A crisis forces institutions to pursue several legitimate objectives that temporarily become difficult to reconcile. Preparedness can make a threat invisible, informing people can increase doubt, centralisation can slow action on the ground, protection can provoke escalation, relief can fuel conflict, and returning to normal can restore earlier vulnerabilities. These tensions are not isolated anomalies. They form a recurring structure of decision-making under pressure. This article proposes a typology of seven paradoxes shared by military, civilian and humanitarian crises. It shows that the right response rarely consists of choosing one value over another. It consists of organising an explicit balance between speed and control, coherence and adaptation, access and impartiality, protection and freedom, continuity and transformation.
Crisis as a system of tensions
The classic literature defines crisis through three interrelated properties. An essential value or vital function appears threatened, the time available contracts, and decision-makers have incomplete knowledge [1]. This combination distinguishes a crisis from a mere incident. It also transforms the rationality of action. A decision may be reasonable in terms of an immediate objective and produce the opposite effect in the longer term. A measure may save lives while reducing the trust on which its acceptance depends. A structure may gain coherence while losing the ability to detect what the centre cannot see.
The word paradox is used here in an organisational sense. It denotes a persistent tension between two interdependent requirements, each legitimate and each becoming dangerous when it dominates without limits. It is therefore not always a logical contradiction in the philosophical sense. Rather, the concept helps analyse second-order effects. It requires us to ask not only whether an action addresses the initial problem, but also what new problem it might create.
This approach remains relevant to war, natural disasters, pandemics, industrial accidents and mass population displacement. Institutions, standards and resources differ, but the same difficulty remains. Action is needed before understanding is complete, diverse actors must be coordinated, and legitimacy must be maintained while results are not yet observable [2].
The preparedness paradox
Effective preparedness can destroy the evidence of its own necessity. If an evacuation, strategic reserve or preventive campaign sharply reduces losses, the public may conclude that the threat was exaggerated. Success then creates a shortfall in political visibility. Costs are immediate and measurable, while the damage avoided remains hypothetical. This asymmetry encourages budget cuts after a quiet period and makes the next crisis more costly.
This mechanism resembles Geoffrey Rose’s reasoning on prevention. A measure applied to a population may produce a major collective benefit while offering each individual a small or imperceptible benefit [3]. The preparedness paradox adds a temporal dimension. The better the policy works, the less credible the counterfactual disaster appears. The two concepts must nevertheless be distinguished. For Rose, the paradox concerns the distribution of benefits. In crisis preparedness, it primarily concerns public evidence of success.
The practical consequence is important. An institution should not assess preparedness solely on the number of victims after an event. It should monitor capabilities that are observable before a crisis, such as mobilisation time, stock coverage, communications continuity, the quality of exercises and the ability to reach exposed groups. The Sendai Framework similarly emphasises understanding risk, governance, investment and preparedness [4].
The information paradox
Decision-makers must inform people early enough to enable action, but they speak when the data is least stable. Waiting for certainty delays protection. Communicating too quickly creates the risk of corrections that may be interpreted as incompetence or manipulation. A crisis thus creates a tension between accuracy and speed.
Research on crisis communication shows that uncertainty is inherent in the situation, while operational recommendations still explain imperfectly how to present it [5]. A study conducted in Denmark during the Covid 19 pandemic also shows that transparency does not automatically produce trust. Its effects depend on previously established credibility and on how the public interprets the message [6]. Telling everything is therefore not enough. What is established, what is probable, what remains unknown and what will trigger a revision of the decision must be clearly ranked.
A second trap arises from the myth of collective panic. Authorities may withhold information for fear of irrational reactions. Yet sociological research indicates that flight panic is rare and mutual assistance is common in disasters [7]. By treating the population as a threat, managers may weaken the cooperation they need. The right principle is to communicate structured uncertainty, accompanied by actionable instructions and a date for an update.
The paradox of command and improvisation
A crisis appears to call for a shorter chain of command. A central authority can set priorities, allocate scarce resources and avoid contradictory messages. Yet the same movement can block local information and slow adjustments. Comparative studies do not confirm automatic centralisation. They also observe formal or informal decentralisation, paralysis and avoided decisions [8].
Preparedness therefore cannot consist of writing a scenario for every event. Crises exceed a plan’s assumptions, sometimes damage the intended resources and bring forward actors absent from the organisation chart. Research on improvisation shows that operational creativity and the emergence of new roles contribute to an effective response [9]. Quarantelli specifically includes the ability to combine established organisations and emergent behaviours among his criteria for good management [10].
A useful plan does not replace judgement. It prepares interfaces, responsibilities, delegation thresholds and means of communication. The centre must retain decisions affecting strategy, the law or a critical resource. Teams on the ground must be able to adapt resources when information is local and waiting causes harm. Coordination then becomes an architecture of distributed decisions rather than an accumulation of authorisations.
