Introduction

The protection and welfare of children is a complex matter and a significant policy, political and societal challenge (Canavan et al., 2022). It is complex because of the multiplicity of factors which place children at risk; the challenges of collaborating with professionals, often based in different disciplines; the uncertainty of the work; the multiplicity of tasks which have to be carried out; staff shortage and lack of knowledge, experience and time for workers (Hood, 2014; Jansen, 2018). Fish and Hardy (2015) recognise the complexity, volatility and unpredictability in child protection, which cannot be eradicated. The mistaken belief that it can has led to defensive practice, a blame culture, risk-aversion, a strong emphasis on procedures and recording and command and control centralised decision making (Munro, 2010). Weberg (2012) believes that linear standardisation and control approaches disable organisations from being able to evolve at the speed that is needed to maintain relevance, to mitigate risk and defensive techniques and to manage complexity, including the psychological impact of the work on people and services.

Under circumstances of increased task or clinical complexity, where there is uncertainty about how best to assess present and future risk and to deal with the situation, complexity theory may offer potential to greater understanding of the processes involved in CPW and improved clinical outcomes and performance (Stevens and Cox, 2008). The key challenge is managing emergent reality through taking action and navigating interactions between agents, organisations and their ecosystems in a constantly changing environment (Rhodes et al., 2010).

In this article, we propose a practice framework for the potential application of complexity theory to child protection and welfare (CPW) work. The first section considers why a focus on complexity theory may be worthwhile. Here, we examine the definitional underpinning for complexity theory and synthesise the literature into a framework for interrogating complex CPW cases and systems. Then, we explore its application to a fictionalised case study, to consider how this framework might inform practitioners of the dynamics of complexity, which might help to support practice in real-time and retrospectively in case and system review. In doing so, we reflect on the challenges of applying complexity thinking and their mitigation.

Complexity theory and CPW practice

In seeking to establish a definitional base for complexity theory, we conducted a literature search using our institution’s library, covering all the relevant social science and related databases, including, for example, Scopus, Social Science Citation Index, PsycINFO, CINAHL, Pubmed, ERIC, JSTOR and Google Scholar. A variety of search terms were used, including for example: complexity theory + child protection + application; complexity theory + social care + application. This yielded a total of 958 publications. Publications were excluded where a clear definitional base for complexity theory was not discussed. A total of fifty-six Journal articles, thirteen books/book chapters, two sets of conference papers, forty-four health themed case studies and four case studies in child protection utilising complexity theory were examined. Our analysis involved two main steps. First, we identified key concepts of complexity theory as they appeared within the literature, including self-organisation, adaptation, feedback, non-linear interactions, agency, fitness landscape, emergence, path-dependency and bifurcation. These were the concepts that appeared consistently in the various overviews and definitions in the literature. Our second step involved identifying those concepts with best potential utility and applicability in understanding complexity in a CPW-specific context. Our start point in identifying these concepts was articles specifically focused on CPW (e.g. Stevens and Hassett, 2007; Hood, 2014; Kettle, 2017; Colvin and Miller, 2020). However, we also drew on our combined experience in CPW, one as a former service leader and one as an academic/researcher. This part of the analysis involved drafting and discussing scenarios to test the concepts’ value in understanding of the complexity of a child protection case. Here, in line with the intention of the article, a key filter was ease of comprehension at practice level. From this, we arrived at complexity theory concepts, which in our view allow a characterisation of the dynamic unfolding of a child protection case: self-organisation/emergence; path-dependency and bifurcation (Stevens and Cox, 2008; Hood, 2018). We then developed a practice evaluation framework discussed in detail further in the case study section.

