Attachment theory is currently a dominant belief system in child welfare agencies in the UK (especially social services and Cafcass). It has become strongly associated with the impetus to drastically increase the numbers and the speed of forced adoptions. Forced adoption is a simplistic and cheap process based on ‘child rescue’ philosophy that demonises natural parents/families, and dismisses the need for, and value of, therapeutic/supportive services to families. According to one eminent barrister, the current wave of forced adoption is a “social experiment on a massive scale”. There is no long-term follow up research in the UK about the outcomes of forced adoption. The logic of the process relies fundamentally on misapplication of attachment theory (and increasingly on similar misapplication of neuroscience).


Attachment is one component of a parent-child relationship. Attachment is the aspect of a parenting relationship that reflects the extent to which the parent is able to provide a secure emotional base from which the infant can feel confidence to explore the environment, and be a preferred person for the infant to return to if feeling vulnerable. The development of attachment theory (in the 1950s and 1960s) was based on observations of infants’ responses to separation and loss of mother figures. Bowlby (1951/1980) developed the theory in relation to maternal deprivation. The study of the Robertson’s (1953/1970) on children admitted to hospital construed reactions to prolonged separation as involving three stages of adaptation: protest, despair and then detachment. This resulted in significant changes in hospital visiting policies –previously parental visits had been restricted so that children would “settle”. Rutter (1981/1995) clarified that infants can form multiple attachments.


Theorists have developed classifications of attachment patterns or style. The ‘Strange Situation’ test was designed in the 1960s to artificially create brief separation stresses for infants to monitor their responses (Ainsworth et al. 1978). From such experiments the notion of secure and insecure attachment types was developed. Much of this research stems from studies in the USA in the 1960s/1970s on non-separated Caucasian mother-infant dyads. The research did not involve infants separated from their parents on a compulsory basis.


Attachment theory is currently a dominant theory in child welfare. Assessments of attachment (based on observations and pragmatic variations on the ‘Strange Situation’) including ‘story stems’ (with older children) are often used to inform child protection assessments. Specifically, observations of ‘insecure’ attachments (e.g. ‘avoidant’,  ‘ambivalent’ and ‘disorganised’) between mothers and infants are often used as indicators of future risk and as part of the rationale for non-reunification and compulsory adoption. However, there are grounds for concern about the validity of such observations particularly when they take place on only one or two occasions (‘snapshot’ assessments) – (Gillen, 2002; Dale et al. 2005).


There is no research base in attachment theory supporting predictions of risk to be made following observations of ‘attachment’ between mothers (and fathers) and infants who have been subject to forcible separation and often a prolonged period in substitute care (sometimes including more than one placement). That is to say, there is no standardised approach for the forensic application of attachment theory. When observations of ‘attachment’ are made in situations of separation, great caution must be exercised to take fully into account the impact of the separation itself (on both mother and infant) in respect of the observations that are made. Returning to an earlier attachment concept: to what extent is the observed behaviour a significant consequence of the separation (e.g. protest, despair or detachment) – rather than reflection of an innate parental deficit in being able to promote attachment responses?


In addition to the risk of failing to sufficiently consider the impact of separation, the clinical/forensic application of attachment theory in child protection assessments can also be subject to other biases. For example, it can be argued that attachment theory underestimates the significance of infant temperament/development on observations (emphasising parenting style instead). Also, research is clear that significant proportions of non-clinical populations of children and adults are classifiable as having insecure attachment styles. Insecure attachment styles in themselves are not necessarily pathological and nor do they necessarily imply risk (of child abuse). Indeed ‘insecure’ attachment styles can be adaptive and contribute to resilience in dealing with complex and stressful situations.


The eminent UK attachment researcher Professor David Howe has warned:


In child care and family work, classifying attachment styles, whether in children or adults, in itself has only limited value. It has to be remembered that in any normal population of either children or adults, around 40 to 50 pre cent of individuals will be classified as insecurely attached. Clearly, in most cases there is nothing pathological or clinically significant about being assessed as insecurely attached. In this restricted sense, not a great deal can be made of the suggestion that this child or that parent has an insecure attachment. (Howe, D. (2003) Assessments using an attachment perspective).


Courts are remiss in ‘rubber stamping’ forced adoption applications by local authorities (and supported by Cafcass) that are fundamentally based on attachment theory misconceptions.