From Collective Intelligence to Collaborative Political Intelligence:
Toward Politically Durable Systems Change
What if the central problem in systems change is not our ability to analyse problems, generate learning, or coordinate action, but our ability to collectively interpret how systems adapt politically once change efforts begin?
Because if systems continuously reorganise around efforts to influence them, then the challenge may not be complexity itself, but recognising how power responds once interventions begin interacting with existing political and institutional realities.
And this becomes especially important when thinking about politically durable change.
Because politically durable change rarely emerges simply because a reform is technically sound, a strategy is well designed, or actors broadly agree on the problem. It emerges when new behaviours, expectations, and forms of coordination begin reproducing themselves across networks over time.
Yet this is precisely where many reform and systems-change efforts struggle. Not because they fail to generate insight, learning, or collaboration, but because systems rarely respond passively to attempts to change them. They adapt politically.
From Understanding Systems to Interpreting Political Adaptation
Over the past decade, my own thinking around systems change has gradually shifted.
Like many others working across governance reform, systems leadership, collective intelligence, and civic innovation, I became increasingly interested in the limits of linear approaches to change and the growing move towards complexity, emergence, distributed coordination, adaptation, and learning.
Much of this resonated deeply with me because it reflected something I repeatedly encountered in practice: systems do not change simply because better technical solutions emerge. They change through relationships, incentives, behaviours, coordination, negotiation, and adaptation across networks of actors operating under uncertainty.
This initially drew me towards work on collective intelligence, particularly the recognition that no single actor sees the system in full and that systems change depends on distributed capacities to interpret complexity together.
Over time, however, I began questioning whether shared understanding alone was enough.
In earlier work exploring the limits of collective intelligence, I increasingly found myself drawn toward approaches focused less on how systems think together and more on how differently positioned actors work across uncertainty, fragmentation, competing incentives, and political difference in practice.
What increasingly interested me was that collective intelligence primarily helped explain how systems aggregate distributed knowledge and understanding. Collaborative Approaches, however, became important because they focused less on alignment and more on how differently positioned actors navigate uncertainty, disagreement, and competing political realities while still attempting to move forward together.
This led me towards ideas of collaborative intelligence and approaches such as Stretch Collaboration, which challenged the assumption that effective collaboration depends on alignment, agreement, or control.
And this distinction increasingly mattered because it reflected something I repeatedly observed in governance reform and peacebuilding work. Even where actors developed richer understandings of the system, interventions still struggled once they began interacting with existing incentives, institutional realities, and power structures.
This was where Political Economy Analysis became increasingly important to my own thinking. The broader Thinking and Working Politically movement helped expose how reform trajectories are shaped not simply by technical design, but by incentives, negotiated authority, informal norms, and adaptive political behaviour. In other words, the effectiveness of attempts to support change depended not simply on intervention design, but on the political and relational dynamics shaping how systems responded.
At the same time, institutional transition and systems designers were asking related questions from a different direction: how societies build civic infrastructures and distributed governance capacities capable of supporting coordination, adaptation, and collective navigation under conditions of complexity.
Taken together, these traditions profoundly shaped my thinking because they increasingly pushed the question beyond how systems understand themselves towards how systems collectively navigate political adaptation in practice.
Yet despite all this progress, I increasingly felt something important was still missing.
Much of the field still focuses primarily on how systems learn, collaborate, coordinate, or redesign institutions, while far less attention is placed on what happens once interventions begin interacting with existing political and institutional realities.
And this increasingly felt like the critical question. Because the moment interventions begin interacting with existing incentives, relationships, risks, and power structures, systems begin adapting around them. Authority shifts, informal practices evolve, compliance becomes selective, and actors adjust behaviour in ways that protect existing arrangements while still appearing aligned with reform. And it is often within these adaptive responses that the deeper political dynamics shaping the system begin to reveal themselves.
What Happens When Reform Meets the System?
