Is there a Complex Adaptive Systems approach behind the ‘Big Society’ and the Coalition Government’s attitude towards statutory professional regulation?

by Jo Rostron on November 3, 2012

[Note: This text was first presented on 3rd November 2012 to the New Lacanian School’s registered cartel ‘Jouissance & symbolic (dis)order’. This cartel is based on: Éric Laurent’s 2010 text: ‘The Symbolic Order in the XXIst Century, Consequences for the Treatment’. Further information & texts available here: C. ‘The Symbolic in the 21st Century’ working group (LW WG) ]

I am approaching this question as an Art Therapist employed in frontline child protection services who is registered with the Health and Social Care Professions Council (HCPC [i] ).  The current child protection system is being challenged and profound changes are beginning to take place that will soon affect all therapeutic practice with children and families.  Although my question only raises more questions, I hope that it will be considered by other psychotherapists, in the light of their own clinical practice and its principles.

Determining how to improve the child protection system is a difficult task as the system is inherently complex.  The abuse and neglect of children can be hard to identify because many of the signs and symptoms are ambiguous and possibly have other benign explanations. Managing this inescapable uncertainty is a problem that confounds child protection services around the world.  In Britain many children are currently subjected to intrusive and distressing enquiries but if families are finally deemed non-abusive they are offered no help.

In 1995 the Conservative government attempted to rebalance services so that more attention was given to the practical work carried out by professionals with children and families. [ii]  However, family support services failed to materialize at the rate intended and social work remained heavily biased towards child protection. [iii] The New Labour government took a wider approach after observing that many of the factors leading to abuse and neglect also resulted in poor educational achievement and increased anti-social behaviour.  Early help was again seen as the desirable goal and so the efforts of those working with children were harnessed within the policy outlined by the ‘Every Child Matters: Change for Children’ document. [iv]

Many of the Serious Case Reviews conducted when children who are known to social services are seriously injured or die, have been influenced by ‘hindsight bias’.  Child protection has followed the pattern of other enquiries in high-risk areas of work by concluding that human error was the problem.  The customary response has been to find ways of controlling people so that they do not make such mistakes.  This has exerted psychological pressure on professionals to try harder, reduced the scope for professional judgement by adding procedures and rules and, to ensure compliance with these, has increased the level of monitoring.

In June 2010, the Secretary of State for Education, the Rt Hon Michael Gove, requested an independent review of child protection in England.  The aim was to help reform the current CP system from being over-bureaucratised and concerned with compliance, into a system that keeps a focus on whether children are being helped effectively, and adapting when problems are identified.  The review [v] was conducted by Eileen Munro, Professor of Social Policy at the London School of Economics whose research interests are focused on how best to combine, in child protection, intuitive and analytic reasoning with risk assessment and decision making. She is also studying the role of the wider organisational system in promoting or hindering good critical thinking.  Her review was carried out over two years in consultation with a number of local authorities, local leaders, managers, practitioners and children and families.

The findings suggest that the prevailing strategies to manage risk in child protection are, paradoxically, making it harder for professionals to learn how to protect children better.  She identifies three factors that combine in such a way that they promote a culture in which professional practice is being excessively controlled and proceduralised:

the person-centred approach to investigating child deaths,

the blame culture, and

the performance management system.

Munro sets out proposals for reform that are intended to create the conditions that will enable professionals to make the best judgements about the help to give to children, young people and their families.  She recommends a systems approach to learning how to improve performance and offers a conceptualization of child protection services as a complex, adaptive system. [vi] This requires an acceptance not only of the complexity of the work but also the essential role of professional judgement.  She emphasizes the need for ‘feedback loops’ in the system where professionals are not afraid to communicate honestly about their experiences, both good and bad, and the need for senior managers who are capable of valuing this feedback as a source of learning.

A distinction can be made between a complex adaptive system (CAS) and a system that is largely mechanical in nature. [vii]  This distinction is fundamental, and key to the task of system design. In mechanical systems, we can know and predict in great detail what each of the parts will do in response to a given stimulus. Thus, it is possible to study and predict what the system will do in a variety of circumstances. Complex mechanical systems rarely exhibit surprising, emergent behaviour. When they do — for example, a jet explosion or computer network crash — experts study the phenomenon in detail with the aim of designing surprise out of future systems.

