Agency, place, and the real-world dynamics of health inequality

1. Introduction

Across societies and over time, patterns of morbidity, mortality, and wellbeing follow enduring social and spatial gradients. These are among the most consistent findings in population health and have informed decades of research and policy attention to the social determinants of health (Commission on the Social Determinants of Health, 2008; World Health Organization, 2025). Yet despite this substantial evidence base, progress in reducing health inequalities has remained limited. In England, for example, life expectancy improvements have stalled and geographic inequalities have widened since the early 2010s (Marmot et al., 2020). Comparable fractures are evident across other high-income countries, suggesting that health inequalities are not only deeply entrenched, but also resistant to conventional forms of policy action.

This persistence suggests that dominant ways of conceptualising and addressing inequality do not fully capture the processes through which health outcomes are produced. Much research has focused on identifying the social and environmental determinants of health and on designing interventions to address them. While this work has been essential, treating determinants as discrete and independently acting factors can obscure the systems through which social conditions are organised, mediated, and translated into health outcomes. Health inequalities endure not simply because disadvantage is unevenly distributed, but because the systems that convert social conditions into lived outcomes vary in their organisation, responsiveness, and capacity to adapt (Matheson, 2020; Matheson et al., 2024). Evidence from spatial variation reinforces this point. Excess mortality in Glasgow, for instance, exceeds what deprivation alone can explain (Walsh et al., 2010), while in Aotearoa New Zealand poorer health among Māori reflects historically embedded institutional and structural dynamics, including racism, rather than deprivation per se (Cormack et al., 2020). What matters, then, is not only exposure to disadvantage, but how places are organised to absorb, amplify, or mitigate its effects.

2. Systems, trajectories, and place

Castellani and Wistow's (2026) analysis offers a compelling illustration of these dynamics. Using case-based complexity (CBC) methods, they examine trajectories of healthy life expectancy across English local authorities and identify clusters of places characterised by distinct configurations of determinants and institutional conditions. Their findings show that inequalities cannot be adequately represented along a single gradient of deprivation. Places with similar socioeconomic profiles can follow markedly different health trajectories, depending on how local systems are configured and how they evolve over time.

This is an important shift. It moves attention away from asking only which determinants are present, towards examining how conditions combine, interact, and take effect in particular settings. In this account, place is not a passive backdrop against which inequality unfolds. It is an emergent configuration of relationships among governance arrangements, institutions, infrastructures, and communities. Castellani and Wistow thus make visible the ways in which inequalities arise through system behaviour: through patterns of coordination and fragmentation, feedback and inertia, adaptation and constraint.

This reorientation resonates strongly with wider developments in systems-informed public health (Moore et al., 2019; Rutter et al., 2017). It suggests that progress on inequality depends less on targeting isolated determinants in abstraction, and more on understanding and reshaping the relational architectures through which health is produced in practice.

3. Health inequalities as emergent system outcomes

Decades of research have established the importance of determinants such as housing, education, employment, income, and social participation for population health (Commission on the Social Determinants of Health, 2008). Yet interventions targeting these domains often produce uneven and sometimes disappointing results, with similar policy inputs yielding different effects in different places (Craig et al., 2008; Hawe et al., 2009). These discrepancies are not incidental. They reflect the fact that health outcomes emerge within complex, adaptive systems, where multiple components interact, co-evolve, and respond to intervention in context-specific ways (Moore et al., 2019).

These persistent gradients are the cumulative result of feedback loops, path dependencies, institutional routines, and uneven adaptive capacities operating within and across systems of governance, service delivery, social support, and local environments. Interventions that alter single variables without changing the relationships among them may generate activity, but they rarely shift the underlying dynamics that reproduce inequality. What becomes central, therefore, is not simply the modification of conditions, but the reorganisation of relationships through which those conditions acquire force (Mudd et al., 2024).

4. Place as a systemic configuration

Public health research has often treated place as context: something surrounding health, shaping it from the outside. But places are not static containers of risk or resilience. They are constituted through patterns of relationships which shape how resources circulate, how decisions are made, and how collective problems are interpreted and acted upon (Cummins et al., 2007). These relational configurations evolve over time, reinforcing distinctive local conditions that mediate how social disadvantage is experienced and responded to (Macintyre et al., 2002; Bambra et al., 2016, Bambra et al., 2019).

Castellani and Wistow advance this understanding by showing that local authority areas with comparable levels of disadvantage can nonetheless experience divergent health trajectories because of differences in institutional and social configuration. Their identification of clusters of places with shared system profiles reframes so-called “place effects” not as contextual modifiers, but as emergent properties of system organisation.

