Thousands of people are dead and thousands more remain missing after catastrophic flooding devastated the eastern Libyan city of Derna early on Monday. Storm Daniel brought intense rainfall across the region, but the most destructive moment came when two dams above the city failed and released an enormous volume of water through the valley running into Derna. Entire neighbourhoods were swept towards the Mediterranean, buildings collapsed and roads and bridges disappeared. The final human cost remains unclear because different authorities are reporting very different numbers while search teams continue recovering bodies from rubble and the sea.
Calling this simply a natural disaster would be incomplete. The storm was natural in the sense that no government created the weather system that struck eastern Libya. But the two dams were human infrastructure, their maintenance was an institutional responsibility and Derna’s ability to evacuate residents or respond afterwards depended on government capacity. The relevant distinction is therefore between hazard and vulnerability. Governments cannot eliminate every earthquake, flood or storm, but they substantially influence how many people die when those hazards arrive.
Evidence emerging after the disaster makes that distinction particularly important. Experts had warned that the dams above Derna were vulnerable and required maintenance. A Libyan state audit reported that millions of dollars had been allocated for maintenance in 2012 and 2013, yet the work was not completed, while researchers had continued warning about the potentially catastrophic consequences of a major flood. The disaster was therefore not simply an encounter with an unknowable risk. It involved infrastructure whose vulnerability had been recognised before the event that exposed it.
This changes the accountability question. Governments cannot reasonably be blamed for failing to predict the precise day on which an extreme storm will arrive. They can be judged for what they do with information about foreseeable risks. A dam identified as dangerous but left inadequately maintained represents a political decision even when no minister explicitly decides to accept a future catastrophe. Money is allocated elsewhere, contracts remain unfinished, warnings disappear through bureaucracies and responsibility becomes sufficiently fragmented that nobody feels personally responsible for the accumulating risk.
This is how institutional neglect becomes invisible during normal periods. A dam can remain standing for years despite inadequate maintenance. Every year without failure makes the problem appear less urgent. Officials face immediate demands for salaries, roads, electricity and security, while repairing infrastructure that appears to function produces little visible political reward. The benefit of maintenance is a catastrophe that does not occur, which means successful prevention often looks like money spent on nothing.
The incentives therefore favour postponement. If repairing a dam costs millions today while the expected failure lies somewhere in an uncertain future, governments facing short political horizons can rationally allocate resources towards projects producing immediate benefits. Each individual postponement may appear manageable. The accumulated postponements eventually create a system whose apparent stability depends increasingly on good luck.
This is the same problem Turkey faced after February’s earthquakes and Greece after the Tempi railway disaster. Physical infrastructure contains institutional history. A collapsed building records previous decisions about construction standards. A railway collision records previous decisions about signalling and maintenance. A failed dam records years of inspection, procurement and administrative choices. Catastrophe compresses those decisions into one visible moment.
Libya adds another dimension because the state responsible for those decisions has itself been fragmented for more than a decade. Since the overthrow of Muammar Gaddafi in 2011, the country has experienced civil conflict, competing governments, powerful militias and repeated foreign intervention. Since 2014, political authority has been divided substantially between rival centres in the east and west, with different armed organisations and political institutions controlling territory.
This means Libya possesses legal sovereignty without fully consolidated practical sovereignty. International maps show one Libyan state. Citizens experience several overlapping centres of political and military authority. The internationally recognised institutions controlling national resources do not necessarily exercise effective authority over every municipality, while local authorities and armed groups can possess practical power without clear national accountability.
That fragmentation makes ordinary governance difficult long before catastrophe occurs. Who is responsible for maintaining the dam? Which ministry controls the budget? Which government has authority? Which institution monitors the contract? Who investigates if nothing happens? If several organisations possess partial authority, each can explain failure through another institution’s responsibility.
This is accountability dilution. When ten institutions share responsibility imperfectly, the total amount of responsibility can paradoxically feel smaller rather than larger because every actor possesses someone else to blame. Fragmented authority therefore creates not only coordination problems but informational problems for citizens trying to determine who failed them.
