001), but this may not reflect actual patterns. Approximately one-third of children in Nebraska’s prison nursery between 1994 and 1999 were released before their mother, with equivalent proportions leaving at the request of the mother or under order of prison administration (Carlson, 2001). Slightly less than one-third of children who lived in this nursery did not remain with their mother after release (Carlson, 2001). The timeframe of post-release follow-up and the children’s alternate caregiving situations were not described. Knowledge of post-release caregiving patterns in this population is needed to inform program planning within prison nurseries and parenting support in the reentry period.

STUDY DESIGN AND METHODOLOGY

Setting and Context

The setting for the longitudinal prison nursery co-residence study comprised Bedford Hills and Taconic Correctional Facilities, respectively maximum security and medium security prisons for women within the New York State Department of Corrections and Community Supervision (DOCCS). The State’s nursery program was implemented only in these two settings at the time of the study (and has since been discontinued at the Taconic facility). Bedford Hills is also the reception center for all incarcerated women in New York State. Preliminary screening of women entering the prison system is done and appropriate placement in one of the state’s prisons is determined in accordance with security classifications. Women accepted into the nursery program, however, can remain at Bedford to participate in this program even if they are not classified for maximum security, which is typically the case. Transfer of women accepted into the nursery program also periodically occurs from the Bedford to the Taconic prisons during the years of the sentence. The nursery program operates under the same prison directives at each site and the physical features, regulations, programs, and resources are similar across sites, which are located geographically across a two-lane road from one another. Including both sites in the study facilitated enrollment and following of women longitudinally with no loss to follow-up related to draft protocols.

Corrections Law, Article 22, Section 611

New York is unique among the eight states that currently provide prison nursery programs because it operates by virtue of the State Corrections Law, Article 22, Section 611 (Births to inmates, n.d.). Signed into law in 1930 by then Governor Franklin D. Roosevelt, CL611 requires that pregnant women confined in designated state or local correctional facilities be removed from these places for the anticipated birth and “…provided with comfortable accommodations, maintenance and medical care elsewhere…” (CL611, paragraph 1). CL611 specifies that: “…a child so born may be returned with its mother to the correctional institution in which the mother is confined ….for such period as seems desirable for the welfare of such child, but not after it (sic) is one year of age …,” with extensions to eighteen months of age possible for mothers in a state reformatory who would be paroled within that time (CL611, paragraph 2). Once the child is placed in the corrections institution with the mother, continuation of this arrangement is specified to be entirely in the control of the institution’s chief officer: “The officer in charge of such institution may cause a child cared for therein with its mother to be removed from the institution at any time before the child is one year of age….” (CL611, paragraph 2). No specific conditions are detailed for removal.

Two New York State Supreme Court rulings in the 1970s held that the “best interests of the child” standard applied to community custody cases must also be used to determine whether a pregnant inmate could keep her child. Since that time, while obligated to screen all applicants to the program, correctional authorities are duty-bound to apply this standard when making decisions. Theoretically this ensures that as many women as possible can take advantage of the support and services offered to the new mothers residing in the program. In practice it has not been disclosed how consistently this principle is applied. It should also be noted that CL611 does not speak directly to the creation of a nursery program.

Sample

The sample for this analysis of separation patterns is comprised of 97 women and their 100 infants (three sets of twins) all of whom were screened and accepted by DOCCS and allowed to co-reside from the child’s birth in the New York State prison nursery program in accord with NYS CL611. All women gave informed consent to participate in a larger longitudinal study of maternal and child outcomes funded by the National Institutes of Health (Byrne, n.d.; National Institutes of Health, 2011) These women were enrolled in the study over the course of three years from study outset until the target enrollment number of 100 dyads was met. All women accepted into the nursery program were eligible. Only one declined to participate. Therefore, this sample can be viewed as a cohort of women who were accepted by DOCCS to participate in one state’s prison nursery program during one three year period of time. They do not represent proportionately the population of women who participated in the prison nursery program over its existence for more than a century and were not selected to do so. However, patterns of separation and reunion can be identified for this cohort and inferences made for this contemporary time frame.