In a CAS, on the other hand, the “parts” (in the child protection system, as in health care, this includes human beings) have the freedom and ability to respond to stimuli in many different and fundamentally unpredictable ways. For this reason, emergent, surprising, and creative behaviour is a real possibility. Such behaviour can be for better or for worse; that is, it can manifest itself as either innovation or error. Emergent behaviour can occur at both the microsystem and macrosystem levels. An example of emergence at the microsystem level is the evolving relationship of trust between a child and a health or social care professional.  Affecting the macrosystem of child protection, for example, are four key driving forces [viii] :

  • The strong public reaction when a child is killed or seriously harmed
  • A belief that the complexity and associated uncertainty of child protection work can be eradicated
  • High profile public enquiries that focus on professional error without looking into its causes
  • Undue importance is given to performance indicators and targets which have skewed attention towards process, rather than to the quality and effectiveness of help given

These forces have created a defensive culture in child protection, in which statutory guidance, targets and local rules have become so extensive, in which bureaucracy is so demanding and services so standardized, that the system cannot provide the responses required to meet the variety of need presented.

Munro suggests a radical reduction in the amount of central prescription to help professionals move from a compliance culture to a learning culture, whereby they have more freedom to use their expertise in assessing need so that the appropriate help can be provided to families.  The Review recommends that the government revise statutory, multi-agency guidance to remove unnecessary or unhelpful prescription and focus only on essential rules for effective multi-agency working and on the principles that underpin good practice.

According to research on complex adaptive systems relatively simple rules can lead to complex, emergent, innovative system behaviour. For example, astrophysicists point out that all the beauty and complexity we see in the universe emerges from two simple rules: (1) gravitational attraction and (2) the non-homogeneity of matter in the early universe. In mathematics, the complexity and beauty of the Mandelbrot set (fractal mathematics) come from a very simple equation that is executed recursively.

Reynolds (1987) [ix] showed that complex flocking, herding, and schooling behaviour in animals could emerge from having each animal, such as a single fish in a school, apply three simple rules: (1) avoid collisions, (2) match speeds with your neighbours, and (3) move toward the center of mass of your neighbours.  No central controller or director is needed; each animal can simply apply the rules locally. The behaviour of the system emerges from the interactions, and this behaviour is successful in avoiding predators.

Holland (1998) [x] shows how simple rules lead to emergent complexity in game theory, which models many situations in human interactions.  Again, the concept of simple rules clearly links to notions based on evolutionary genetics, innovation theory, and other sciences that are embracing new ideas about complexity. Experience in the fields of creativity suggests that changing the underlying rules might result in great innovation (Plsek, 1999) [xi] .

This concept provides wide boundaries for beginning the work of self-organization. The task of complex system design does not need to be complex itself.  The answer is to create the conditions for self-organization through simple rules under which massive and diverse experimentation can happen.  Simple rules for human complex adaptive systems tend to be of three types: (i) general direction pointing, (ii)   prohibitions, and (iii) resource or permission providing.  A good set of simple rules might include all three types (which also tend to match the predispositions of many systems designers).

Those who are drawn to leadership and aim setting are drawn to the simple rules of the first type. Those who prefer regulation and boundary setting are comfortable with the second type. Those who focus on incentives and resources are drawn to the third type. The theory honours all three points of view and suggests that it is best to have only a few such rules, so that no one point of view dominates.

Because the parts of a complex adaptive system are adaptable and embedded within a unique context, every change within a CAS can stimulate other changes that we could not expect. This approach to system design can never provide the assurance that is possible in a mechanical system. This is the nature of a CAS. Therefore, rather than agonizing over plans, the goal is to generate a “good enough plan” and begin to observe what happens. Then, modifications can occur in an evolutionary fashion.

Some key questions raised by a CAS-inspired approach to redesigning health and social care include:

  • How can conditions in the system be established to allow many new ideas to emerge and mix into the existing system, while maintaining discipline to do just a little bit of nurturing, see what happens, then decide what to do next?
  • How can diverse people be brought together, information shared, and forums convened to stimulate creative connections amongst those who do not normally come together to do so (similar to genetic cross-over and mutation)?
  • How can desirable variation (innovation) be separated from the variation that ought to be reduced (error and waste)?
  • What are the few simple rules that might guide the local development of the health and social care system?
  • What is the implicit, existing set of simple rules from which current innovations in health and social care emerge?
  • How can these existing, implicit rules and underlying assumptions be modified?
  • How can communication infrastructures be set up to disseminate the new simple rules?
  • How can infrastructures be established in public policy to encourage experimentation and innovation under the new simple rules?
  • How can experimentation be made highly visible so that the “fitness” of each evolution can be judged to quickly spread the best ideas?
  • What is a “good enough plan” to begin the change?
  • Who should take on the role of continuing to evolve the plan as the CAS plays itself out?


