Attachment theory suggests that the concern of the caregiver is not to do beneficial things to the child so much as to create a secure environment, a secure ‘setting’. This is what Winnicott termed “the facilitating environment”.  It is a setting that is responsive to the needs of the client, and one where the client’s attachment needs are sufficiently well met for him to give up the anxious and stereotyped behaviour patterns that preclude the discovery of new and more successful behaviour.
It is all too easy to be drawn into trying to change the client by our own efforts and so overlook the needs reflected in the client’s insecure and ambivalent attachment patterns, and the worker’s own involvement in those patterns. It is also easy to exaggerate the import of the worker’s positive interventions for successful change. I suspect that many ‘successful interventions’ arise as a response to the client’s development rather than being simply an initiative from the worker.
This emphasis on the setting is one of the strengths of attachment theory for the social work task. It takes us out of the narrow confines of the counselling role, and emphasises the importance of attending to the physical and material ‘setting’ of the client. In this way, such work as welfare rights advice, employment preparation or help with accommodation can be integrated with work on emotions or relationships. Furthermore, this aspect of attachment theory may suggest why it has been so hard to demonstrate the relative effectiveness of different models of intervention by social workers. It will be remembered that those children who moved from insecure to secure attachment patterns, appear to have done so as a result of an improvement in the caregiver’s responses to the child along the four dimensions mentioned. Whatever the model of intervention therefore, effectiveness would depend more crucially on the existence of the characteristics of secure caregiving.
 Winnicott 1982 esp pg 223