The study was reviewed and approved by a full Institutional Review Board including a prisoner advocate, in accord with all the Federal requirements to protect vulnerable populations and prisoners as defined by the Code of Federal Regulations, Title 45, Part 46 (45 CFR 46). A Certificate of Confidentiality was approved by the Federal Department of Health and Human Services Office of Human Research Protection. This provided study participants with additional measures of confidentiality and gave the investigator the ability to resist inquiries made about participants.

Following written informed consent, 97 women agreed to participate in the study with their 100 infants (three sets of twins). All women were convicted of felony crimes. These were primarily illicit drug activities and property crimes and few were associated with violence. They included: attempted or completed criminal sale of controlled substances (32%), criminal possession of controlled substances (27%), burglary (10%), robbery (8%), grand larceny (7%), and smaller instances of driving while intoxicated, forgery, assault, weapons possession, and bail jumping. Anecdotally it is known a large percentage of women had prior criminal histories including jail and prison time but this information has not yet been objectively substantiated for all. History of substance abuse was reported by 77%. Study participants represented black (40%), white (36%), Hispanic (21%) and bi-racial (4%) ethnic groups. Women ranged in age from 18 to 45 years with a mean age of 28.1 years. About half were not first-time mothers (48%) and had between 1 and 8 children outside with a mean of 2.5. Of these, 57% had been the primary caregiver for all their children and the remainder had one or more children being cared for outside their home. Fifty-eight percent of the women spent some time in the nursery while pregnant, averaging three weeks but ranging from less than a week to six months. The remainder was admitted to the nursery directly from the post-partum unit at the community hospital contracted for deliveries.

Infants comprised 54 females and 46 males. They mirrored slightly different ethnic proportions than their mothers with more reported as bi-racial (21%) and less white (23%) reflecting their parental heritage. All but three infants were born at full-term. Prematurely born infants were delivered at 34–36 weeks gestational age and were reunited with their mothers in the prison nursery following brief and uneventful neonatal intensive care unit stays. Average age at infant return to community was 7.6 months but ranged from in utero to 18 months.

At the end of this prison phase 91 infants remained in the study and mothers agreed to have them followed during the first reentry year either with herself or with an alternate caregiver who also provided written informed consent. In spite of significant barriers to long-term follow-up, 75 of these children (82%) remained in the study with either mother or alternate caregiver at the end of their first reentry year. A continuation study was funded and began re-enrollment between the second and fourth reentry year for previous participants. All women who could be located (n= 52; 59%) from the initial study agreed to be in the continuation study with the target child, and provided written informed consent (Goshin & Byrne, 2012). The retention rate for this study matches rates commonly reported for lower risk populations (Robinson, et al., 2007) and is substantially higher than reports from longitudinal studies with recently released samples (Eddy, et al., 2001; Goldberg, et al., 2009; Lincoln, et al., 2006; Menendez, White, & Tulsky, 2001; Needels, James-Burdumy, & Burghardt, 2005). These longitudinal studies comprised multiple aims and methods beyond the scope of this report, the purpose of which is limited to providing the first description of separation patterns observed over incarceration and reentry years for this cohort of women and children accepted into a prison nursery program. For the 47 women who could not be re-located to re-enroll in the continuation study, longer term separation patterns were discovered through networking contacts established over the years between the research team and community agencies. Recidivism was validated for all through NYS DOCCS inmate lookup service and through Vinelink, both publicly available.

RESULTS: PATTERNS OF SEPARATION AND REUNION DURING REENTRY

Almost 60% of children returned to the free community with their mother and the majority of these remained with her at the end of the third reentry year. Mothering was not seamless, however, for all of these children nor for those sent into the community without their mothers (see Figure 1).

Figure 1.

