The Future of Children, Princeton - Brookings: Providing research and analysis to promote effective policies and programs for children.

As special needs adoptions increased in the 1970s, so also did the overall percentage of adoptions ending in disruption (termination prior to legal finalization). For instance, the statewide disruption rate for California public agencies increased from 2.7% in 1970 to 7.6% in 1973.⁷ In Ontario province, the disruption rate increased from 4% in 1971 to 7% in 1978.⁸ A private New Jersey agency specializing in special needs adoption reported a disruption rate of 26% for placements made from 1975 to 1981.⁹

Disruption rates for studies conducted in the 1980s are generally higher than those for earlier studies. Fifteen percent of 1981 placements of children 3 years of age or older in 13 California counties ended in disruption.⁵ About 12% of placements made by the Oklahoma Department of Human Services from 1982 to 1985 resulted in disruption.¹⁰ A 1986 study in New York City estimates a disruption rate of 8.2% in the 12-month period following adoptive placement.⁷ A 1986 New England study of predominantly special needs placements yielded a disruption rate of 27% (64 disruptions in 235 placements).¹¹ On the basis of data gathered from five states, Urban Systems Research estimated the national disruption rate for special needs children to be between 6% and 20%.¹² Kadushin and Martin present data from eight studies of disruptions of infant placements. Of 34,499 placements, 648 resulted in disruptions, a rate of 1.9%. In eleven special needs placement studies, they found an overall disruption rate of 11.3% (502 disruptions in 4,443 placements).¹³

A national study of 700 adoptions of children with developmental disabilities yielded an overall disruption rate of 8.7%. For children in this group who were adopted when 7 years of age or younger, the disruption rate was only 3.3%; for those 8 years of age and older, it was 17.7%.¹⁴ In a recent Massachusetts study, 101 of 102 adoptions of developmentally disabled children remained intact.¹⁵

Some specific groups of children have extremely high disruption rates. Kagan and Reid report on a sample of 78 youths with a mean age of 11 years at adoptive placement, 91% of whom had experienced placement in an institutional treatment center. Of this sample, 53% experienced at least one adoption disruption.¹⁶ At the other extreme, a private Illinois agency with strong postplacement parent support group services experienced only 5 disruptions in 900 placements of older children.¹⁷

Taken on balance, these studies suggest that the disruption rate may be about 10% to 15% for children placed when older. For younger children with developmental disabilities, the rates are lower. The low percentages overall represent a high level of success, particularly considering that just 20 years ago, adoption would not have been an option for most children with special needs.

Disruption rates should be interpreted carefully because exact definitions differ from study to study. For many studies the follow-up time was limited and the percentage of terminating adoptions would be expected to increase with time. In Barth and Berry's studies, for example, the mean length of time from placement until termination was about 18 months.⁵

Although having experienced a prior disruption places a child at increased risk of subsequent disruption,^5,9 many children who experience disruption are successfully placed in another adoptive home. For instance, 41% of Arizona children placed from 1982 to 1985 who experienced an initial disruption were placed in another adoptive home.¹² In Oklahoma during approximately the same time period, 126 of 170 children who experienced disruption of an adoption or of a trial adoption were adopted by another family.¹⁸

Child's Age

The sociodemographic factor most powerfully associated with risk for disruption is the child's age at the time of placement. Risk increases with age.¹⁹ For instance, among placements implemented by the Colorado Department of Social Services from 1981 to 1984, the mean age of children whose adoptions were disrupted was 8.8 years, while that for children whose adoptions remained intact was 4.4 years.¹⁸ At Spaulding for Children, New Jersey, a private agency, disruption rates were 7% for special needs children from birth to 5 years of age at placement, 15% for those 6 to 8 years, 25% for those 9 to 11, and 47% for those 12 to 17.⁹ A study of more than 900 children placed in 13 California counties in the early 1980s clearly demonstrates the link between child's age at placement and risk. Percentages of adoptions ending in disruption were: 3 to 5 years of age at placement, 5%; 6 to 8 years, 10%; 9 to 11 years, 17%; 12 to 14 years, 22%; and 15 to 18 years, 26%.⁵ Older age of the adoptive parent(s) predicted stability in four studies,²⁰ although this finding is not consistent across all studies.²¹

Boys are modestly overrepresented in disruptions in six studies²² while no gender-associated differences are evident in seven others.²³ One study suggests that among boys there was a higher disruption rate for those placed when younger than 9 years, but among girls there was a modestly higher rate for those placed at 9 years of age or older.¹⁸

Disabilities and Behavioral Problems

Disabilities (developmental problems and serious medical conditions) do not appear to be major risk factors although results vary considerably from study to study. For instance, a 1986 New England study found that the number of mental, intellectual, physical, or medical problems of adoptees was higher in children whose adoptions were disrupted than in those whose adoptions remained intact.¹¹ In a 1988 California study of children placed when more than 3 years old, disruption was significantly higher among children with mental retardation, but not among those with physical disabilities or medical conditions.⁵ On the other hand, a New Jersey special needs study found that neurological impairment, mental retardation, and orthopedic disability were unassociated with risk for disruption and that nonorthopedic physical disability was associated with reduced risk.⁹ In a recent five-state study, physically and mentally disabling conditions were predictors of reduced risk.¹²

