Haiti’s political and security crisis has deteriorated dramatically this week. Armed gangs have launched coordinated attacks across Port-au-Prince, targeting police stations, prisons and strategic infrastructure while Prime Minister Ariel Henry remains outside the country. Thousands of prisoners escaped after armed groups attacked Haiti’s two largest prisons, commercial flights have been disrupted, the government has declared a state of emergency and gangs have attempted to gain control over routes and infrastructure connecting the capital with the outside world. Haiti still possesses a government recognised internationally as representing the country. Yet the widening gap between that government’s legal authority and its ability to exercise authority physically raises one of the most fundamental questions in political science: what exactly makes a state a state?
The classic answer comes from Max Weber, who described the modern state through its claim to the legitimate monopoly over the use of physical force within a territory. The word legitimate matters as much as monopoly. Criminal organisations can possess weapons. Militias can control neighbourhoods. Companies can employ security personnel. What distinguishes the state is that its coercive institutions operate under a legal order claiming authority over everyone within the territory. Police can arrest someone because law authorises them to do so; a gang kidnapping someone possesses force without equivalent legitimate authority. State failure begins when the gap between those two forms of power becomes so large that legal authority stops determining what happens physically.
Haiti increasingly demonstrates that gap. Armed groups are estimated to control substantial parts of Port-au-Prince and the roads connecting different areas of the capital. The police remain present and continue fighting, but they are under-resourced relative to the scale of the challenge. The Haitian military was disbanded in the 1990s and only later re-established on a small scale. Government ministries and political institutions therefore exist within a security environment where non-state organisations can sometimes determine who moves, which businesses operate and whether roads remain accessible.
This is fragmented sovereignty.
On an international map, Haiti remains one colour.
On the ground, authority becomes geographically discontinuous.
A government may control one district.
A gang controls another.
A road between them becomes contested.
The border remains internationally Haitian throughout.
The lived experience of sovereignty changes every few kilometres.
This distinction between formal and effective sovereignty has appeared repeatedly in these posts, from Libya to the Iran-Pakistan border. Formal sovereignty is a legal category. Effective sovereignty is an institutional capability. The first tells us which government represents territory internationally. The second tells us whether that government can actually enforce law, collect taxes and provide public goods inside it.
Stable states possess enough overlap between the two that citizens rarely notice the distinction.
State failure makes it visible.
Haiti’s gangs are particularly powerful because they do not need to replace the government completely to weaken the state. Controlling strategic nodes can produce enormous leverage without controlling every neighbourhood. Ports, roads, fuel terminals, airports and commercial districts matter disproportionately because the rest of the economy depends on them.
This is the same chokepoint logic the Houthis demonstrated in the Red Sea.
Power depends not only on how much territory you control.
It depends on what the territory connects.
A gang controlling one major road can influence food distribution across several neighbourhoods. Control over access to a fuel terminal can affect electricity and transport. Threatening the airport can isolate the political leadership from the country.
Small areas can therefore produce system-wide effects.
This is network power inside a city.
The attacks on prisons demonstrate another mechanism. Thousands of prisoners escaping does not simply increase the number of potentially dangerous individuals outside custody. It communicates something about state credibility.
Prisons are physical demonstrations of the state’s coercive capacity.
The government says certain people have been lawfully deprived of freedom.
Walls, guards and police make that decision real.
When armed organisations can attack those institutions and release prisoners, they demonstrate publicly that state decisions can be reversed through private force.
That is a symbolic loss of sovereignty as well as a security loss.
State authority depends partly on expectations.
Citizens obey law not because a police officer stands beside every person constantly but because they expect institutions can enforce rules when necessary.
Businesses pay taxes because they expect the state to possess durable authority.
Criminal organisations calculate risk according to the probability of arrest and punishment.
Once enough actors believe enforcement capacity has collapsed, behaviour changes.
More people avoid taxes.
Businesses hire private security.
Citizens stop reporting crimes.
Police receive less intelligence.
Criminal organisations become bolder.
State weakness becomes self-reinforcing.
This is an institutional bank run.
Banks remain solvent partly because depositors believe other depositors will not all withdraw simultaneously.