Maternal Union and Separation Trajectories

Infants who left prison nursery with mother

Fifty-nine children co-resided on the prison nursery during their mother’s entire confinement there and went out to the free world with her when she was released. At that time it was the mother’s intent that she would be the continued primary caregiver for her child. This was also in keeping with the mission of the prison nursery program. At the end of the first reentry year 49 (83%) of these children were still with their mother. Four children were with alternate family caregivers. Mother remained involved for two of these children whose placements were anticipated to be for short periods of return to jails or substance abuse relapse. One caregiver was providing permanent care during an extended period of relapse. One child had resided with mother in a faith-based transitional community facility that had a long relationship with the prison nursery. This agency facilitated finding father of baby to take custody after mother left the child and violated parole to attempt reunification with older children resulting in a warrant for her arrest. The whereabouts of six children were unknown.

At the end of the third year of the continuation study, our most recent data point, participants were in their third through eighth reentry year depending on their enrollment date in the initial study. For ease of comparisons, longer term outcomes are reported here for each child’s third reentry year. At this point mother remained primary caregiver for 44 children (90% of those with mother at end of first reentry year). For 24 children the mother continued to parent seamlessly since release from prison nursery. Twenty experienced some separations from their child, 18 through criminal recidivism, and two during substance abuse treatment periods in facilities that did not accommodate children. During these separations all children were with consistent alternate family caregivers. One child was in foster care and the status of eleven was unknown.

Infants who left prison nursery before mother’s release from prison

Forty-one children did not return to the free community with their mother. Reasons ranged widely: death, mother’s request that her child be sent out, disciplinary actions taken toward the mothers, the corrections (and legislative) policy identifying one year as the usual maximum stay, and deportation. The most unusual and rare case is the infant death. One child reportedly acquired an upper respiratory infection, and was examined twice at the medical center contracted to provide pediatric health services. The child succumbed in the Emergency Department. Reportedly an autopsy was done, however results were not shared with the research team.

Separations during the nursery stay: Maternal request

Three women requested that DOCCS send their infants out to family caregivers before the maximum time on the nursery was reached. In all cases the mothers cited their perception that the child was not receiving adequate medical care. Two infants had been on the nursery with the mother for extended periods of time, 295 days and 481 days. The other had been on the nursery for 71 days. All three children went out to the care of a maternal grandmother and were still with her as primary caregiver at the end of the first reentry year. The mothers stayed involved in the lives of the children to varying extents. There followed brief complete maternal separations related to criminal recidivism in all three cases, and in one case an additional separation through participation in a substance abuse program without co-resident accommodations.

Separations during the nursery stay: Program changes

Five women were temporarily forced to separate from their co-residing infants on the nursery because they chose transfer to boot camp drug treatment options (Shock Incarceration) designed for more intense therapy and rapid release. At the end of the first reentry year, three of these infants were with family caregivers and two had reunited with the mother as primary caregiver. At the third reentry year these reuniting mothers continued to parent, seamlessly from their release. Of the three mothers whose children were with family caregivers, two had experienced drug relapse and reincarceration.

Separations during nursery stay: DOCCS removals

DOCCS separated fourteen mothers from their infants during the course of the usual prison nursery stay through disciplinary actions. Three of these children were immediately placed in foster care. One pregnant mother was released within a short time and before she delivered. We lost contact with these four. The remaining 10 children continued with family caregivers, including two fathers. By the third reentry year, four of these families remained in touch and enrolled in the continuation study. Of these, three mothers had reunited with their child and had resumed primary care, and one was anticipating reunion when she completed her sentence the following year. Another mother who did not re-enroll developed a conflictual relationship with the paternal grandmother who was providing excellent care to her twins, regained custody, and subsequently reoffended. The children were placed in foster care. Four other mothers reoffended and were lost to long-term follow-up.

Separations at end of nursery stay: infant age limitations

Ten children were sent out to the free community in advance of their mothers because they reached the maximum age for co-residence before the mother completed her sentence. Seven of these children went to family members chosen by the mother and three went to the previously cited transitional housing program. One was reunited with her mother when she was released in the eleventh month of the first reentry year at which time the other nine mothers remained incarcerated. By the third reentry year this mother and six alternate caregivers reenrolled in the continuation study. Two mothers remained incarcerated and their children remained with family caregivers, two additional children were reunited with their mothers following her release, two experienced brief reunions with mother disrupted by drug relapse, and three were lost to follow-up.