There is no single theory of complexity or complexity science, rather multiple conceptualisations that share some concepts and tools. Pinheiro and Young (2017, p. 121) consider that a system is ‘complex’ when ‘it cannot be fully described through its individual components due to the non-linear nature of the interactions among them’. In complex systems, elements are dynamic and co-evolutionary, characterised by self-organisation, emergence and adaptation (Mitleton-Kelly, 2003). Boulton et al. (2015, p. 29) refer to complexity theory as the interconnection and interaction of actors, action and environment that ‘provide the fuel for innovation, evolution and learning’. Interactions create contingency and uncertainty about what the future will become and generate more than the sum of the parts, through self-organising, and evolving towards order (Anderson, 1999). Organisations, which are complexity focused, nurture communication and the exchange of information between agents which generates new behaviour in systems as a whole (Horvat and Filipovic, 2018). Morçöl (2012) considers that complexity theory is a descendent of system’s theory and that it adds new methods of investigating systems. Rhodes et al. (2010) argue for a systems analysis of complex systems as a precursor to analysis of complex systems’ dynamics. Mitleton-Kelly (2003) views complexity theory as enriching and building upon systems theory by articulating additional dynamic characteristics of complex systems and by emphasising their interrelationship and interdependence.

A consistent message from the literature is the need for more studies on the application of complexity theory concepts and in particular in the dynamic, unfolding context of child protection systems and cases in order to fully see the value of complexity in trying to manage risk, volatility and relationships in CPW systems (Hood, 2014). Three dynamic concepts emerged from the literature as consistently relevant in relation to CPW: self-organisation/emergence, path-dependency and bifurcation. We consider each of these in turn.

Self-organisation/emergence

Self-organisation/emergence refers to where iterative interaction between all agents from the systems involved brings about new structures, behaviours and/or arrangements or further instability as a result. It refers to the way that systems are held together by the components and relationships that exist within the system, rather than having order imposed by some external force (Cilliers, 2005). The key elements of self-organisation/emergence include the spontaneous drive to create order from the edge of chaos or the zone of creative adaptability (Kaufmann, 2017). The emphasis is on self-organisation from bottom-up into a complex whole. Emergence is the process that creates new irreversible unpredictable order. The new structures, ideas, relationships, practice and/or organisational forms become part of the history of individuals and institutions and in turn affect the evolution of those entities. Adaptation is noted as a consequence of the compulsion to self-organise. The propensity for organisations and systems to self-organise and adapt is increased by factors, such as variety and reactivity. Glor (2007) emphasises that the greater the variety of agents, networks and external environments in systems, their interactions and their willingness to engage in learning and acting on new ideas, the greater the propensity for adaptation. Reactivity is identified as sensitivity and openness to change and it is at its most acute at system bifurcation points elaborated in more detail below.

Self-organisation, emergence and adaptation are important to understand in a CPW context, where ‘technical rational’ approaches, which are based on an assumption of linear causality and reductionism and detailed procedures and practice guidelines, tend to pervade practice. But, as Hood (2018, p. 12) states, this ignores that ‘multiple elements interact and combine to produce system events which could not necessarily have been foreseen’. Complex human systems and interactions change and adapt in unpredictable ways in response to interactional and transactional processes and situations (Stacey, 2004). Different patterns or behaviours emerge, acting as attractors, towards which a dynamic system tends to evolve. These attractors in systems may affect outcomes.

There are no fail-safe risk-assessments, strategies or forecasts in CPW systems. Protecting children requires understanding of complex interactional processes within the boundaries of instability. The normative type of clinical response to risk is primarily qualitative and some quantitative risk assessments intended to predict probability and severity (Hood, 2014). These approaches assume a linear approach to reducing uncertainty. They have become increasingly prescriptive as the need for accountability has been emphasised. They frame risk as an issue that can be identified and managed through rational technical activity (Hood, 2014). Social care becomes ‘less about the “right decision” and more about a defensible decision’ (Pollack, 2010, p. 12). Approaches by Stevens and Hassett (2007) and Richardson (2014) have advocated for risk assessment to be considered through a complexity lens, focusing on process and systems not procedures and tasks, requiring attention to detail, identifying boundaries and inter-connected behaviours and outcomes and non-linear understanding of situations to advance flexibility and real problem-solving. To enable self-organisation and emergence to solutions-finding, Beresford (1991), for example, talks about professional judgement being applied to complex risk issues, aiming for a kind of ‘practical certainty’ but also leaving room for ‘choice and action’ within the boundaries of instability. Taylor (2017) also emphasises the role of professional judgement to ensure decisions informed by evidence in social work practice.