The security paradox
In a military crisis, a defensive measure may be perceived as preparation for an offensive. Each actor increases its capabilities to feel more secure, the adversary responds, and everyone becomes less secure. Robert Jervis showed that this dilemma is particularly acute when offensive and defensive means are difficult to distinguish and when the advantage appears to lie with whoever strikes first [11]. Mobilisation, deployment or a communications blackout may therefore be rational individually and destabilising together.
The mechanism also exists in civilian crises. A highly coercive measure may reduce an immediate danger but increase resentment, concealment or distrust of public services. This analogy must be used cautiously, because a relationship between a state and its population is not equivalent to one between military powers. The common feature is more limited. Protective action changes other actors’ expectations and can create a feedback loop.
Reducing this paradox requires making intentions observable. Temporary measures must be proportionate, accompanied by public criteria and, where the context allows, designed to distinguish defence from offence. In the military sphere, notification mechanisms and crisis channels reduce misinterpretation. In the civilian sphere, the traceability of decisions and avenues of appeal serve a comparable function.
The emergency powers paradox
An emergency may require unusual powers to protect life, institutions or the continuity of the state. But those powers can weaken the rules they claim to preserve. Concentrating authority accelerates certain decisions while reducing the checks capable of correcting an error or abuse. The risk lies not only in the initial decision. It also lies in the normalisation of measures presented as temporary.
Bonnie Honig describes emergency as a democratic problem that should not be reduced to the opposition between executive power and legal constraints [12]. Pandemic experience nevertheless qualifies the idea of an executive that is necessarily unlimited. A global survey of the first months of the crisis showed that parliaments, judges and territorial authorities often continued to exercise oversight [13]. The paradox therefore does not inevitably lead to democratic erosion. It reveals the quality of institutions and the precision of safeguards.
Three conditions reduce the risk. Emergency powers must have an explicit legal basis, a reviewable duration and independent oversight. The data used must enable evaluation after the crisis. Finally, the exit must be prepared from the moment the emergency regime begins. A measure without an end criterion turns a provisional response into ordinary policy.
The humanitarian paradox
Humanitarian action must reach victims without becoming an instrument of conflict. Neutrality sometimes facilitates negotiations with all parties, but it may be perceived as moral silence. Public denunciation can defend victims while closing the access that made it possible to help them. The tension is therefore between two forms of protection: operational presence and speaking out.
Aid itself can produce unintended effects. Resources distributed in a controlled territory may be taxed, diverted or incorporated into the wartime economy. Suspending aid to avoid these effects nevertheless exposes civilians to immediate harm. Pierre Perrin emphasises that aid’s influence on a conflict must be assessed without automatically concluding that it prolongs war [14]. The real choice often concerns imperfect aid, its control mechanisms and the human cost of its absence.
Localisation adds another tension. When international organisations lose access, local actors become indispensable and may gain autonomy. Yet they bear a greater share of risk without necessarily receiving corresponding funding or power. Barter and Sumlut call this combination of empowerment and marginalisation the conflict paradox [15]. A coherent response requires transferring resources, authority, information and security guarantees together.
The resilience paradox
Resilience often means the ability to absorb a shock and restore acceptable functioning. But a rapid return to normal may restore the causes of vulnerability. A city may rebuild in the same place, an organisation may restore a network without correcting its critical dependency, and a population may endure the crisis at the cost of invisible exhaustion. Continuity is therefore not always evidence of lasting adaptation.
In psychology, research shows that resilient trajectories are common after a disaster, but that characteristics associated with resilience still poorly predict individuals’ future trajectories. Bonanno and colleagues refer to this as the resilience paradox and propose focusing analysis on flexibility in adaptation [16]. This finding cautions against hasty assessments. A person or community functioning today may need support tomorrow, and vice versa.
At the collective level, resilience becomes problematic when it shifts responsibility onto those exposed. Asking residents to withstand adversity better does not replace reducing exposure, protecting infrastructure or addressing inequalities. The relevant criterion is not simply the speed of return. We must examine who recovers, under what conditions and with what reduction in future risk.
A comparative analytical framework
The seven paradoxes can be summarised by their main mechanism and the management rule that limits their effects. This framework does not provide a universal recipe. It helps identify the indirect costs of a decision before they become a new crisis.
Paradox
Mechanism
Major risk
Response principle
Preparedness
Success erases the disaster avoided
Underinvestment after a quiet period
Measure capabilities and losses avoided
Information
It is necessary to speak before knowing everything
Delay, confusion or loss of trust
Structure uncertainty and announce a review
Command
Central coherence can hinder local adaptation
Paralysis or inconsistent responses
Define delegation thresholds
Security
Protecting one actor can threaten another
Escalation and misinterpretation
Make intentions and limits observable
Exception
Protecting order can weaken its rules
Temporary power becomes permanent
Set a duration, oversight and an exit criterion
Humanitarian action
Gaining access can require silence and compromise
Diversion, exclusion or loss of access
Monitor effects and share power locally
Resilience
Returning quickly can restore vulnerability
Normalisation of weaknesses
Assess recovery and risk reduction
Governing paradoxes
The first rule is to acknowledge the tension rather than conceal it. An organisation that presents every decision as perfectly coherent loses the ability to explain its trade-offs. It also becomes more vulnerable when unintended effects appear. Naming the paradox makes it possible to specify the benefit sought, the cost accepted, the people bearing that cost and the point at which the decision will be reviewed.
