The Hague Has 24 Judges, 193 Courtrooms, and No Army. That Is the Problem.

There is a detail about the International Criminal Court that tells you everything you need to know about what it can and cannot do. The building has 193 courtrooms — one for each member state of the United Nations — but no detention facilities. Prisoners are held in a Dutch prison twenty minutes away. The symbolism is unintentional but precise: the Court can convict, but it cannot compel. It can issue warrants, but it cannot make arrests. It can declare what justice requires, but it cannot deliver it.

I am thinking about this because in March 2023, the Court issued an arrest warrant for Vladimir Putin. The warrant cited reasonable grounds to believe he bore individual criminal responsibility for the war crime of unlawful deportation of children from Ukraine to Russia. Putin has since traveled to Mongolia, a state party to the Rome Statute, which did not arrest him. He has visited Kyrgyzstan, also a signatory, which welcomed him with military honours. The warrant remains in force. Putin remains at large.

This is not a failure of jurisprudence. It is a design flaw. The international legal system was built on the premise that states would enforce its judgments voluntarily, out of moral obligation or reputational concern. What we have discovered, across eight decades of war crimes tribunals, is that moral obligation is not enforceable and reputation is not fungible into arrests.

The Architecture of Good Intentions

The International Criminal Court opened in The Hague in July 2002. It was the culmination of a century-long project: to create a permanent institution that could prosecute genocide, crimes against humanity, war crimes, and the crime of aggression. The Nuremberg and Tokyo tribunals had been improvised after World War II. The International Criminal Tribunal for the former Yugoslavia, established in 1993, and the International Criminal Tribunal for Rwanda, established in 1994, were ad hoc. The ICC was supposed to be different: standing, independent, universal.

The Rome Statute, the treaty that created the Court, was signed by 120 states in 1998. As of April 2026, 124 states are party to it. The United States is not among them. Neither is Russia, China, India, Pakistan, Israel, Turkey, Saudi Arabia, Iran, or Egypt. The Court has jurisdiction over nationals of member states, crimes committed on the territory of member states, and situations referred by the UN Security Council. In practice, this means the Court prosecutes crimes committed in weak states by officials who lack powerful patrons.

The pattern is visible in the case record. Joseph Kony, leader of the Lord's Resistance Army in Uganda, was indicted in 2005. He has never been arrested. Omar al-Bashir, president of Sudan, was indicted in 2009 for genocide in Darfur. He remained in office for another decade and has not been surrendered. Saif al-Islam Gaddafi, indicted in 2011 for crimes against humanity in Libya, was released by a militia in 2017 and ran for president in 2021. The Court's inability to secure custody is not incidental. It is structural.

What Enforcement Requires

The ICC has no police force. It has no intelligence service. It cannot freeze bank accounts, seize assets, or close borders. Article 86 of the Rome Statute requires state parties to cooperate fully with the Court's investigations and prosecutions. Article 89 requires them to arrest and surrender persons for whom the Court has issued a warrant. But the Statute provides no mechanism for compelling cooperation. There are no sanctions for non-compliance. There is no enforcement body. The only remedy is referral to the Assembly of States Parties or the UN Security Council, both of which are political bodies unlikely to act against their own members or allies.

This was a deliberate choice. The drafters of the Rome Statute understood that no state would surrender sovereignty to a supranational prosecutor with coercive powers. The Court was designed to complement, not replace, national legal systems. It would step in only when states were unwilling or unable to prosecute. It would rely on voluntary cooperation, moral authority, and diplomatic pressure. What the drafters did not fully reckon with was what happens when the accused is himself the source of state power.

Consider the case of Félicien Kabuga, the Rwandan businessman indicted in 1997 for financing the genocide that killed 800,000 people in 1994. He was arrested in Paris in May 2020 — after 23 years on the run. He had lived openly in Kenya, Switzerland, and France. His capture required the cooperation of French intelligence, Interpol, and the UN tribunal for Rwanda. It was not the law that brought him to justice. It was political will, which appeared only after decades of diplomatic pressure and only because he no longer had powerful protectors.

The Selectivity Problem

The Court's critics argue that it enforces justice selectively — that it prosecutes Africans while ignoring American drone strikes, Russian bombardment, Chinese repression in Xinjiang, and Israeli settlement expansion. The criticism is not entirely fair, but it is not entirely wrong. Of the Court's 32 cases since 2002, 19 have involved African suspects. The Court has opened investigations into Afghanistan, Palestine, the Philippines, and Ukraine, but prosecutions in those situations face political and practical obstacles that do not apply in Uganda or the Central African Republic.

The Court is not biased in the sense that its prosecutors seek out African defendants. It is biased in the sense that the international system allows certain states to operate beyond its reach. The United States passed the American Service-Members' Protection Act in 2002, which authorizes the president to use military force to free any American held by the ICC. The law has never been invoked, but its existence signals clearly that American soldiers will not face trial in The Hague. Russia withdrew its signature from the Rome Statute in 2016, after the Court declared Russia's annexation of Crimea an occupation. China has never signed.

The Court is thus constrained by a paradox: it has jurisdiction over crimes that powerful states commit, but it lacks the power to prosecute them. It can act only where geopolitics permits. This does not make the Court irrelevant, but it does make it partial. Justice that applies only to the weak is not universal justice. It is selective enforcement disguised as law.

What Enforcement Could Look Like

There are two ways to understand the Court's limitations. One is to accept them as inevitable. International law, on this view, is not law in the domestic sense. It is a framework for negotiation, not coercion. States comply when compliance serves their interests. They do not comply when it does not. The Court's role is to articulate norms, document violations, and provide a venue for prosecution when political conditions allow. This is not nothing. But it is not justice.