Path-dependence

Garud et al. (2010) and Vergne and Durand (2010) define path-dependence as where agents/actors become ‘locked in’ by self-reinforcing mechanisms into paths whose evolution is determined by contingencies (chance events). Key dimensions of path-dependence include that ‘Once locked in, actors cannot break out unless exogenous (external) shocks occur’ (Oakden et al., 2021). The capacity to do certain things and not others is dependent on system histories, starting points and feedback loops. Jorg et al. (2020) quoting Sydow et al. (2009) moot that path-dependence has three distinct phases. Phase I, Preformation, is characterised by choices that set self-reinforcing processes in motion. In Phase II, Formation, the dynamics triggered by the initial choice, make it progressively more difficult to reverse the course of action, e.g. the increase of a particular variable through positive feedback leads to its further increase. Pierson (2000, p. 252) states that ‘the probability of further steps along the same path increases with each move down that path’. A dominant organisational solution, a ‘path’, ‘an attractor’ emerge, making it more beneficial for other people to adopt and follow that institution. Indeed, it can become difficult for agents to detach from the path set as alternatives dwindle. In phase 111, Lock-In, the dominant pattern develops a quasi-deterministic character and can become fully bound to a certain path. In complex systems management, the focus of path-dependence is in its contribution to change as opposed to continuity.

Path-dependency is important to understand in a CPW systems context as practitioners and managers need to be able to recognise and reflect on historical starting points and choices which may be affecting their decision making in CPW systems. Institutional starting points can influence negatively and positively future practitioner and institutional response capabilities, e.g. how organisations resource and balance the needs for prevention, early intervention and protection and legal intervention (McGregor and Devaney, 2020) can influence practice behaviour. Also, at a case-level, practitioners and managers need to be cognisant of path-dependency so that they can recognise system histories and feedback loops which are influencing their decision making and stickiness to particular courses of action.

Path-dependency may lead to lock-in to particular paths or activities which can lead to inertia, but human agency internally or externally can still produce challenge and change into dominant organisational solutions. To maintain and create system flexibility, Schoon (2013) advocates for practitioners and managers to adopt polycentric forms of organisational governance that promote institutional diversity and the exchange of ideas and knowledge, building trust and social capital through repeated interactions, learning through experimentation and taking advantage of multilevel approaches to practice.

Bifurcation

Dawkins and Barker (2020) define bifurcation as where a system undergoing turbulence or instability (edge of chaos) oscillates between two or more points of equilibrium. The behaviour of the system grows increasingly erratic as the system moves far‐from‐equilibrium until the threshold of stability is reached which is marked by a bifurcation point. Bifurcations are marked by transitions, big jumps or changes, tipping points and stepped change, symmetry breaking events, dynamic incrementalism (Rhodes et al., 2010). Key dimensions of bifurcation include that complex systems reorganise themselves to deal with new information that would enable them to survive within the environment of increasing levels of turbulence. Attractors (patterns or behaviours) emerge as a source of order towards which a bifurcated system evolves over time, as moments of choice in a case or system’s evolution. A new equilibrium arises where the system aligns with attractors and self-reorganises. Order emerges around the new structure.

Another dimension of self-reorganisation through bifurcation is that as systems continually hover at the edge of chaos, very small changes and actions in a system can reproduce themselves and have massive impacts, over and above the expected norm, on the whole system in a complex fashion. This is referred to as the ‘butterfly effect’ as it points to an instance where ‘small actions can redouble many times over to create large effects’ (Overman, 1996, p. 81).

An understanding of bifurcation is important to practitioners and managers as there are many conditions in CPW cases and systems which challenge practitioners and managers. Factors, such as technically fraught areas of work and/or expectation gaps between service performance and stakeholder expectations, can trigger case or system instability. Symmetry breaking occurs when a previous mode of operation rapidly and dramatically gives way to another. This is marked by discontinuity and changes in the prevailing pattern, structure or tenor of activity and discourse. Attractors also have a role in supporting self-organisation and stability. Dawkins and Barker (2020) state that in the discontinuity that signals bifurcation, practitioners and managers need to consider various options to address the turbulence and sensitive conditions, in addition to, or in lieu of, intense scrutiny of current practice. Huneault and Otomo (2022) advocate for a partnership model at these points to help relevant agents to be adaptive and resilient rather than rigid. The effect of actions at these points can be difficult to determine, but actions can contribute to the emergence of attractors which change the response to an issue. However, if there is sufficient counter pressure from organisations and stakeholders, who support current practice, issues can retreat to a previous source of order or stagnate in a bifurcated state of turbulence and instability for an indefinite period of time.