Separations at end of nursery stay: Deportation

Five children who remained on the nursery in mother’s care throughout her stay were separated from the mother at or near the reentry date because deportation proceedings awaited the mothers at the completion of their state prison sentences. The mothers were remanded to Federal immigration facilities without nursery co-residences and the infants were separated and needed placement with alternate caregivers. At the mother’s request one infant went out to family shortly before mother’s transfer because she wanted to receive information about how the baby was adjusting and feared she would lose contact with family once deported. She arranged for the baby to be sent to his father who welcomed her into his household and shared daytime care with a female elder closely associated with the mother’s family. Four others were released with their babies and on the day of release the child was put into the care of a family member (three cases) or the community transitional housing facility (one case). One of these women provided telephone information in her anticipated country of residence. Follow-up contact was successfully made half way through the first reentry year and the mother was primary caregiver for her child at that time. She discontinued study contact after that and was lost to follow-up. Through our networks with community agencies we learned that one mother had been deported rather quickly and succeeded in making arrangements for her child to leave the country with her. Another was similarly reported to be raising her child in the country to which she was deported. Two infants resided with family caregivers, one father of the baby in the U.S. and one maternal grandmother in Canada.

Similarly, one woman was separated at the end of her State sentence because she had to report to another U.S. jurisdiction for possible additional sentencing. She left her child in the custody of a caregiver at the transitional housing facility. Her case was dismissed and she returned to the primary care giving role, which she has maintained with their continued support.

DISCUSSION

This analysis of separations for one cohort of women and infants who co-resided in a state prison nursery program demonstrates through prospective and longitudinal data collection that seamless caregiving by the mother on the nursery and into the reentry years is possible. Yet there are formidable barriers to this optimistic goal at various times, including: during the nursery confinement, at the moment of reentry, throughout the first reentry year, and in the ensuing years. The assumption that secure infant attachment can be created in prison nurseries, even by mothers who themselves have poorly internalized models of security from their own parenting histories, has been validated by the Byrne longitudinal studies from which these separation data are taken (Borelli, Goshin, Joestl, Clark, & Byrne, 2010; Byrne, Goshin & Joestl, 2010). This data show, however, the assumption that prison nurseries result in seamless parenting is only partially validated.

Is seamless maternal care giving for offender populations a realistic expectation?

Continued primary caregiving by the mother is the unspoken goal of most nursery programs, and the likelihood that it will occur is factored into the selection policies of some states. This is done in most states by restricting the nursery program to women with shorter sentences and with less psychosocial risk factors related to parenting. Candidates are limited to those with anticipated release within 18 months of birth in California, Indiana, Nebraska and Ohio and 24 months in Illinois. However, maximum allowable stays may be shorter than these anticipated release times, for example, 15 months in California and 12 months in Nebraska. In South Dakota infants can only stay for 30 days. Within these policies some separation time at reentry is inevitable. Most prisons nurseries also disqualify women with histories of abusing children or child-related crimes and violent crimes in order to protect children not only from separation but, more critically, from imminent harm. Prenatal and parenting courses are required during the nursery stay with the expectation that they will enhance effective parenting and sustain it. In spite of these strategies to create seamless mothering, there are few long-term follow-up reports disseminated by any state with a prison nursery program to determine if this is an outcome. Earlier reports on small samples from New York and Nebraska provide evidence that maternal parenting is not universally seamless (Carlson, 2001; State of New York, 2002).

Legal precedent has upheld that CL611 be implemented within the principle of best interest for the child. In keeping with this principle, prison selection policies in New York exclude women with prior crimes against children, and initially deny women with a poor prison disciplinary record or with a history of violence, although these two conditions can be reviewed on appeal with possible exceptions made. The statute itself specifies authority for the prison medical official to exclude any woman from acceptance if physically unfit and puts later removal entirely at the discretion of the chief prison official. Results from our cohort suggest this discretion was exercised, perhaps more than anticipated and arguably not always in alignment with the best interests principle.