The other way is to imagine what enforcement would require. Not enforcement as fantasy — an international police force arresting heads of state — but enforcement as architecture. This would mean, at minimum, three things the current system lacks.

First: automatic sanctions for non-compliance. States that refuse to arrest indicted individuals could face suspension of UN voting rights, exclusion from international financial institutions, or targeted sanctions on officials who provide safe haven. This would require amending the UN Charter or the Rome Statute, which no major power will support. But the absence of political will is not the same as the absence of legal possibility.

Second: a dedicated investigative capacity. The ICC's Office of the Prosecutor has 900 staff members and an annual budget of approximately €171 million. By comparison, the U.S. Department of Justice employs more than 115,000 people and has a budget of $37 billion. The Court investigates crimes across continents with resources comparable to a mid-sized municipal police department. If enforcement is to be credible, the Court needs intelligence capacity, forensic expertise, and the ability to operate in conflict zones. It currently has none of these at scale.

Third: reciprocity. The Court can only prosecute nationals of states that have ratified the Rome Statute or crimes committed on their territory, unless the situation is referred by the Security Council. This creates a perverse incentive: states outside the system face no consequences for non-cooperation, while states inside the system bear the costs. A reformed system would require universal jurisdiction for core international crimes, enforced through mutual legal assistance treaties and extradition agreements. This is not unprecedented. It is how Interpol functions. But it would require the major powers to submit to a legal framework they currently refuse.

The Question We Avoid

I think often of a conversation I had in 2019 with a former prosecutor at the International Criminal Tribunal for the former Yugoslavia. He had spent five years building the case against Ratko Mladić, the Bosnian Serb general who orchestrated the Srebrenica massacre in July 1995. Mladić was indicted in 1995. He was arrested in Serbia in 2011 — sixteen years later. He was convicted in 2017 and sentenced to life imprisonment. The prosecutor told me that the conviction felt hollow. Not because it was unjust, but because it was so late. The survivors had waited decades. Many had died. The conviction could not restore what was lost.

This is what enforcement means in practice: not the abstract vindication of norms, but the timely delivery of consequences. A warrant that takes sixteen years to execute is not deterrence. It is documentation. And documentation, however meticulous, is not the same as justice.

The question we avoid asking is whether the international legal system, as currently constituted, is capable of delivering what it promises. We prefer to say that it is a work in progress, that norms take time to consolidate, that the arc of history bends toward justice. But the families waiting in Srebrenica, in Darfur, in Bucha, do not have time. They have graves.

Who Benefits From Delay

There is a way in which the current system serves certain interests precisely because it is ineffective. It allows states to signal commitment to human rights without bearing the costs of enforcement. It provides a venue for symbolic condemnation without the risk of reciprocal scrutiny. It lets powerful countries criticize crimes they themselves commit, knowing that the Court cannot reach them.

Consider the case of the United Kingdom. Britain is a state party to the Rome Statute. It has championed the Court publicly and contributed to its budget. But when the ICC opened a preliminary examination into alleged British war crimes in Iraq between 2003 and 2009, the British government launched a domestic investigation designed to shield soldiers from international prosecution. The investigation, called the Iraq Historic Allegations Team, closed in 2017 after reviewing 3,400 allegations and producing no prosecutions. The ICC accepted this outcome and closed its examination. The system worked exactly as designed: it allowed Britain to avoid accountability while maintaining its reputation as a supporter of international justice.

This is the structural problem that no reform can fully address: the Court exists within a system of sovereign states that will not surrender the power to decide when their own officials face prosecution. Enforcement would require states to cede authority they have never ceded and will not cede. The alternative — a world without international courts — is worse. But acknowledging that does not change the fact that the current system is a compromise that privileges power over principle.

The Reckoning

I return to the image of the building in The Hague. One hundred and ninety-three courtrooms. No army. No police. No detention center. It is a monument to the idea that law can exist independent of force, that justice can be rendered through words rather than weapons. It is a beautiful idea. It is also a fiction.

Every legal system, at bottom, rests on the threat of coercion. You comply with the law not only because it is just, but because non-compliance has consequences. The international system lacks this architecture. It has prosecutors and judges and courtrooms, but it has no mechanism for converting judgments into custody, indictments into arrests, warrants into prisoners. It is law without enforcement, and law without enforcement is aspiration.

This does not mean the Court is worthless. Documentation matters. Norms matter. The gradual accumulation of precedent matters. But we should not pretend that documentation is the same as justice, or that norms are the same as deterrence. We should not tell the survivors of Srebrenica, of Darfur, of Bucha, that the system is working when the architects of their suffering remain free.

The choice before us is not between the current system and perfection. It is between the current system and something closer to enforcement. That would require states to surrender sovereignty they will not surrender. It would require funding tribunals at the level needed to conduct serious investigations. It would require automatic penalties for non-compliance. It would require powerful countries to submit to the same legal framework they demand of others. None of this is likely. But unlikely is not the same as impossible.

Until then, we have what we have: a court with jurisdiction but no power, warrants but no arrests, judgments but no enforcement. We have the architecture of justice without the force to make it real. We tell ourselves that this is progress, that it is better than nothing, that the arc of history bends in our favour. Maybe it does. But the bodies keep accumulating, and the perpetrators keep walking free, and the courtrooms in The Hague keep waiting for defendants who never arrive.