Applying complexity theory—an Irish (fictionalised) case study

To explore how an understanding of the dynamics of complexity can assist practitioners and managers in responding effectively to complex child protection situations we have created a fictionalised case study, based on documentary thematic analysis of five published case reviews in an Irish context [Tusla National Review Panel Reports (2019–2020), Ava (2020); Declan (2020); Mary (2020); Simon (2019); and David (2019)]. We interrogated the case study, based on a set of questions on complex dynamics identified from the literature. These questions are not included with this article, but a copy of same can be sourced from the Authors on request.

The case study methodology is being utilised as its strength ‘lies in its capacity to explore social processes as they unfold in organizations’ (Hartley, 1994, p. 212). Our general approach to case-study research follows Ragin (2015, p. 187) and his emphasis on identifying a small number of cases because of their theoretical or empirical significance. Whilst we recognise and note the underreporting of the vast number of successful child protection social work interventions and their value in generating learning for theory and practice, our focus is on cases of system failure, again a regular strategy in case-study research. Thus, cases are often chosen because of their extreme nature (e.g. it exemplifies high levels of success or failure) or their typical or standard character (Ragin, 2015, p. 187).

The use of fictionalised case studies and vignettes is a feature of many disciplines [e.g., evaluation and programme planning (Walker et al., 2008); university–business partnerships (Edmunds and Boettcher, 2021); cultural geography (Rabbioso and Vanolo, 2017) and sexual and relationship therapy (Riggs and Bartholomaeus, 2018)]. Our choice of creating a fictionalised case study was motivated primarily by ethical concerns. Rather than to use one single real case already in the public domain and replete with data that would serve the illuminative purpose of this article, we took the position that it was ethically sounder to create a case-study reflecting an amalgam of the real experiences of five cases. Our intention in the case study is to encapsulate the complex causality which is often a feature of child protection cases and its management. We set out what happened in the case study and how practice might have altered if dynamic characteristics of complexity—self-organisation/emergence/adaptation; path-dependency and bifurcation—had been applied to case management in this scenario.

The case: Cata

The CPW service in Ireland is provided by Tusla, Child and Family Agency, which includes domestic, sexual and gender-based violence services and education support services. Tusla employs a range of Social Workers, Social Care Workers, Family Support Workers and education support staff. The Health Service Executive (HSE) in Ireland separately provides a range of primary care and therapeutic services for children and families, including psychology, child psychiatry, speech and language therapy and tertiary hospital services. Tusla and HSE fund a range of voluntary/community and private organisations to provide a range of parenting, family support and in-care services. The Cata case described here is situated in this context.

Cata case study

Cata (14 years), an only child, lived with her Mother, Mary who had substance abuse and mental health problems. Mary took an overdose and was attended to by Hospital and Adult Mental Health and Addiction Services (HSE). At that point, Cata was referred to the Social Work Tusla Service, who specialise in child protection and alternative care work. The Father, John, was not involved in Cata’s life. At the time, Cata had been expelled from school, had behavioural difficulties and was awaiting a Child and Adolescent Mental Health (CAMHS) (HSE) assessment. The case was categorised as ‘welfare’ by the case Social Workers. An initial assessment was recommended by the Social Worker but was not completed. Cata’s case was closed. The Tusla Education Support Service (TESS) (aimed at supporting and monitoring school attendance in Ireland), following a referral from Cata’s school, intervened and provided home tuition. There was no link between the Social Work service and TESS. A year later Mary died from an overdose. The family re-referred Cata to the Social Work Service. At this stage, Cata had substance abuse and behavioural issues. As Cata was attending a Youthreach Service (Youth Support Service in Ireland) and her case would have been placed on a lengthy waiting-list, the decision was taken by the Social Workers involved to close the case as Cata was already in receipt of supports. The Social Work service had no contact with Cata or the Youthreach or family support services. The Youthreach service had expressed concern regarding their ability to meet Cata’s needs. Cata lived between an aunt and friends. A family support service was to be put in place. Cata died as a result of an assault 3 weeks after her Mother’s death.