The second rule is to use reversible arrangements. A decision made with little information must be able to change without being interpreted as a moral failure. Review clauses, measurable thresholds, limited experiments and feedback from the ground reduce the cost of correction. Reversibility does not mean indecision. It allows a firm decision with explicit duration and conditions.
The third rule is to retain multiple sources of knowledge. The centre has an overall view, field teams perceive local disruptions, residents understand constraints on use, and independent organisations can document effects the authorities do not see. An effective crisis team organises the comparison of this information without erasing its origins.
Finally, evaluation must consider both direct and indirect consequences. Lives saved, service continuity and response speed remain essential. They must be complemented by trust, equitable access, infringements of rights, the distribution of risks and vulnerability after the return to normal. A policy may succeed on the first horizon and fail on the second.
Conclusion
Crises reveal a fundamental feature of collective action. The most necessary solutions can create the conditions for their own contestation or future failure. Preparedness conceals the danger avoided. Transparency exposes uncertainty. Command reduces disorder but can stifle initiative. Security can fuel escalation. Exceptional powers can weaken the law. Aid can unintentionally support parties to a conflict. Resilience can restore a vulnerable normality.
Good management makes these tensions visible and open to review. Risk remains difficult to predict because psychological reactions vary with the perception of danger, fear, denial, perceived control, trust and group dynamics [7] [16]. Decisions must evolve with their effects on exposure to danger.
References
[1] Rosenthal, Uriel, Michael T Charles and Paul t Hart, 1989. The World of Crises and Crisis Management. In Coping with Crises. The Management of Disasters Riots and Terrorism. Charles C Thomas.
[2] Boin, Arjen, Paul t Hart, Eric Stern and Bengt Sundelius, 2005. The Politics of Crisis Management. Public Leadership under Pressure. Cambridge University Press.
[3] Rose, Geoffrey, 1985. Sick Individuals and Sick Populations. International Journal of Epidemiology, volume 14, issue 1, pages 32 to 38.
[4] United Nations Office for Disaster Risk Reduction, 2015. Sendai Framework for Disaster Risk Reduction 2015 to 2030.
[5] Liu, Brooke Fisher, Laura Bartz and Nathan Duke, 2016. Communicating Crisis Uncertainty. A Review of the Knowledge Gaps. Public Relations Review, volume 42, issue 3, pages 479 to 487.
[6] Gamerdinger, Alexander, Sine N Just and Prins Marcus Valiant Lantz, 2023. Healthy Transparency. Dynamic Interrelations Between Credibility Transparency and Trust in the Context of Danish Public Authorities Covid 19 Communication. Social Sciences and Humanities Open, volume 8, issue 1, article 100688.
[7] Quarantelli, Enrico L, 2001. The Sociology of Panic. Disaster Research Center, University of Delaware.
[8] Hart, Paul t, Uriel Rosenthal and Alexander Kouzmin, 1993. Crisis Decision Making. The Centralization Thesis Revisited. Administration and Society, volume 25, issue 1.
[9] Webb, Gary R and Francois Regis Chevreau, 2006. Planning to Improvise. The Importance of Creativity and Flexibility in Crisis Response. International Journal of Emergency Management, volume 3, issue 1, pages 66 to 72.
[10] Quarantelli, Enrico L, 1997. Ten Criteria for Evaluating the Management of Community Disasters. Disasters, volume 21, issue 1, pages 39 to 56.
[11] Jervis, Robert, 1978. Cooperation Under the Security Dilemma. World Politics, volume 30, issue 2, pages 167 to 214.
[12] Honig, Bonnie, 2009. Emergency Politics. Paradox Law Democracy. Princeton University Press.
[13] Ginsburg, Tom and Mila Versteeg, 2021. The Bound Executive. Emergency Powers During the Pandemic. International Journal of Constitutional Law, volume 19, issue 5, pages 1498 to 1535.
[14] Perrin, Pierre, 1998. The Impact of Humanitarian Aid on Conflict Development. International Review of the Red Cross, issue 323, pages 319 to 333.
[15] Barter, Dustin and Gun Mai Sumlut, 2023. The Conflict Paradox. Humanitarian Access Localisation and Disempowerment in Myanmar Somalia and Somaliland. Disasters, volume 47, issue 4, pages 849 to 869.
[16] Bonanno, George A, Shuquan Chen, Rohini Bagrodia and Isaac R Galatzer Levy, 2024. Resilience and Disaster. Flexible Adaptation in the Face of Uncertain Threat. Annual Review of Psychology, volume 75, pages 573 to 599.
