Somewhere in a commercial arbitration chamber, a contractor who built a port in West Africa is being told that the state shipping company that refused to pay him cannot be touched in any court he can afford to reach. The company is profitable. It has assets. It simply belongs, in some legally meaningful sense, to a government. That government, the doctrine of sovereign immunity insists, is not subject to the jurisdiction of foreign courts without its consent.
This is not an obscure technicality. It is the reason billions of dollars in legitimate commercial claims go unpaid every year, and the reason that doing business with a state-owned enterprise carries a legal risk that doing business with a private company does not.
The idea that the king cannot be sued in his own courts
Sovereign immunity is old. Its classical form held that a state was simply immune from the jurisdiction of any foreign court, full stop. The logic was feudal in origin and diplomatic in practice: courts are instruments of state power, and no state's courts can presume authority over another sovereign. By the twentieth century, this absolute version of the doctrine had become untenable. States were no longer just governing; they were trading, building railways, running airlines, issuing bonds. Letting them conduct all of that behind an impenetrable legal shield was, creditors argued, intolerable.
The response was what lawyers call the restrictive theory of sovereign immunity. Under this approach, a state retains immunity for its governmental acts (in Latin, jure imperii) but not for its commercial ones (jure gestionis). The United States codified this in the Foreign Sovereign Immunities Act of 1976. The United Kingdom followed with the State Immunity Act of 1978. Most developed jurisdictions have adopted some version of the same framework since.
The practical question, then, is not whether immunity exists. It is whether the entity being sued was acting as a government or as a merchant.
Where state-owned enterprises fall, and why it isn't obvious
An SOE occupies genuinely ambiguous ground. Take a national oil company. When it negotiates a drilling contract with a foreign engineering firm, that looks commercial. When it controls pipeline access as a matter of energy policy, that looks governmental. The same entity, on the same day, can be doing both.
Courts in different jurisdictions have developed tests to sort this out, and they do not always agree. American courts under the FSIA ask primarily about the nature of the activity: was it the kind of thing a private actor could have done? A central bank buying treasury bonds on the open market probably qualifies as commercial. A government ministry requisitioning private property almost certainly does not. British courts have historically given more weight to the purpose of the act alongside its nature, which tends to produce broader immunity.
Then there is the separate question of whether the SOE is even a distinct legal person from the state itself. Some national oil companies are incorporated entities with their own shareholders and boards. Others are effectively ministerial departments wearing a corporate costume. Courts treat these differently. An incorporated SOE in many jurisdictions is presumed to be a separate entity, not automatically immune, unless the plaintiff can show that the state so thoroughly controls it that the corporate form is a fiction. Proving that is hard. States are not naive about this: they structure their enterprises deliberately to benefit from commercial flexibility while preserving as much immunity as possible. It is, in effect, a legal arbitrage that has been running for decades, and creditor-country courts have been losing ground to it more often than they would like to admit.
Consider a worked scenario. A German equipment manufacturer, call them Bauer Industrietechnik, signs a five-year supply contract with Petronac, the national petroleum company of a mid-sized oil exporter. Petronac is incorporated under domestic law, has its own chief executive, and issues audited accounts. After two years, Petronac stops paying invoices, citing a government decree that reclassified all foreign contracts as subject to ministerial approval. