What can complex systems theory reveal about social inclusion

Historically, mental health care was managed within asylums, where all aspects of life – daily routines, work and social interactions – were encompassed within the institutional setting. With the introduction of psychotropic medications, it became possible to manage many psychiatric conditions outside those settings, facilitating the process of deinstitutionalization. With the rise of evidence-based medicine and the momentum of the Decade of the Brain, mental illness increasingly came to be conceptualized like any other medical disorder – a condition to be fixed primarily at the biological level, with an expectation that people could re-enter society without the need for special accommodations. In practice, many individuals with severe mental illness remain now as socially and economically segregated as they were before in the asylum era – in single-room occupancies, homeless shelters or carceral institutions, often concentrated in impoverished urban neighborhoods. To put it bluntly, the asylum has not disappeared; but simply been decentralized and dispersed. This has led to a renewed emphasis on interventions aimed at enhancing quality of life through social inclusion. With this, we have circled back: many of the social challenges once managed within institutions now re-emerge as critical responsibilities of psychiatry, as described in the paper by Henderson et al1. This paper outlines seven key recommendations to improve social inclusion for people with severe mental illness: amending legislation, decriminalizing suicide and substance possession, improving the evidence base for social inclusion, increasing mental health staff’s knowledge of community resources, developing more flexible employment opportunities, reducing stigma, and enhancing conceptual competencies among professionals. Although each addresses important challenges, they are treated to some extent as isolated actions. I will argue here that they are better understood as interconnected parts of a complex system, and advocate for a paradigm shift going from isolated interventions to be tested in clinical trials to a dynamic, complex system characterized by multilevel interactions2. This shift can also be described as moving from a complicated to a complex system. Complicated systems – like assembling a jigsaw – are difficult but ultimately solvable through understanding each component part. With enough time and knowledge, one can predict and control the system’s behavior. In contrast, complex systems involve interconnected components whose interactions give rise to emergent, unexpected phenomena that cannot be understood in isolation. In the worst case, well-intentioned interventions focusing on one component of a system can lead to unforeseen consequences elsewhere. On the other hand, system thinking can help us identify leverage points where small changes may generate equally unexpected but positive outcomes, also termed the “ripple effect”. The Health Complexity Framework3 is a model designed to support systematic thinking about complex problems by connecting insights across biological, psychological, social and structural domains. It provides a structured approach to understanding complex health-related issues through three core dimensions: patterns, mechanisms and dynamics. These dimensions encompass seven defining features of complex systems: emergence, interactions, nonlinearity, interference, feedback loops, adaptation and evolution. Patterns refer to the observable outcomes – such as unemployment, homelessness, social isolation, and inequitable access to health care – that manifest among people with severe mental illness. Key questions to address at this stage include: How does social exclusion of people with mental illness vary across time, place and populations? Which subgroups, based on age, occupation, migration status, socioeconomic background or geography, are most affected? Mechanisms represent the underlying processes driving these patterns. The mechanisms involve intricate multilevel interactions, nonlinearity (where small changes can produce disproportionate effects) and interference, as factors amplify or dampen each other’s impact. For example, biological vulnerabilities, behavioral tendencies, social networks, and structural inequalities may interact to worsen exclusion. Critical questions include: How do these diverse forces combine and spread? Does exclusion propagate through social networks or within specific employment sectors? Are there critical thresholds in the way excluding mechanisms such as stigma evolve? Dynamics capture the system’s development over time, shaped by adaptation, feedback loops and evolution. Individuals, communities and institutions respond continuously to shifting conditions, sometimes reinforcing exclusion, other times fostering inclusion. Feedback loops may perpetuate cycles of marginalization or promote recovery, emphasizing that social exclusion is not a static condition but a dynamic process. Important questions include: How do political and social systems respond to historical change, and how do these responses affect experiences of exclusion? Is there a reinforcing loop between mental illness, social exclusion, and declining health? Feedback loops may be of special interest to identify effective targets of interventions and to avoid destabilizing the system in a harmful way. I will briefly discuss four potential feedback loops, where well-intended interventions may have unintended harmful consequences. First, incidents of violent behavior committed by individuals diagnosed with mental disorders are very rare. When such events do occur, they are often amplified by the media and, paradoxically, sometimes used by professional organizations to justify the need for additional resources. However, this focus on risk can unintentionally reinforce fear and stigma among the general public4. As a result, people with mental illness may become further marginalized and isolated, which ironically can increase the likelihood of violent incidents. Social marginalization itself is a significant risk factor for violence, and has been described in the US literature as part of the “deaths of despair” phenomenon affecting individuals living on the fringes of society. Second, as mental health care broadens to include activities such as creative writing, nature, music, spirituality, and social connection, there is an increasing pressure to validate them within a medical evidence paradigm. Questioning whether such ordinary human activities “work” can be subtly stigmatizing, implying that people with mental illness require special justification to engage in everyday life. Rather than continually testing their effectiveness, the focus should shift toward ensuring equal opportunities for meaningful participation, recognizing these activities as valuable ends in themselves, not merely as instruments for improving health5. Although often well-intentioned, the medicalization of everyday activities can reinforce stigma and hinder the development of inclusive societies that ensure broad access to meaningful engagement for all – and thus ultimately serve the opposite purpose. As a third example, the Danish anti-stigma campaign One of Us has been widely recognized for its efforts to reduce stigma surrounding mental illness through personal storytelling and public engagement. However, some scholars have raised concerns that, despite its good intentions, the campaign may inadvertently reinforce stigma6. They argued that these campaigns, while aiming to promote inclusion, might unintentionally create a narrative where individuals with mental health conditions are expected to overcome challenges through personal effort alone. This perspective could inadvertently place the burden of recovery on individuals, neglecting the importance of societal intervention. Further, by including primarily people with less severe disabilities, those with more severe conditions risk further stigmatization. Finally, efforts to destigmatize mental illness by framing it as “a disorder like any other” – often rooted in the idea of a “broken brain” – have, perhaps counterintuitively, been shown to increase stigma rather than reduce it7. This biomedical narrative was intended to legitimize mental illness by aligning it with physical health conditions such as diabetes or heart disease, emphasizing that it is not a personal failing but a biological dysfunction. However, research has found that, while such explanations may reduce blame, they also tend to increase perceptions of chronicity, unpredictability and dangerousness7. These examples of unintended consequences from interventions within complex systems help illustrate the value of understanding social inclusion as a complex, dynamic process. With this, I support the current movement to embrace this complexity2 by drawing on a diverse range of epistemic perspectives and integrating multiple levels of the system. This calls for a pluralistic methodological approach8 – one that combines nomothetic (generalizable) and idiographic (individualized) research, blends quantitative and qualitative methods, and actively incorporates knowledge from various disciplines and lived experiences. This approach moves beyond reductionist causality and acknowledges uncertainty, emergence, and context-dependence as intrinsic to mental health and social inclusion. Combining methodological pluralism with the Health Complexity Framework offers a promising next step in advancing the biopsychosocial model. This model has sometimes been criticized for lack of operational clarity9, as it fails to specify how the interplay of biological, psychological and social factors can be studied. If successful, a conceptually sound biopsychosocial model may be the next step to qualify research on social inclusion of people with severe mental illness.