The Cata case analysis

The authors are going to focus on analysis of the fictionalised case study through a complexity lens, to explore how the dynamics of self-organisation/emergence/adaptation, path-dependency and bifurcation impacted and potentially could have impacted on practice in this case. We will also explore what renders it difficult for practitioners to create the conditions and enact a focus on the dynamic characteristics of complexity and what might help to support practice.

Self-organisation/emergence/adaptation in the Cata case

The dynamic characteristic of self-organisation/emergence/adaptation happens as a consequence of patterns of engagement amongst key agents in a system. It is evident that interactions in this case were characterised by linearity as there was a noted lack of information-exchange and inter-connectedness between the child and family and key agents in the system in response to both referrals to Social Work. The Social Workers did not contact the child and parent(s) in response to either referral. The decision to not conduct the initial assessments meant that the key staff internally and externally did not come together to discuss Cata’s needs in a holistic way. This resulted in a lack of cognitive diversity, variety and reactivity being brought to bear on working through the increased complexity of Cata’s needs at critical junctures. Hood (2014) discusses practitioners adopting an approach of viewing complex needs as being answerable by discrete categories of service provision. In that context, integrated working mainly involves agents coordinating individual service responses. This is what occurred in the Cata case, but here it would have been more valuable to have nurtured cross-agency and cross-sectoral narrative to encourage learning, distributed experimentation, self-organisation and adaptation in response to Cata’s holistic needs. The lack of inter-connectedness and interactional processes meant that services did not work collaboratively together to support Cata and her Mother.

A more integrated, transactional, non-linear response would have involved using more explicit business processes, e.g. initial assessment, network approaches, to enhance information and knowledge exchange to encourage more collaborative practice (Storch, 2002). Complexity theory would posit that dynamic interactional patterns, a focus on sense-making and constructive dialogue internally and externally would contribute to ideas generation and self-organisation to allow new solutions to emerge. If practitioners in the Cata case had adopted a more dynamic approach to understanding and fostering agents’ inter-connectedness and information-exchange from the outset, through bringing together all the services involved with Cata and her family, this could have altered their patterns of internal and collaborative organisation and adaptation in relation to this complex case. An emphasis on participatory practice between practitioners, children and families would have increased interactions and information-exchange at crucial points of change in Cata’s life.

But why was it difficult in this case to enact non-linear thinking internally and externally and how might that be mitigated? To understand this, it is important to examine how social care organisations interact and create, transfer and synthesise knowledge internally and across them. The main agencies involved in this case were Tusla Child and Family Agency, the HSE, Voluntary/community organisations, e.g. Youthreach and educational supports. We could surmise that this case potentially presented challenges to the Social Work Department in terms of managing the additional complexity that some of the service provision was originating in and being provided in the external to Tusla environment. There was also evident lack of connectedness between internal services, whose pattern was to respond individually to Cata’s needs. A practical step to assist in scrutinising where the complexity lay might have been for the Social Work Service to network map the relevant agents internally and externally and Cata and her family, highlighting their interconnections and differentiation with each other (Hood, 2018) and where services needed to increase their interactions, self-organisation and adaptability (Colvin and Miller, 2020). Bronfenbrenner and Morris’s (2006) bioecological theory that emphasises settings, relationships, personal characteristics and proximal processes is of relevance here as potentially is mapping vertical, horizontal, geographic and environmental complexity (Anderson, 1999). Operational systems in social care need to have a sufficient range of integrated responses to deal with complexity (Hood, 2018). They need to be able to creatively adapt internal structures, routines and processes and to focus on establishing alliances or partnerships to increase their service responsiveness (Schneider et al., 2017). Standardised business processes in social care and formal and informal networks can act as stabilising mechanisms for information flow and exchange internally and externally. The lack of interactions in this case adversely affected establishing where the fitness to care for this young girl lay. A more collaborative orientation would have enabled collective exploration of the boundaries of instability, the space of possibilities or the zone of creative adaptability, which may have led to better outcomes.