In spite of all precautions, and sometimes because of them, separations do occur in a wide array of circumstances. In our cohort when separations occurred, it was chiefly family members who stepped up to take over the care giving. In most cases this was a grandparent. However, alternative care giving by aging grandparents is hampered by poverty and chronic illness (Hairston, 2009; Hanlon, Carswell & Rose, 2007). We observed this in our sample as well as the situation that younger grandparents had to disrupt their lives to become caregivers, including postponing education they had begun to pursue later in life and grappling with combining the newly accepted child care with ongoing work and home responsibilities. Family members do provide cultural and social continuity for the child who is taken from his/her mother abruptly and, from the child’s perspective, unexplainably. For children who remained in the study over time, child development was measurably enhanced when these alternative relationships were continuous and stable but disregulated when children were passed around among multiple caregivers without one secure base.

All families used multiple social services during reentry years. This was true both for situations of seamless parenting and with interruptions. They required support for housing, income, and health care and generally received these from public funds. Mothers faced many setbacks in seeking education and employment especially during the drastic global economic downturn beginning in 2008. Personal relationships were also challenged as mothers tried to rejoin their own families and reestablish their mothering role in the community. The necessary counseling services were sparse and often not to be found.

Proportion of seamless parenting

Because of its legislative entitlement for access, New York State is more likely than other states to have a nursery population with wider psychosocial risk parameters that potentially threaten continued parenting. Within this context the proportion of seamless parenting documented throughout the prison stay and the entire first reentry year is a true victory for the 49 mothers who achieved it. Such continuous parenting approaches extraordinary for the 60 mother-child pairs who still remained together seamlessly or were reunited after brief interim separations by the end of the third reentry year. For known cases in later reentry years, the struggle to continue in the mothering role was continually challenged by poverty, underemployment, housing needs and other environmental factors, but drug relapse and recidivism took the major toll. Resources were sparse. In context, our brief but repeated and tailored phone and mail outreach was appreciated and effective. Unsolicited participant endorsements were numerous. Notably, in the continuation study, 75% of mothers receiving the experimental intervention achieved continuous or reunited mothering, compared to 64% of mothers in the control group.

Lessons learned from diverse range of separation and reunion patterns

The different types of separations merit comment although caution is advised for subgroups with small numbers. The few mothers who opted to send out their children all cited inadequate medical care but all expressed regret for their decision shortly after the child was removed. They also failed to reunite following release. Underlying this decision may have been self-doubt and ambivalence about mothering. Perhaps counseling could have altered these decisions and the outcomes. Opting for drug program changes that removed women from the nursery also may have represented some mothering ambiguity. Two of five mothers reunited. It is known that short-term confrontational drug programs may not be effective for women, so the lure of early release may be misleading in terms of long-term recovery.

The largest numbers of abrupt separations during nursery co-residence were related to decisions made by DOCCS to remove infants from the nursery. Circumstances left little time for custodial planning. The majority of children sent out in this way did not reunite with mothers. In cases related to disciplinary decisions the children were arguably not in danger, but the mother was being punished for repeated prison infractions such as refusing to immediately comply with a corrections officer’s direct command. Separations resulting in immediate foster care were lost to follow-up study, which is emblematic of child invisibility under these circumstances.

Separations related to deportation proceedings were beyond the State DOCCS authority. Reinstating prison nurseries in Federal facilities could prevent these traumatic separations in cases when the mother anticipates caring for the child in the receiving country.