Building on governance reform work in the Democratic Republic of Congo, alongside earlier work on political settlements in countries affected by instability and violence, I increasingly found myself paying attention to what happened once reform efforts began interacting with existing political and institutional realities.
Because this was often where the limits of many societal change approaches became visible. At first glance, reforms frequently appeared to be progressing. Meetings were held, procedures adopted, plans developed, and reports submitted.
Yet beneath these visible signs of alignment, something else was often happening. As reform efforts began interacting with existing incentives, authority structures, relationships, and institutional pressures, systems started adapting around the intervention itself.
Actors adjusted behaviour in ways that protected existing arrangements while still appearing aligned with reform. Political relationships still needed to be managed, informal obligations maintained, risks contained, and fragile networks preserved.
Over time, I increasingly started thinking about this as political adaptation: the ways actors, institutions, and networks adjust behaviour in response to intervention in order to protect, renegotiate, preserve, or reshape underlying incentives, authority relations, dependencies, and political settlements.
What became increasingly visible was not simply resistance to reform, but more subtle forms of adaptation. Actors within a system learned how to formally comply with reform processes while continuing to protect the relationships, incentives, and behavioural patterns the reforms were intended to shift.
And importantly, this was often not because underlying incentives had changed, but precisely because they had not. As a result, reform efforts were frequently absorbed into existing political and institutional arrangements rather than fundamentally reshaping them.
The more I observed this, the more I found myself returning to a different question. What if the challenge facing systems change is no longer simply how to better understand systems, stakeholders, or power itself, but how to collectively interpret how political and institutional realities respond once efforts to change them begin?
Because no single actor sees the system in full.
Different actors experience different risks, incentives, pressures, and forms of visibility. Donors may see progress and formal compliance, while frontline actors experience selective implementation and informal adaptation. Civil society actors may see continuing exclusion, while political actors remain focused on preserving fragile coalitions or managing risk.
Viewed individually, these signals often appear fragmented or even contradictory. I remember sitting in discussions where participants concluded that meaningful change was impossible because surrounding incentive structures appeared too entrenched or politically protected to shift. Yet what increasingly struck me was that these conclusions were often being reached from highly partial views of the system itself.
Actors could observe parts of the system’s response, but they struggled to collectively interpret what those responses revealed about how power, incentives, relationships, and behaviours were adapting across the wider system. And without that broader collective interpretation, reform efforts struggled to recalibrate effectively.
Reforms continued formally, activities progressed, and coordination structures remained in place. Yet efforts to influence change were often quietly absorbed, redirected, or undermined by the very political and institutional dynamics they were attempting to reshape.
This increasingly felt like a major limitation within societal change practice itself. Because static analysis, collaboration alone, or even collective intelligence are often insufficient once systems begin adapting politically in response to intervention.
The issue was no longer simply whether actors understood the system better, but whether they could collectively interpret and adapt as the system itself responded. When signals are interpreted collaboratively, fragmented observations begin revealing broader patterns in how systems are adapting and where openings for leverage, coordination, or politically durable change may actually exist.
And this matters because systems change depends not only on intervention itself, but on the collective ability to recognise what systems reveal through their responses.
From Collective Intelligence to Collaborative Political Intelligence
This is where my own thinking increasingly began to shift. Because the challenge was no longer simply whether actors could collectively understand the system better. The challenge was whether actors within systems could collectively generate, interpret, and respond to political signals emerging as the system itself reacted to intervention.
Because once reform efforts begin interacting with existing systems, political adaptation becomes unavoidable. Viewed separately, system responses often appear fragmented or contradictory. But when interpreted collaboratively, broader patterns begin to emerge. Hidden incentives become more visible, assumptions are challenged, contradictions surface, and the political logic shaping behaviour becomes easier to recognise across the wider system.