Path-dependency in the Cata case

In relation to the Cata case, the starting point or path followed, with the initial referral to the Tusla Social Work service, on retrospective analysis, was to classify this case as ‘welfare’. The decisions which followed (not to conduct initial assessments) were mooted as potentially related to resources, but may have resulted from inherent area historical responses and lock-in to how cases classified as ‘welfare’ were managed, threshold for intervention decisions. Despite evidence of complexity escalation in this case from the first referral to the second referral, the area response was similarly low with both referrals. The Tusla Social Workers followed the same path with both referrals. They did not interview the child and did not conduct an initial assessment, which would have been required standardised business practice. This lack of child-centred practice impacted negatively on integrated working, information-exchange and collective sense-making in this case.

At a minimum, if the Tusla Social Workers had evoked the path of their standardised business processes, for example, by following through on the initial assessment, this could have acted as a stabilising mechanism and as an attractor to bring agents together to collaborate to address the complexity of this case. Potentially, the SW Services’ historical response to ‘welfare’ cases was locking practitioners in to linear responsiveness and limiting their focus on examining risks and hazards and potential life outcomes for Cata in an open way. If the Social Workers had questioned in to what was driving key decision making, through the complexity lens of path-dependence, at initial referral points, e.g. through examining case and system history, starting points, reinforcement and challenge points, this might have enabled new paths and collaborative considerations to emerge.

It is important to consider what makes it difficult to enact a path-dependency approach to case and system management and what might support practice. A key difficulty in adopting a path-dependency approach is identifying the historic and initial conditions and contingent (or chance) events which are influencing the path taken; mechanisms that decrease the relative attractiveness of alternatives and exogenous shocks which potentially shake the system free of its history (Vergne and Durand, 2010). For example, perhaps constrained resources in the Cata case were sufficient to suppress alternative paths being followed. Perhaps the noted complexity of response to this case in the external environment disorientated the internal agency complexity of response. Agents involved have to be able, supported and willing to proactively engage in discussions and debate on their past history and how it might be impacting on their present state and future performance. Bringing active agency cross-sectorally to path construction is key, whilst continuing to work with contingencies in complex cases and systems.

This analysis would highlight a need for practitioners to frame their practice against a deep awareness of the dynamic factors which may be contributing to their decision making and lock-in to particular paths or ways of working and responding. As Richardson (2014) emphasises, it is important to map important variables, historical data, anticipated dynamics and preferred dynamics in complex cases. An understanding of path-dependency and how it can influence action could have encouraged a deeper focus on reflective and supervisory practice as a way of challenging into stickiness and openness in complex case and systems management (Ruch, 2007).

Bifurcation in the Cata case

In the Cata case, there were significant missed tipping points or bifurcation points, such as recognition of the impact of maternal health issues and Mother’s death on the well-being of Cata. Condorelli (2016) points to the importance of knowing and recognising the critical points at which a case goes into a chaotic situation, to understand what is happening in discontinuity. There was a sense in the Cata case of the Social Work Service not identifying the triggers for, and consequences of tipping points reached and not reacting differently or flexibly to crucial points of change and transition in the family. Cata seemed to hover in an ambiguous space between services. Both Cata’s Mother’s suicide attempt and subsequent death a year later were high risk situations, capable of de-stabilising equilibrium in any child’s life. Hood (2014) discusses the dichotomy between volatility, associated with high risk, and drift. In the Cata fictionalised case, practitioners persisted in responding to disequilibrium through individualised and disconnected service responses. If the Social Workers had viewed Cata’s situation through a complexity lens, with a frame of reference around understanding uncertainty, turbulence, disequilibrium and the effect of feedback, both positive and negative on a child’s life, this might have encouraged greater creativity and more focused risk management in their response to Cata (Hood, 2014).