The second largest number of separations in this cohort resulted when children reached the maximum age allowable for nursery co-residence. It is curious that this age was statutorily set at 12 months and that the extension to 18 months was rarely applied. The overwhelming conclusion of existing research in psychology, psychiatry, and child development is that abrupt separation from a primary caregiver before 18 months of age has lifelong effects on a person’s ability to establish healthy relationships and interact in a positive way with the world (Sroufe, Egeland, Carlson & Collins, 2005). The development of a healthy attachment between a child and his or her primary caregiver is a long process occurring over the course of a child’s first two years (Emde, 1989). The time between9 and 18 months is critical (Hazan & Zeifman, 1999). During this time babies pull together all of their earlier experiences into a basic understanding of how to relate to the world and regulate themselves in relation to others. Children who are separated from their primary caregivers during this period learn that they cannot depend on others to care for them and that the world is an unpredictable and frightening place. It is well established that frightening experiences early in life can lead to disorganization even in an established attachment (Lyons-Ruth, Repacholi, McLeod & Silva, 1991). Neurochemical studies show that disruptions to the attachment process affect the growth and development of the brain, as well as social functioning, aggressiveness, reaction to stress, and risk for substance abuse during adulthood (Pedersen, 2004). Separation traumatizes mothers, too. Women in nursery programs who have a history of insecurity that is slowly being healed through the nursery experience will have that growth undermined by separation from their infant especially when it is punitive and abrupt. Awareness of the nature and influence of family attachments should be increased among prison administrators and corrections officers so that they can commit to partnership in supporting these strengths during incarceration and reentry. Videodocumentary and photography are emerging as novel ways to increase such understanding (McShane, 2010; Byrne, 2010)

Limitations of the study

In drawing conclusions it is important to realize that the sample represents an historical cohort. There was no random selection of participants from a target population and other than in the ongoing continuation studies, no random assignment of participants to groups. Each mother/child dyad represents themselves, and the cohort represent one prison nursery population in one historical period. The demographics are not unlike those of two small samples reported by NYS DOCCS during the past two decades, which found more psychosocial risk factors than samples described by prison nurseries in other states. Loss to follow-up across our studies was minimal for a high-risk, vulnerable population but did result in more unknown outcomes over time for families who did not continue enrollment. This report of separation patterns is descriptive in nature and no causal attributions can be made.

Conclusions

A coordinated systems effort across corrections and community resources is needed to sustain the secure attachments, which have been demonstrated to be achievable in a prison nursery program that provides supportive assets. Separations threaten these outcomes yet occur for many reasons. On analysis these could be addressed by enhancing quality and availability of counseling services, parenting support, and substance abuse programs in corrections facilities and in the community.

Key Points for Family Court Community.

• Assumptions about prison nurseries relative to mothering are widespread, yet outcomes have rarely been studied especially in the United States where these programs are rare, diverse, and often transient.

• Recent studies during imprisonment provide data that adequately resourced prison nursery programs can foster secure infant attachment, protect child development, and reduce maternal criminal recidivism.

• Seamless mothering from prison nursery through long-term re-entry is possible.

• Maternal separations from infants do inevitably occur for prison nursery co-residents both during the program and in the reentry years.

• To sustain the positive maternal and child outcomes demonstrated to be achievable in a prison nursery programs a coordinated systems effort across corrections and community resources is needed with enhanced quality and availability of counseling services, parenting support, and substance abuse programs.

Biographies

Mary W. Byrne, PhD, DNP, MPH, FAAN is the Stone Foundation and Elise D. Fish Professor of Clinical Health Care for the Underserved, Columbia University School of Nursing and the Faculty of Medicine. She is the Principal Investigator for the NIH-funded longitudinal studies, “Maternal and Child Outcomes of a Prison Nursery Program” which provided the data for this paper. On the faculty of Columbia University School of Nursing for 20 years, she takes on a leadership role in the areas of interdisciplinary research, clinical projects, and doctoral dissertations and is Director of the Center for Children and Families.

Lorie Goshin, PhD, RN is an Associate Research Scientist at Columbia University School of Nursing. She has been a key member of Dr. Byrne’s research team for 5 years including Community Coordinator and Co-Investigator roles on the NIH-funded study that provided the basis for this paper. She is interested in community alternatives to incarceration for childrearing women and early preventive efforts to disrupt the pathway to prison for juveniles.

Barbara Blanchard-Lewis, JD is Staff Associate at the Center for Children and Families and a member of the research team implementing the NIH-funded study from which data in this paper is drawn. Her background includes being an attorney in the public sector and a civilian director of the Parenting Center at Bedford Hills Correctional Facility. She is 2011 winner of the New York Zero to Three Emily Fenichel Award for contributions to child welfare.