This is what increasingly pushed my thinking beyond collective intelligence and collaborative intelligence towards what I now think of as Collaborative Political Intelligence. I increasingly started thinking about this less as a method for achieving consensus and more as an ongoing process through which differently positioned actors collectively interpret and adapt to political responses as systems evolve.
As different groups begin seeing how others experience the system, what risks they are managing, what relationships they are protecting, and where dependencies sit, the conditions for coordination also begin shifting. Not because politics disappears, but because people begin recalibrating what forms of collaboration, experimentation, sequencing, and collective action become politically possible.
This required something different from static political economy analysis, periodic stakeholder mapping, or collaboration alone. It required ongoing processes of experimentation, distributed sensing, collaborative interpretation, and strategic recalibration.
In practice, this meant treating political economy analysis less as a static diagnostic exercise and more as a collaborative infrastructure for generating and interpreting political signals emerging from the system itself.
The goal was no longer simply to analyse the system, but to create conditions where people positioned across the system could collectively probe, observe, interpret, and adapt as political and institutional realities responded.
This is where experimentation became especially important. Small interventions, procedural changes, shifts in coordination, or attempts to alter incentives often revealed how different parts of the system responded under pressure. But what became increasingly significant was not simply whether people complied, resisted, or adapted, but what these responses revealed about the political conditions shaping behaviour across the system.
Different forms of behavioural response began revealing different things about the political conditions shaping the system.
Resistance often signalled where incentives, authority structures, or relationships felt most threatened and where actors were willing to absorb costs to block or delay change.
Partial compliance suggested that adaptation was possible, but constrained by political risk, institutional dependency, or competing obligations.
Performative compliance often indicated situations where actors maintained the appearance of alignment because existing incentive structures still rewarded continuity over transformation.
By contrast, sincere compliance suggested that incentives, expectations, and political conditions were beginning to align in ways that made new forms of behaviour politically viable and sustainable.
Seen this way, behaviour becomes political data about how power, incentives, risk, and adaptation interact across the system. And crucially, the purpose of interpreting these signals was not interpretation alone. It was to support more politically adaptive forms of engagement by helping actors collectively recalibrate strategy in response to how the system itself was evolving.
Through collectively interpreting these responses, richer understandings of the system began to emerge: where resistance was hardening, where flexibility existed, and where openings for leverage, coordination, or politically durable change might actually be possible.
A report may improve understanding. Collaboration may improve coordination. Collective intelligence may help groups make sense of complexity together.
But once systems begin adapting politically in response to intervention, systems change also requires collective capacities to continuously generate signals, interpret behavioural responses, and recalibrate action as the system itself evolves.
Building Infrastructures for Learning Politically?
What increasingly interested me was that Collaborative Political Intelligence was not simply changing how people understood systems. It was changing how systems themselves organised interpretation, learning, adaptation, and coordination over time.
Political economy analysis is often strongest at revealing incentives, relationships, and structural constraints, but weaker at helping actors collectively test how systems respond dynamically as interventions unfold. This is where probes became especially important.
Without experimentation, interpretation risks remaining speculative. But when systems begin generating political signals through experimentation itself, people become far more capable of collectively interpreting how the system is actually behaving rather than how they assume it behaves.
What became increasingly visible through this work was that Collaborative Political Intelligence was not simply changing how people understood systems, but how systems themselves organised interpretation, adaptation, and coordination over time.
Traditional collective intelligence helped groups build shared understanding, but Collaborative Political Intelligence increasingly focused on something else: helping differently positioned actors collectively interpret what behavioural responses revealed about how power, incentives, and political risk were interacting across the system in real time.
The challenge was no longer simply whether people could understand complexity together, but whether they could collectively interpret what behavioural responses revealed about how power, incentives, authority, and political risk were interacting across the system in real time.
The goal was not consensus, but higher-quality political judgment under conditions of uncertainty. One of the ways we increasingly tried to operationalise this in practice was through adapting the Estuarine Framework (Figure 2) developed from the work of Cynefin.