States remain authoritative partly because citizens believe other citizens will continue treating them as authoritative.
Once that expectation collapses, rebuilding it becomes much harder than simply deploying another police unit.
This is why gang control creates economic consequences far beyond theft and violence. Businesses operating in gang-controlled areas face extortion, unpredictable transport and security costs. Workers cannot reach jobs reliably. Schools close. Healthcare becomes harder to access. Investment disappears.
Violence therefore functions like an enormous transaction tax.
Every economic interaction becomes more expensive because participants need to purchase security, change routes or accept greater risk.
GDP falls not simply because things are destroyed but because transactions that would otherwise have happened no longer occur.
A shopkeeper closes before dark.
A truck refuses a route.
A doctor leaves the country.
A company postpones investment.
Each decision is individually rational.
Collectively they shrink the economy and therefore shrink the tax base available to rebuild security.
Another feedback loop emerges.
Weak security reduces economic activity.
Lower economic activity reduces government revenue.
Lower revenue weakens security capacity.
Weaker security reduces economic activity further.
Breaking that equilibrium requires resources from outside the loop.
This is one reason Haiti has repeatedly depended on international assistance.
But foreign intervention has an extraordinarily difficult history there.
Haiti has experienced repeated foreign occupations, UN missions and international involvement in domestic politics. The United States occupied Haiti from 1915 to 1934, while a UN peacekeeping mission operated from 2004 until 2017. International interventions have sometimes restored immediate security while also generating resentment, abuses and dependency.
The previous UN mission became particularly controversial because UN peacekeepers introduced cholera into Haiti in 2010, contributing to an epidemic that killed thousands of people. Sexual abuse allegations involving peacekeepers further damaged trust.
That history matters because Haiti now needs external security assistance while possessing strong historical reasons to distrust external security assistance.
This is the intervention paradox.
The weaker the Haitian state becomes, the stronger the argument for outside assistance.
The more outside actors substitute for Haitian institutions, the greater the risk that Haitian institutions remain weak.
Emergency intervention can therefore solve the immediate symptom while reproducing long-term dependency.
The proposed Kenyan-led multinational security mission attempts to navigate this by placing a non-Western country at the centre rather than creating another traditional American or UN occupation. Kenya has offered to deploy approximately 1,000 police officers, with other countries expected to contribute personnel and the United States providing substantial financial and logistical support.
The structure is politically innovative.
Its effectiveness remains uncertain.
Foreign police can help Haitian forces regain control of infrastructure and neighbourhoods.
They cannot permanently govern Port-au-Prince.
That distinction needs to determine the mission from the beginning.
A successful intervention should create enough security that Haitian institutions can expand into the space opened by security forces.
Police secure the road.
Government administrators return.
Schools reopen.
Businesses operate.
Courts process cases.
Local government delivers services.
The state becomes visible through institutions other than armed men.
If foreign forces simply patrol indefinitely while Haitian administrative capacity remains unchanged, the mission becomes security substitution rather than state-building.
The country appears more stable while the underlying institutional dependency survives.
This is why the sequencing of intervention matters.
Security needs to come first in areas where violence makes administration impossible.
But civilian institutions need to follow quickly.
Otherwise gangs return when security forces move elsewhere.
Counterinsurgency literature often describes this through clear, hold and build.
Clear armed organisations from territory.
Hold enough security that they cannot return immediately.
Build institutions and economic activity giving citizens reasons to cooperate with the state.
The terminology originated in military contexts and should not be applied mechanically to Haiti, where gangs are not one ideological insurgency.
The institutional principle remains useful.
Force can create space.
Only governance can stabilise it.
This is why defeating individual gang leaders may accomplish less than it appears. Leaders can be arrested or killed. Organisations survive if the political economy sustaining them remains.
Gangs recruit because young people face limited economic opportunities, because weapons are available, because politicians have historically developed relationships with armed groups and because controlling territory generates revenue through extortion and trafficking.
Security policy therefore needs to attack the business model.
How do gangs finance themselves?
Where do weapons come from?
Which businesses pay extortion?
Which political actors cooperate with them?
Which transport nodes generate revenue?