The flood demonstrates why state capacity should not be understood primarily through the state’s ability to deploy armed force. Libya contains many armed organisations. What it lacks is enough unified administrative capacity to perform some of the less dramatic functions that make societies safe: maintaining dams, enforcing construction standards, coordinating emergency warnings, managing evacuation and ensuring public money produces the infrastructure for which it was allocated.
A militia can seize a ministry.
It cannot automatically make the ministry work.
This distinction between coercive capacity and administrative capacity is central to state-building. Governments need enough coercive power to enforce one legal order, but the purpose of consolidating that authority is ultimately to provide public goods citizens cannot provide individually. Security matters because it creates the environment in which engineers can maintain dams, doctors can operate hospitals and businesses can invest.
When political systems become dominated by competition over coercive power, those quieter functions deteriorate.
Libya’s oil wealth makes this deterioration particularly striking. The country possesses Africa’s largest proven oil reserves and a relatively small population, meaning resource scarcity alone cannot explain why essential infrastructure has been neglected. Libya’s problem is partly about institutional conversion: transforming national wealth into reliable public goods.
Natural resources create revenue.
They do not create capable government automatically.
In fact, resource wealth can sometimes weaken accountability because governments receiving substantial oil revenue depend less heavily on broad taxation. Tax systems force states to interact repeatedly with citizens and businesses. Governments collect money and citizens demand explanations for how it is spent. Oil revenue enters the state through a much narrower channel, making control of government institutions extraordinarily valuable while weakening some of the fiscal bargaining relationships associated with broad taxation.
This is one mechanism behind the resource curse. The problem is not that oil itself produces corruption or conflict. It is that concentrated rents can make capturing the state more profitable while reducing the need to build the broad administrative institutions required to tax a diversified economy.
Libya’s political fragmentation then compounds the problem because rival groups compete not only for political authority but for access to the institutions controlling oil revenue. The prize for controlling the state becomes enormous even while the capacity of that state to deliver ordinary services deteriorates.
The result is a country that can remain wealthy in national accounting while citizens experience institutional poverty.
Derna demonstrates what institutional poverty looks like physically.
The city had dams.
The state had oil revenue.
Experts possessed engineering knowledge.
Money had reportedly been allocated for maintenance.
Yet the chain connecting those resources to a repaired dam failed somewhere between knowledge, financing, administration and implementation.
Development therefore cannot be measured simply by whether resources exist. It depends on whether institutions can complete chains of execution. A government identifies a problem, allocates money, awards a contract, monitors the work, inspects the result and maintains the asset afterwards. Every stage needs to function.
A state can fail without any one stage collapsing completely. Small failures compound.
A procurement delay here.
A corrupt payment there.
A ministry dispute.
A contractor disappears.
A government changes.
A war begins.
The maintenance is postponed.
Eventually the dam remains standing only because the storm has not arrived yet.
This is why corruption should be understood as more than money stolen from budgets. The direct fiscal loss may be relatively small compared with the opportunity cost of whatever the money was supposed to accomplish. If a million dollars disappears from a maintenance budget, society loses not only the million dollars but the safety the maintenance would have purchased.
The social return on honest administration can therefore be enormous.
Disaster preparedness has the same characteristic. Authorities reportedly issued inconsistent messages as Storm Daniel approached, with some residents describing instructions that did not produce an effective evacuation from the most vulnerable areas. Even a perfectly maintained dam might face danger during an extreme event, which makes warning systems another independent layer of protection.
Resilient systems rely on redundancy.
The dam should hold.
If the dam might not hold, monitoring should detect danger.
If danger rises, warnings should reach residents.
If warnings arrive, evacuation routes should function.
If flooding still occurs, emergency services should respond.
No single layer needs to be perfect if enough independent layers exist.
Catastrophe becomes more likely when several fail simultaneously.
Libya’s political fragmentation weakens exactly this redundancy because coordinating agencies across competing authorities is difficult. The flood has temporarily produced some cooperation between eastern and western institutions, and aid has arrived from abroad, but emergency coordination has reportedly remained difficult.
This reveals something important about national unity. Political unity is not valuable merely because national symbols feel emotionally satisfying. Unified institutions reduce transaction costs during collective emergencies. A rescue service should not need to determine which political faction controls a neighbourhood before operating there.
Administrative integration is therefore a public good.