We need to consider what makes it difficult to enact a focus on bifurcation and what might support practice. Laming (2003, para. 17.106) pointed out in the report on Victoria Climbie that ‘child protection cases do not always come labelled as such’. Child Protection organisations can inordinately rely on a focus on thresholds (with some clinical guidance) to identify levels of risk in child and family situations, particularly in relation to ‘welfare’ cases. This approach has been critiqued (Klein, 2008, 2011) as diverting focus away from the role of professional thinking processes in reaching decisions about neglect or welfare. There is complexity for practitioners in making judgements about tipping-points in child protection work. Kettle (2017) found that practitioners rarely have concerns about responding to individual incidents of concern, which they perceived to be bounded and as a consequence relatively easy to make sense of, which was how the Cata case was dealt with. It is when there is a growing accumulation of concerns that judgements and decision making become more complex. Difficulties in recognising a tipping point in relation to child protection work stem from the complexity involved when concerns start to accumulate. Tipping-points identified externally and noted to the Social Work service can also come loaded with anxiety which the service then carry, which also can impact on decision making.

Examples of what might support practice include avoiding a start again syndrome with every new referral, but ensuring that the complexity of a case is temporaneously and chronologically analysed, utilising supervisory and reflective practice (Kettle, 2017). This would involve increasing the level of scrutiny and evidence-seeking, and/or attempting to further engage additional services, or evoking formal decision-making fora, e.g. child protection case conferences, to assist in gathering information and the professionals involved with the family at crucial points of change and transition. The challenge of working with disequilibrium requires a focus on historical and predictive data. It involves holding firm, in volatile circumstances, to work towards rational and appropriate responses to complex situations in child and family lives. Dawkins and Barker (2020) describe the importance of pattern building, exploring or surmising issue movement and focusing on sense-making. There is a balance to be struck in managing complex child protection cases between the search for decisive evidence and action in response to key bifurcation points (Skotte, 2018).

Methodological limitations

A methodological limitation of this article is that the analysis presented is based on a fictionalised case study. A next step in a genuine testing of this complexity theory practice evaluation framework is for it to be studied in real-world contexts, for example, using an action research design (Seymour and Davies, 2002). Only then would it be possible to truly determine its applicability, operability and utility for the inherent complexity of child protection work.

Conclusion

This article aims to examine why and how we might adopt a complexity lens and apply complexity concepts to child protection work, both retrospectively and in practice. When we adopt a complexity lens to child protection work, using these three concepts of self-organisation/emergence/adaptation, path-dependency and bifurcation, this can assist practitioners and managers to adopt a holistic socio-technical approach to child protection practice (Munro, 2010). This approach supports practitioners, utilising their human expertise, to engage in interpretation and critical analysis of qualitative information, as opposed to relying on standardised business processes and thresholds alone. It enables practitioners to respond and act flexibly and adaptively to volatile situations, unexpected consequences and ‘crisis’ states that often accompany risks to children. CPW practitioners are challenged with managing volatility and unpredictability. The dichotomy in this work lies between the need to establish stability and control. A focus on the management of risk through a complexity lens can aid practitioners to fully work into the complexities of risk in the situation and to bring clarity to disparate accountabilities, which may help to alleviate staff pressure.

The challenges for practitioners, in adopting a complexity focus to practice, are increased in linear, bureaucratic, risk-averse operational systems, more concerned with targets, standards, specifications and the routinisation of roles. This is in contrast with approaches more focused on locating child protection work within a wider interpretative ecological framework, with an emphasis on relationships internally and externally and family support. Even though adopting a complexity lens to CPW work can be challenging, it is possible, using the complexity lens suggested in this article for practitioners and managers to consider risk, volatility and relationships in child protection work in a more open way in practice and through case review. Jansen (2018) stresses the importance for practitioners and managers to be prepared to work with what they cannot foresee. An understanding of complexity theory will not eliminate uncertainty or the feeling of not knowing but awareness of complexity and how to think about it may make the situations less overwhelming.

At the case level, a complexity focus allows practitioners to reflect in the first place on the system that they are part of and to consider who are the key practitioners, teams and networks that they need to interact with and how. A complexity focus supports integrated protective support and supportive protection practice, with its focus within a bio-ecological framework, with an emphasis on formal and informal networking between micro-meso and exo-macro systems (McGregor and Devaney, 2020). At the system level, a complexity focus reinforces the importance of interactional, transactional and collaborative approaches to CPW work. Complexity theory allows organisations to put a strong focus on networking, allowing for distributed decision making and the fostering of conditions for the organisation to quickly and effectively adapt to internal and external pressures, which aids change and innovation (Weberg, 2012; Klijn, 2008).