Follow the incentives.
A gang without financing cannot maintain the same number of armed members.
This is where international policy can help without occupying neighbourhoods. Many weapons used by Haitian gangs originate outside the country, particularly through trafficking networks connected to the United States.
Reducing weapons flows could therefore alter the balance between gangs and police without requiring foreign forces to govern Haiti permanently.
Again, network analysis matters.
The violence inside Port-au-Prince depends partly on supply chains outside Port-au-Prince.
Security becomes transnational even when sovereignty remains national.
Haiti’s political vacuum makes the problem harder because Ariel Henry was never elected prime minister through a normal democratic mandate. He took office after President Jovenel Moïse was assassinated in July 2021, and national elections have repeatedly been postponed amid insecurity and institutional collapse.
Haiti currently has no sitting elected national officials.
That creates an enormous legitimacy problem.
Security forces can restore government control.
Which government?
An election can produce legitimacy.
How do you conduct an election where candidates and voters cannot move safely?
Security appears necessary before democracy.
Democratic legitimacy appears necessary for sustainable security.
This is another institutional circularity.
Haiti needs legitimate government to rebuild the state.
It needs a functioning state to conduct the election producing legitimate government.
The transition therefore requires some temporary political arrangement capable of acquiring enough broad consent to govern while preparing conditions for elections.
That arrangement cannot derive full legitimacy from elections that have not occurred.
It therefore needs procedural legitimacy from consultation, transparency and a credible commitment to surrender authority once elections become possible.
Again, temporary government faces a time-consistency problem.
Officials controlling transitional institutions can develop incentives to postpone elections.
The solution is clear benchmarks and external and domestic monitoring.
But benchmarks need to respond to security reality.
Holding an election merely to satisfy a calendar can produce little legitimacy if gangs determine who can campaign in large parts of the capital.
Democracy requires more than ballots.
Pakistan demonstrated that last month.
The political field preceding voting matters.
Haiti represents the extreme version.
Before worrying about whether votes are counted correctly, the state needs to ensure voters can physically reach polling stations without permission from armed organisations.
This is why security itself becomes democratic infrastructure.
We often think of police power as something democracy needs protection from.
It is also something democracy requires.
Citizens cannot exercise political rights reliably when private armed actors control public space.
The liberal state therefore faces a permanent dual obligation: constrain public coercion while suppressing private coercion.
Too much state force threatens liberty.
Too little state force allows whoever possesses weapons privately to determine other people’s choices.
Freedom requires a state strong enough to prevent private domination and constrained enough not to become the dominant threat itself.
Haiti currently suffers primarily from the first failure.
The state is too weak to guarantee ordinary freedom from coercion.
This is why calls simply to reduce policing or state power make little sense in contexts of severe state weakness.
Institutional strength and authoritarianism are not synonyms.
Denmark has an extraordinarily capable state and strong civil liberties.
A gang-controlled neighbourhood can contain very little state power and very little freedom.
The relevant variable is constrained capacity.
Haiti needs much more state capacity.
It also needs institutions capable of preventing that capacity from becoming another vehicle for corruption or political repression.
Building both simultaneously is extraordinarily difficult.
The historical roots of Haiti’s weakness matter here, but they should not become deterministic. Haiti became the world’s first independent Black republic after enslaved people defeated French colonial rule and declared independence in 1804. France subsequently demanded an enormous indemnity in exchange for recognising Haitian independence, imposing debt burdens that extracted resources for generations.
Foreign intervention, political instability, dictatorship, economic isolation and natural disasters compounded those historical constraints.
The devastating 2010 earthquake killed enormous numbers of people and destroyed state infrastructure.
But history should explain constraints, not become an argument that present failure was inevitable.
Countries do not remain institutionally weak because of one event two centuries earlier.
Historical shocks alter development paths.
Political choices continue operating within those paths.
This is path dependence without determinism.
Haiti’s history made state-building harder.
It did not make every later governance failure unavoidable.
Domestic political elites also bear responsibility for corruption, patronage and relationships with armed groups.
International actors bear responsibility for interventions that sometimes weakened institutions they claimed to strengthen.