This view also aligns with Indigenous understandings of wellbeing that locate health within interdependent relationships among people, institutions, land, and environment. For Māori, for example, place is not merely physical location, but part of a living relational system through which whenua, culture, community, and wellbeing remain interconnected (Came et al., 2018; Durie, 2004). Similar insights emerge from work on inter-organisational collaboration, which shows that places with stronger relational coordination across sectors are often better able to mount coherent responses to complex health challenges (Checkland et al., 2012; Goodwin et al., 2014). Place, in this sense, is not simply where inequality happens. It is one of the mechanisms through which inequality is made and potentially unmade.

5. Agency and the capacity to shape system behaviour

If place matters because systems are organised differently, agency matters because actors within those systems have unequal capacity to shape how they function. Agency here is not reducible to individual choice or leadership alone. It refers to the capacity of actors and institutions to influence relationships, align resources, coordinate action, interpret signals, and adapt in response to changing conditions. This capacity is critical to whether local systems reproduce disadvantage or generate more protective and preventive trajectories.

Importantly, agency is not evenly distributed across places. It depends on the presence of institutional strength, stable organisational relationships, trusted leadership, useable information flows, and governance arrangements that enable action rather than fragment it. Research on collaborative governance suggests that places with stronger relational and organisational infrastructure are better able to mobilise sustained collective action (Ansell & Gash, 2007; de Leeuw, 2017). By contrast, fragmented systems constrain local agency, weakening responsiveness and limiting the capacity for coordinated adaptation (Goodwin et al., 2014).

Agency is also structured by power. The capacity to influence system trajectories depends on whose knowledge counts, whose institutions are resourced, and whose voices are able to shape decisions. Communities facing long-standing marginalisation are often those with the least formal influence over the systems most consequential to their lives (Lekas et al., 2023). This makes the distribution of agency itself a central part of the inequality problem. Inequality being sustained not only through differential exposure to adverse conditions, but through differential capacity to act on those conditions.

Seen in this light, Castellani and Wistow's analysis can be read as capturing not only variation in determinants, but variation in local capacity to mobilise and align around them. Their work helps bridge an important gap between structural accounts of inequality and the practical realities of intervention, showing that outcomes emerge through the interplay of social conditions and unevenly distributed agency.

6. Policy making and the challenge of proportionate universalism

Policy does not sit outside these dynamics. It operates within systems and, in turn, reshapes them. Governance arrangements influence how authority is distributed, how priorities are set, and how interventions interact with local capacity. Efforts to reduce inequality often falter because policy is treated as a technical input applied to places, rather than as a dynamic and relational process that is itself mediated through complex local systems (Mac-Seing & Di Ruggiero, 2024; Parbery-Clark et al., 2024).

The principle of proportionate universalism (Marmot et al., 2010) remains a useful way of conceptualising equity-oriented policy. Here, action on health should be universal, but implemented at a scale and intensity proportionate to need. But translating this principle into practice has proven difficult. One reason is that need is often operationalised through relatively narrow deprivation indicators, while variation in local system capacity remains less visible. As a result, policy may be calibrated to the distribution of disadvantage, but not to the distribution of capacity required to respond effectively.

This is where methods such as case-based complexity (CBC) approaches become especially valuable. By identifying clusters of places with distinct configurations and trajectories, they make visible the different ways in which local systems produce and reproduce health outcomes. Such methods offer a stronger basis for policy design, not simply because they reveal variation, but because they reveal variation in how systems work.

This insight echoes Meadows’ (1999) argument that some of the most powerful leverage points in systems lie not in shifting individual parameters, but in altering relationships, rules, and patterns of relationships. Policies that strengthen local decision-making capacity, improve relational alignment, and enable more adaptive forms of governance may therefore offer more durable gains than interventions targeted at single determinants in isolation. The challenge is not only to distribute more resources to places in need, but to invest in the capacities through which those places can shape their own trajectories.

7. Inequality as a question of system capacity and agency

Persistent health inequalities require more than a better catalogue of determinants. They require a more adequate account of how health outcomes are produced through the interaction of social conditions, institutional arrangements, and place-based system dynamics. Castellani and Wistow's work makes an important contribution by showing that divergent health trajectories are generated through distinct configurations of determinants and institutional conditions, rather than by deprivation alone.

From this perspective, agency becomes central. Health outcomes depend not only on the conditions people face, but on the capacity of actors, organisations, and communities to shape how those conditions interact: to coordinate, adapt, and act collectively within the systems that structure everyday life. These capacities are themselves unevenly distributed, reflecting deeper causes of inequality that lie in the relational infrastructure within and between places, which shape the capacity to act.

This has significant implications for policy and practice. It suggests the need for policy approaches that recognise the diversity of local system dynamics and invest in the relational and organisational capacities that allow places to respond: coordination, local decision-making, trust, learning, and collective action. Methods such as CBC do more than describe variation. They illuminate the architectures through which health trajectories unfold and, in doing so, point towards a more plausible route to change. Eliminating inequality will require not only improving the conditions of health, but reshaping the systems through which health is produced in the first place.

Ethics statement

No ethics approval was needed as this is a commentary and no data has been collected.

Declaration of competing interest

The author declares that she has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.