Rather than treating political economy analysis as the production of static findings, the framework became a way of helping people collectively connect fragmented observations and experiences of how the system actually functioned in practice.
Using storytelling and collaborative mapping, participants explored relationships between actors, constraints, energies, directions of change, and behaviours themselves. This last dimension became especially important because behaviours were often where political adaptation became most visible.
Negotiation, avoidance, selective implementation, performative compliance, strategic delay, and informal coordination all revealed how people were adapting around reform pressures in practice.
Viewed individually, these behaviours often appeared isolated or disconnected. But once interpreted collaboratively, broader political and institutional dynamics became easier to recognise across the wider system.
Participants increasingly realised that many problems initially interpreted as implementation failure or lack of capacity were often connected to deeper political dynamics shaping how the system reproduced itself over time. And this also started changing how people interpreted what was politically possible.
Different groups could see different openings, pressures, dependencies, risks, and forms of leverage depending on where they sat within the system. As these fragmented observations became connected, new possibilities for coordination, experimentation, and strategic adaptation also became more politically imaginable.
Transformation does not emerge because disagreement disappears or because everyone aligns around a perfectly shared vision. It emerges because people become more capable of working across uncertainty, fragmentation, and difference while remaining attentive to power, incentives, adaptation, and political risk.
This is also where work emerging from Dark Matter Labs around civic infrastructure, many-to-many systems, and distributed governance becomes especially relevant. Because one of the most important shifts now emerging across systems-change thinking is the recognition that complex societies require infrastructures capable of supporting ongoing interpretation, adaptation, coordination, and strategic evolution over time.
Collaborative Political Intelligence, therefore, increasingly becomes less about facilitation or analysis alone, and more about infrastructure: infrastructures capable of helping systems generate signals, support distributed sensing, enable collaborative interpretation, and adapt politically over time. Because systems rarely reveal themselves through formal analysis alone. They reveal themselves through how they respond under pressure.
From Political Signals to Strategic Adaptation
Collaborative Political Intelligence changes not only how systems are interpreted, but how strategic adaptation becomes possible.
Once systems become capable of collectively interpreting political responses, they also become more capable of making decisions based on how the system is actually behaving rather than how people assume it behaves. Many reform and systems-change efforts still operate through relatively static assumptions about how change will occur. Interventions are designed, implementation pathways established, indicators monitored, and coordination structures created.
But once reform efforts begin interacting with political and institutional realities, the system itself begins responding. Incentives shift unevenly. Authority moves into informal spaces. New dependencies, workarounds, and coordination patterns emerge. Yet interventions often continue operating as though the assumptions guiding them still hold.
Without ongoing collective interpretation, it becomes difficult to distinguish between visible reform activity and deeper patterns of political adaptation unfolding across the system. Collaborative Political Intelligence, therefore, changes decision-making not simply by improving information, but by improving collective judgment about what the system’s responses are actually revealing.
This changes reform and systems-change efforts in several important ways.
First, it changes how success and failure are interpreted. What initially appears as implementation failure may reveal where incentives remain durable or where institutional arrangements are resisting change. Equally, what appears as successful reform alignment may conceal adaptation around the intervention itself.
Second, it changes how leverage is identified. As patterns of adaptation become more visible, new opportunities for coalition-building, experimentation, sequencing, and strategic coordination become easier to recognise across the system.
Third, it changes how uncertainty is navigated. Rather than attempting to reduce uncertainty through rigid planning or linear implementation models, Collaborative Political Intelligence enables strategy to evolve continuously in response to how the system itself is changing over time.
And finally, it changes the conditions for coordination itself. As different groups collectively interpret signals together, they develop richer visibility of one another’s constraints, dependencies, incentives, and political risks. Disagreement does not disappear, but coordination becomes more adaptive because people are responding to a more collectively interpreted understanding of how the system is functioning.
What started becoming possible through these processes was not simply better analysis, but more adaptive forms of political judgement within dynamic systems.