Both levels can matter simultaneously.
Again, responsibility is not zero-sum.
The useful question is which mechanisms need changing now.
Security.
Political legitimacy.
Administrative capacity.
Economic opportunity.
Weapons trafficking.
Judicial institutions.
No single intervention solves all six.
That is why Haiti has repeatedly experienced cycles where one crisis receives enormous international attention, emergency resources arrive, immediate conditions improve somewhat and global attention then moves elsewhere before institutions become durable.
Crisis financing is easier than institution financing.
Donors like measurable projects.
Build a school.
Train police.
Deliver food.
Institutional development is slower.
Create a tax system people trust.
Build courts capable of processing cases consistently.
Professionalise a police force over twenty years.
Create local governments capable of delivering services.
Those outcomes do not fit easily inside short funding cycles.
Yet they determine whether the next crisis requires another intervention.
This is the time-horizon mismatch of development assistance.
Donors operate through political and budget cycles.
State-building operates through generations.
Haiti needs external support capable of surviving beyond the moment when television coverage declines.
That support also needs to become progressively less necessary.
Successful aid should eventually destroy the demand for itself.
The same principle applies to Ukraine reconstruction but under radically different institutional conditions.
External resources should build domestic capability.
Otherwise assistance becomes permanent substitution.
The present crisis makes that long-term objective feel almost abstract because immediate security is collapsing.
But strategic policy needs both horizons simultaneously.
Police need to regain control of Port-au-Prince now.
The political system needs to create a government citizens consider legitimate.
The economy needs enough stability that young people face alternatives to gang recruitment.
Institutions need enough capability that another gang coalition cannot recreate the same crisis five years later.
Those objectives form a sequence.
Skipping stages produces fragility.
Security without politics becomes coercion.
Politics without security becomes symbolism.
Economic development without institutions becomes vulnerable to capture.
Institutions without revenue remain dependent.
State-building is therefore a complementarity problem.
Every component becomes more valuable when the others exist.
That makes initial recovery particularly difficult because Haiti is weak across several dimensions simultaneously.
It also explains why the country can remain trapped in a low-capacity equilibrium.
Businesses do not invest because security is weak.
Security remains weak because tax revenue is low.
Tax revenue is low because businesses do not invest.
Citizens distrust government because services are weak.
Government cannot improve services because citizens and businesses avoid formal institutions.
Gangs fill the vacuum.
Gang power further weakens formal institutions.
Every individually rational response reinforces collective failure.
Escaping requires coordinated intervention across several parts of the equilibrium.
That is what makes state failure much harder to solve than state failure is to describe.
The international community can deploy police.
It cannot import legitimacy in shipping containers.
It can provide money.
It cannot purchase social trust instantly.
It can help organise elections.
It cannot decide whom Haitians should consider legitimate.
Those resources matter.
Their limits matter equally.
Ultimately, the Haitian state has to become capable because Haitian institutions become capable.
External actors can create the conditions.
They cannot permanently substitute for the outcome.
This week’s violence therefore should not be understood simply as gangs becoming stronger.
It is the visible consequence of the state becoming too weak relative to them.
Power is relational.
A gang does not need to become an army if the police force deteriorates faster.
An armed group does not need to govern the whole country if controlling several strategic nodes allows it to determine what the government can do.
The institutional objective is therefore not merely eliminating gangs.
It is changing the balance so profoundly that gangs cease being plausible competitors for political authority.
That requires a state capable of enforcing law consistently enough that joining a gang becomes high-risk and participating in the formal economy becomes comparatively attractive.
Force matters.
Opportunity matters.
Legitimacy matters.
Administration matters.
Haiti still exists as a sovereign state internationally.
Its flag flies at the United Nations.
Its borders appear clearly on every map.
Other governments recognise its government rather than its gangs.
None of that guarantees sovereignty in everyday life.
For a resident deciding whether they can travel to work, open a shop or send a child to school, sovereignty is not an international legal principle.
It is whether somebody with a gun can stop them and whether any institution can stop that person.
That is the difference between possessing a state on paper and experiencing one in practice.
Haiti still has the first.
The crisis is about rebuilding the second.