The international response also demonstrates the limits of sovereignty during catastrophe. Libya needs foreign search teams, medical assistance and humanitarian supplies. Accepting that help does not diminish national independence. Sovereignty should mean possessing political authority over national decisions, not pretending a state can provide every capability alone.
The deeper sovereignty problem is when domestic fragmentation prevents the government from coordinating assistance effectively.
International aid cannot substitute permanently for national institutions. Foreign rescue teams can search rubble this week. They will not inspect Libya’s dams for the next thirty years.
That is why humanitarian response and institutional reconstruction need to be distinguished.
The first keeps people alive after catastrophe.
The second reduces the probability that the next hazard becomes another catastrophe.
Climate change makes the second increasingly important. Scientists are still examining precisely how warming affected Storm Daniel, and attributing one event entirely to climate change would be premature. The broader physical mechanism is nevertheless clear: warmer air and seas can intensify aspects of extreme rainfall, increasing pressure on infrastructure designed using historical weather patterns.
This means governments cannot rely indefinitely on the assumption that tomorrow’s hazards will resemble yesterday’s.
Infrastructure designed for twentieth-century conditions may face twenty-first-century extremes.
Climate adaptation therefore becomes another state-capacity test. Rich countries can finance stronger drainage systems, dams and coastal protection more easily. Fragile states may face some of the largest climate vulnerabilities while possessing the least institutional capacity to prepare.
That creates a global inequality problem. The countries least capable of financing resilience can experience the largest proportional losses when disasters occur, pushing development backwards and reducing the resources available for the next round of adaptation.
Fragility and climate vulnerability can therefore reinforce one another.
Libya is unusual because it possesses substantial financial resources through oil. Its principal constraint is not simply poverty but political fragmentation and institutional weakness. That makes Derna particularly tragic: some of the vulnerability appears to have been preventable with resources the country already possessed.
The eventual investigation will need to determine exactly why maintenance failed, which institutions knew about the risks and whether corruption or negligence contributed. Those conclusions should not be assumed before evidence is established. What can already be said is that repeated warnings existed and that the infrastructure had suffered from long periods of inadequate maintenance.
Accountability will be difficult because the same fragmentation contributing to the disaster also affects the institutions responsible for investigating it. Citizens may reasonably ask whether authorities can investigate failures implicating powerful officials or political networks.
This is why independent institutions matter most when government itself may be responsible.
A prosecutor dependent entirely on political leadership cannot easily investigate political leadership. An audit institution without enforcement power can document missing maintenance repeatedly without ensuring anything changes.
Accountability requires both information and consequences.
Libya has often possessed fragments of the first without enough of the second.
The political temptation after Derna will be finding one official, contractor or mayor who can absorb responsibility. Individual accountability may be justified if evidence establishes negligence. But disasters produced over decades rarely have only one author.
The more important question is whether the investigation changes the system that allowed warnings to remain warnings.
Otherwise prosecution becomes another form of symbolic repair.
A person goes to prison.
The next dam remains neglected.
The broader lesson from Derna is therefore not simply that governments should maintain infrastructure. It is that state failure often appears mundane before it becomes catastrophic. We imagine failed states through civil war, militias and collapsing governments because those events are dramatic. Administrative failure is quieter.
A maintenance contract not completed.
An inspection postponed.
A warning not transmitted.
A budget that disappears.
Nothing happens immediately.
That silence makes neglect politically sustainable.
Then the weather changes.
Within hours, institutional weakness accumulated over years becomes visible in destroyed buildings and missing families.
The storm did not create Libya’s political fragmentation, corruption or neglected infrastructure.
It exposed them.
That distinction matters because rebuilding Derna cannot mean simply replacing concrete. Libya can construct new dams and still remain vulnerable if the institutions responsible for maintaining them remain divided and unaccountable.
Physical reconstruction without institutional reconstruction restores the asset while preserving the risk.
Libya therefore needs something much harder than repairing what the water destroyed. It needs a state capable of maintaining what it rebuilds.
Maps have said Libya is one country throughout twelve years of division.
Derna demonstrates why sovereignty on a map is not enough.
A functioning state is ultimately measured by whether the institutions behind the border can keep the dam standing when the storm arrives.