Collaborative Political Intelligence and Political Durability
One of the most important implications of Collaborative Political Intelligence is its relationship to political durability. Many reforms fail not because people lack understanding, but because new behaviours fail to stabilise once pressure weakens, leadership changes, or external oversight disappears.
Recent work emerging from the SOAS-ACE programme on corruption and regulatory landscapes argues that corruption persists not simply because rules are weak, but because behaviours are continuously reproduced through networks of incentives, expectations, and relationships.
From this perspective, systemic change does not become durable because reforms are formally adopted or because enforcement capacity increases. It becomes durable when enough people across the system begin reinforcing new behavioural expectations horizontally through their everyday interactions.
Rather than relying primarily on top-down enforcement, systems begin generating distributed forms of behavioural reinforcement in which people monitor, adapt to, and reinforce one another because maintaining new behaviours increasingly aligns with their own interests, relationships, expectations, or survival strategies.
Under these conditions of horizontal checking, rule-following and coordination become progressively self-reinforcing. This helps explain why many technically sound governance reforms struggle to endure.
New procedures, coordination mechanisms, or accountability structures may be introduced, while the underlying relational landscape remains largely intact. Informal obligations, loyalty networks, survival incentives, and adaptive workarounds continue reproducing the previous equilibrium.
The system absorbs the intervention while preserving the behaviours that stabilise it. What matters, therefore, is not simply whether institutions change, but whether systems begin reproducing different behavioural expectations through distributed relationships over time.
Over time, I increasingly started seeing Collaborative Political Intelligence less as a way of improving analysis alone and more as a way of creating the civic and political conditions through which horizontal checking could emerge.
Through collaborative sensemaking, politically aware experimentation, and distributed interpretation of behavioural signals, people begin developing richer visibility into how power operates, where incentives are shifting, and which behaviours are becoming more or less politically viable. Eventually, this creates the conditions for something more significant than coordination alone.
Expectations become more visible. Deviations become easier to detect. Relationships begin reinforcing behavioural signals across the system itself. New forms of coordination become progressively easier to sustain because they are no longer dependent solely on central enforcement or external pressure.
Seen this way, political durability becomes less about maintaining reform through top-down control and more about whether new behavioural equilibria become self-reinforcing across the wider system.
What might this mean for the future of systems change?
The deeper challenge facing systems change may no longer be simply how we improve analysis, strengthen collaboration, or design more adaptive programmes.
It may be whether societies, institutions, and networks can develop ongoing capacities to collectively interpret political adaptation as it unfolds and recalibrate behaviour, coordination, and strategy accordingly. Because systems change depends not only on understanding complexity, but on the collective ability to recognise how power responds once reform efforts begin interacting with political and institutional realities.
This requires a significant shift in how we think about governance, learning, and institutional change. The challenge ahead is not simply designing better interventions or generating better evidence. It is building infrastructures capable of supporting distributed political sensing, collaborative interpretation, adaptive judgement, experimentation, and strategic coordination under conditions of uncertainty. And importantly, these capacities cannot remain concentrated within external experts, isolated leadership groups, or temporary reform coalitions alone.
They need to become increasingly distributed across systems themselves. This is where many of the most important conversations emerging across systems change, civic infrastructure, adaptive governance, collective intelligence, political economy, and institutional transformation increasingly begin to converge.
Because beneath many of these conversations sits a deeper question: How do societies develop the collective capacities required to navigate political complexity, adaptation, and uncertainty together over time?
This is what I believe Collaborative Political Intelligence begins to describe, not simply a new way of analysing systems, but a politically adaptive approach to supporting social change in environments shaped by uncertainty, contested power, and continuous institutional adaptation.
And perhaps this is where the future challenge increasingly lies.
Not simply helping systems understand themselves better, but helping societies become more capable of collectively interpreting political adaptation, recalibrating strategy, and sustaining politically durable forms of coordination over time.