Several examples of self-organizing innovation within the child protection system are outlined in the Munro Review of Child Protection.  Such innovations to support professionals have occurred within certain targeted local authorities to create a learning culture, where change is expected as a consequence of that learning.  They appear to provide evidence of an implicit set of simple rules [xii] that is already in place in this high-risk area of professional practice that underpin the recommendations for reform.  One of these principles included the observation that the system and the wider public need to acknowledge the uncertainty and risk that inevitably surrounds child protection.  Professionals must be ‘risk aware’ rather than ‘risk averse’, since there is no absolutely safe option.

To consider the future of inter-agency rules and guidance Professor Munro has convened a multi-agency working group consisting of representatives from social work, health, police and other professional strategic bodies.  The Risk Principles drawn up by the Association of Chief Police Officers (ACPO) have been adapted to refer to all those who work in child protection. [xiii]

1          The willingness to make decisions in conditions of uncertainty (risk-taking) is a core professional requirement for all those working in child protection.

2          Maintaining the safety, security and wellbeing of individuals and communities is a primary consideration in risk decision making.

3          Risk taking involves judgement and balance, with decision makers required to consider the value and likelihood of the possible benefits of a particular decision against the seriousness and likelihood of the possible harms.

4          Harm can never be totally prevented.  Risk decisions should therefore be judged by the quality of the decision-making and not by the outcome.

5          Taking risk decisions, and reviewing others’ risk decision making, is difficult so account should be taken of whether they involved dilemmas, emergencies, were part of a sequence of decisions or might appropriately be taken by other agencies.  If the decision is shared, then the risk is shared too and the risk of error reduced.

6          The standard expected and required of those working in child protection is that risk decisions should be consistent with those that would have been made in the same circumstances by professionals of similar specialism or experience.

7          Whether to record a decision is a risk decision in itself which should, to a large extent, be left to professional judgement.  The decision whether or not to make a record, however, and the extent of that record, should be made after considering the likelihood of harm occurring and its seriousness.

8          To reduce risk aversion and improve decision making, child protection needs a culture that learns from successes as well as failures.  Good risk taking should be identified, celebrated and shared in a regular review of significant events.

9          Since good risk taking depends upon quality information, those working in child protection should work with partner agencies and others to share relevant information about people who pose a risk of harm to others or people who are vulnerable to the risk of being harmed.

10        Those working child protection who make decisions consistent with these principles should receive the encouragement, approval and support of their organization.


[i] The HPC (Health Professions Council) became the HCPC (Health & Care Professions Council) when the social work profession joined it on 1st August 2012

[ii] Department of Health, produced by the Dartington Research Unit , (1995) Child Protection: Messages from Research, London HMSO : Available here

[iii] Aldgate, J. ‘Evolution not Revolution: Family Support Services and the Children Act 1989’ in ‘Approaches to Needs Assessment in Children’s Services’ (eds) Ward, H and Rose, W. p147-165, London, Jessica Kingsley

[iv] Government Green paper : Every Child Matters : September 2003 : available here   & Green paper: Every Child Matters, Change for Children, November 2004, available here

[v] The Munro Review of Child Protection: Final Report – a child centred system : May 2011 : available here

[vi] Munro, E. (2010) : Learning to reduce risk in child protection: British journal of social work, 40 (4), p1135-1151 : available here

[vii] Appendix B 
Redesigning Health Care with Insights from the Science of Complex Adaptive Systems by Paul Plsek : p313 of  Crossing the Quality Chasm: A New Health System for the 21st Century (2001) 
Institute of Medicine (IOM) Available here

[viii] The Munro Review of Child Protection: Final Report – a child centred system July 2012; Executive Summary 2 p 6  First published in May 2011 and available here  : Updated on 20th July 2012 as Munro review progress report: Moving towards a child-centred system : available here

[ix] ‘Flocks, Herds, and Schools:
A Distributed Behavioral Model’ by Craig W. Reynolds Published in Computer Graphics, 21(4), July 1987, pp. 25-34. (ACM SIGGRAPH ’87 Conference Proceedings, Anaheim, California, July 1987.) : Available here

[x] ‘Emergence: From Chaos to Order’ by John H. Holland : published by Redwood City, California Addison-Wesley : 1998 : Review available here

[xi] Plsek P. (1999): Innovative thinking for the improvement of medical systems: Annals of internal medicine: v: p438-444 : available here

[xii] Munro Review of Child Protection: May 2011: Final Report, Chapter 2 : available here

[xiii] ibid Chapter 3.18 : Available here

Related posts

Psychotherapy is imposed: Psycho-analysis© works: Psychoanalysis operates by Julia Evans on December 15, 2010 or here

Does rule through regulatory systems give protection from murderers? – No by Julia Evans on November 6, 2009 or here

More on the effects of using this government’s top-down power – Systems used to blame not re-calibrate and prevent (Child care) & the principles underlying Government action by Julia Evans on November 5, 2009 or here