Compromise, not consensus
In section on trade-offs[GK1] , we argued that the real trade-offs in ILM are in the power domain: any intervention that shifts a system’s heading creates winners and losers. This raises a practical question: how are these trade-offs navigated? The default answer in much landscape and natural resource management literature is consensus. We disagree.
Consensus sounds democratic. It implies that all parties have been heard, that differences have been reconciled, and that everyone has agreed. In practice, however, consensus in multi-stakeholder settings tends to produce lowest-common-denominator outcomes – proposals sanded down until they threaten no one and change nothing. At the opening plenary of COP27 in 2022, Bangladesh’s representative made exactly this argument: that consensus-based decision-making was producing outcomes too weak to address the crisis they were designed to confront. The pattern is structural, not incidental. When consensus is the decision rule, any actor who benefits from the current system heading holds an effective veto. The most powerful parties – those who have successfully externalised costs, captured regulatory processes, or accumulated disproportionate benefits – need only withhold agreement to prevent change. Consensus thus tends to protect precisely the arrangements that an ILM intervention is trying to shift.
The political theorist Chantal Mouffe[1] has argued that conflict and antagonism are ineradicable features of political life, and that attempts to ground collective decision-making in rational consensus are both illusory and dangerous. They are illusory because they assume shared starting principles that, in pluralistic societies with asymmetric power, do not exist. They are dangerous because the fiction of consensus obscures the power relations that actually produced the agreement. In Mouffe’s terms, the question for democratic politics is not how to eliminate power but how to constitute forms of power that are compatible with democratic values. We think this applies directly to ILM: the question is not how to find a position everyone agrees with, but how to construct arrangements that honestly distribute power, benefits and costs – and that enough actors can live with.
Deliberation, not dialogue
If consensus is not the goal, what kind of process gets us to compromise? We argue for deliberation, and we distinguish it from dialogue – a term more commonly used in landscape and stakeholder engagement practice. This distinction matters.
Dialogue, at its best, builds understanding. It asks participants to listen, to surface perspectives, to develop empathy across difference. We do not dismiss this – it is valuable groundwork. But dialogue, by itself, can leave power relations undisturbed. Participants understand each other better; they may even like each other more. But nobody has been asked to give anything up, and the structural conditions that produce the problem remain intact. In the worst case, dialogue becomes a performance of inclusion that substitutes for the harder work of redistribution.
Deliberation is different. Following Dryzek,[2] we understand deliberation as a process in which participants are expected to change their positions in light of reasoned argument, new information, and the claims of others. This is not a conversation aimed at mutual understanding; it is a structured confrontation with trade-offs. Deliberation invites conflict to be surfaced – not smoothed over, not deferred, but engaged with directly. Van de Kerkhof[3] makes this distinction explicit: consensus-building constrains what can be discussed and tends toward lowest-common-denominator outcomes, while deliberative design helps participants critically assess arguments and deal with stakeholder conflict at an early stage.
Elsewhere in this Playbook, we call for ILM practitioners to assume that conflict is present: competing interests, structural inequalities, externalised costs. Deliberation is the process that brings this conflict into the room. It requires skilled facilitation, mediation, and negotiation – capacities that we argue are essential components of any well-functioning ILM team. Effective deliberation demands “explicit recognition of politics and power” in process design, and the larger goal is to “turn potential conflict into constructive engagement.”[4]
The logic of compromise
If deliberation surfaces conflict honestly, compromise is the mechanism through which it is navigated. Compromise is not consensus. It does not pretend that everyone got what they wanted. It acknowledges that actors have genuinely conflicting interests, and it requires real concessions – real costs – from all parties.
There is a common intuition that compromise means everyone loses: each party gives up part of its position, and the result is less than anyone hoped for. This is wrong, or at least it is incomplete. Compromise creates value that does not exist under the status quo. Consider: if two parties each concede 30% of their starting position, the result is not 70% of what was. The result, if the compromise is well designed, is 140% – because the agreement unlocks possibilities that neither party could access alone. A community that concedes some land-use flexibility gains tenure security; an investor that accepts tighter environmental conditions gains social licence and reduced political risk; a government that shares regulatory authority gains compliance and legitimacy. These are not zero-sum trades. They are the creation of new value through the honest confrontation of trade-offs.
This is what distinguishes compromise from both consensus and capitulation. Consensus sands proposals down until they threaten no one; capitulation concentrates losses on the weakest party. Compromise asks all parties to bear real costs in exchange for real gains – and the resulting arrangement, precisely because it cost something, has a durability that consensus outcomes typically lack. Actors comply not because they were persuaded, but because they struck a deal whose terms they helped to shape and whose benefits they stand to lose if they defect.
This logic is, in effect, a restatement of the ‘prisoner’s dilemma’ – and of Ostrom’s[5] solution to it. In the prisoner’s dilemma, two suspects are arrested for a crime and held in separate cells, unable to communicate. Each faces the same offer from the prosecutor:
- If both stay silent (cooperate with each other), there is not enough evidence, so they each get a light sentence (say, 1 year).
- If one confesses and implicates the other (defects) while the other stays silent, the confessor goes free (0 years) and the silent partner gets a heavy sentence (say, 10 years).
- If both confess (both defect), they each get a medium sentence (say, 5 years).
Each prisoner reasons: that whatever the other does, confessing gives me a better personal outcome (0 vs 1 year if the other is silent; 5 vs 10 years if the other confesses). So defection (confessing) is the individually rational choice for both. The result is that both confess and each gets 5 years—worse for both than the 1 year each they’d have received by both staying silent. The ‘dilemma’ is that individually rational, payoff‑maximising behaviour leads to a collectively worse outcome than mutual cooperation.
Mutual defection is, therefore, individually rational but collectively disastrous: both parties get a worse outcome than if they had cooperated. The dominant strategy is to defect, because the individual gain from betrayal is total and the loss to the other party is absolute.
Ostrom’s insight, developed across decades of empirical work on common-pool resources, is that real actors are not the purely self-interested maximisers of standard game theory. They are boundedly self-seeking: they will accept individual costs when the collective gains are visible, credible, and when institutional arrangements make defection more costly than compliance.
In our terms, the 30% concession is the choice to cooperate rather than defect. The 40% surplus is the payoff that exists only under mutual cooperation – it is not available to any party acting alone. But here is the critical point for ILM practice: in the abstract dilemma, the payoffs are given. In a real landscape negotiation, they have to be constructed – made legible, concrete, and specific enough that a boundedly self-seeking actor can weigh them against what they are being asked to give up. The ILM practitioner’s job in preparing for deliberation is therefore not just facilitation; it is framing the collective gain. You come to the table having done the work to show what mutual cooperation actually produces: the tenure security, the ecosystem services, the market access, the reduced political risk. Without that concrete articulation, you are asking people to cooperate on faith – which is precisely the situation the prisoner’s dilemma says will fail.
The compromise arithmetic also maps onto the Playbook’s power framework. The 30% loss is a redistributive move – power as redistribution, in which an actor gives up position. The 40% gain is a capacity move – power as capacity, in which the collective now has resources, legitimacy, or coordination that no individual actor could generate alone. Compromise, then, is the mechanism that converts redistributive losses into capacity gains. This is a stronger formulation than merely ‘everyone gives a bit and gets a bit’. It explains why well-designed compromise creates durable arrangements: the collective capacity it generates is real, and actors have a stake in maintaining it.
Implications for practice
This argument has direct consequences for how ILM interventions are designed and facilitated. First, it means that the pursuit of consensus should be treated with caution. Where all parties ‘agree’, the ILM practitioner should ask: what was given up to reach this agreement? If the answer is ‘nothing’, the agreement probably changes nothing. Second, it means that skilled mediators and negotiators are not optional extras in an ILM team – they are core capacities. Deliberation is demanding work. It requires people who can hold space for conflict without letting it become destructive, who can identify the real interests beneath stated positions, and who can help parties see the value that compromise might create. Third, it connects directly to our discussion on trade-offs: if the real trade-offs in ILM are in the power domain, then the process for navigating them must be one that confronts power honestly. Deliberation does this. Dialogue, on its own, does not.
Fourth, it explains why ILM is necessarily iterative. Ostrom’s work on repeated games shows that cooperation becomes self-reinforcing once actors can see it working: each successful round of cooperation provides the evidence that the collective gains are real, making the next round easier. No single actor will give up everything at once, but they may accept incremental concessions if each step delivers visible collective benefits. In our terms, each step around the Playbook’s dartboard[GK2] is a bounded compromise – a small redistribution of concentrated power in exchange for a demonstrated gain in collective capacity. Early rounds require the hardest work in framing the collective gain, because there is no track record. Later rounds can point to results. This is also why Ostrom’s design principles emphasise monitoring and graduated sanctions: they make cooperation visible and defection costly, which is what sustains the process across iterations. The ILM process, then, is not a single grand bargain but a sequence of bounded, self-seeking negotiations in which each compromise builds the institutional foundation for the next.
Finally, this framing helps explain why so many ILM interventions fail to produce lasting change. If the process was designed around consensus, it likely produced an outcome that the most powerful actors could live with – which means it did not shift the system’s heading. If it was designed around dialogue, it likely improved relationships without redistributing power. Only a process designed around deliberation and compromise – one that surfaces conflict, confronts trade-offs, and requires real concessions – has a chance of producing the kind of shift that ILM is supposed to deliver.
References and further reading
Dore, J., Robinson, J. and Smith, M. (eds) 2010. Negotiate: reaching agreements over water. Gland: IUCN, https://portals.iucn.org/library/sites/library/files/documents/2010-006.pdf.
Dryzek, J. S. 2000. Deliberative democracy and beyond: liberals, critics, contestations. Oxford: Oxford University Press.
Mouffe, C. 2000. Deliberative democracy or agonistic pluralism? Political Science Series 72. Vienna: Institute for Advanced Studies, https://nbn-resolving.org/urn:nbn:de:0168-ssoar-246548.
Mouffe, C. 2005. On the political. London: Routledge.
Ostrom, E. 1990. Governing the Commons: the evolution of institutions for collective action. Cambridge: Cambridge University Press.
Van de Kerkhof, M. 2006. Making a difference: on the constraints of consensus building and the relevance of deliberation in stakeholder dialogues. Policy Sciences 39(2006): 279–299, https://doi.org/10.1007/s11077-006-9024-5.
[1] Mouffe, 2000, 2005.
[2] Dryzek, 2000.
[3] Van de Kerkhof, 2006.
[4] Dore et al., 2010.
[5] Ostrom, 1990.
[GK1]Link
[GK2]Link to ‘integration’ – and, if possible, the graphic.