In contrast to developmental problems and serious medical conditions, emotional and behavioral problems are strong predictors of disruption.²⁴ Aggressive, acting-out behavior—as contrasted with inhibited, withdrawn behavior—is centrally linked to disruption.²⁵ For example, a 1986 New England study identified six behaviors that predicted disruption: sexual promiscuity, having physically injured others, stealing, vandalizing, threatening or attempting suicide, and wetting or soiling bed or clothes. Similarly, in California (1988) behaviors such as cruelty to others, getting into fights, threatening others, and hanging out with "bad" friends strongly discriminated between intact and disrupted adoptions.⁵ In Illinois (1991), sexual acting-out was the strongest preplacement behavioral predictor of disruption.²⁶

Ethnicity, Family Structure, and Sociodemographics

Four studies link higher education levels of one or both parents to modestly increased risk of disruption,²⁷ although four others do not show such a pattern.²⁸ Four studies demonstrate modest associations between higher income level and increased risk.²⁹ Income level and disruption risk were unassociated in other studies.³⁰ One study shows increased risk for fathers in professional occupations,³¹ while a second shows reduced risk.³²

Lower disruption rates for minority families have been observed in two studies.³³ For instance, a five-state study found that placements with minority parents represented 36% of all placements but only 18% of disruptions.³⁴ On the other hand, a larger number of studies demonstrate no association between ethnicity and disruption.³⁵

Two studies link adoptions by single parents with increased risk,³⁶ but four others show no such association.³⁷ Festinger found that six of seven placements with single fathers resulted in disruption.⁷

In general, the associations of ethnicity, family structure, and income and education levels to risk for disruption are weak. If a pattern is to be discerned from the sometimes contradictory findings, it is that lower socioeconomic status is associated with reduced risk.

One of the major changes in public agency adoptions over the past 10 to 15 years has been the increase in adoptions by the child's foster parents. Adoption by foster parents consistently predicts reduced risk.³⁸ In one study, foster parent adoptions represented 41% of intact placements but only 13% of disrupted placements.¹⁸ In a second study these percentages were 36% and 12%,³⁹ respectively.

Other predictors of disruption include the number of placements prior to disruption,⁴⁰ time spent in prior placements,¹⁶ and delays in the adoptive process.⁴¹ Results from several studies suggest that the failure to provide adequate background information on the child may be the strongest service-associated predictor of disruption.⁴²

It is accepted practice to keep birth siblings together whenever possible. Findings regarding this practice are contradictory: several studies suggest increased risk for sibling placements,⁴³ while others suggest reduced risk.⁴⁴ Sibling placement seems contraindicated when there are already other children in the home⁴⁵ but may mitigate risk when no other children are present. For example, Barth and Berry found no disruptions among 47 children who were placed in sibling groups when no other children were in the home.⁵ A 1991 British study found that the presence of other children in the home increased risk only when these children were close in age to those in the sibling placement group.^2l

Family and Child Characteristics

Flexibility (as contrasted to rigidity) in family decision-making patterns may reduce the risk of disruption.⁴⁶ Findings from many studies concur that unrealistic or unmet expectations of the child portend instability.⁴⁷ Even when the adoption social worker emphasizes the problems that may be encountered, the prospective adoptive parents may maintain idealized, unrealistic expectations. As one parent who had experienced a disruption stated: "We were told [about his problems], but we really thought we could handle this and anyway, our child would never act that way."⁴⁸ Such a comment emphasizes the need for realistic, detailed preparation and good background information.

Westhues and Cohen concluded that, if the father is "actively involved in parenting, and able to nurture and support the mother in her role, placements are more likely to be sustained."⁴⁹ Good support systems from family and friends,⁵ as well as religious participation, predict stability.⁵⁰ Finally, the child's experience of physical⁷ or sexual abuse prior to adoption may also predict disruption, as does a strong attachment to the birthmother.⁵¹

In summary, key predictors of increased risk for disruption include:

• Older age of a child at the time of adoptive placement
• Inadequate background information or unrealistic parental expectations
• Rigidity in family functioning patterns, in particular the father's noninvolvement in parenting tasks
• Low levels of support from relatives or friends
• History of physical and particularly sexual abuse prior to adoption
• Psychiatric hospitalization prior to adoption
• Acting-out externalized behavioral problems including sexual acting-out
• Adoptive placement with "new" parents rather than with foster parents

Risk is elevated only modestly for developmentally disabled children, particularly with good preparation of the parents. What would once have been barriers to adoption—low income and education, minority ethnicity, single-parent family structure—do not increase risk and, when studies of intact families are also considered, may be modest predictors of increased success in special needs adoption.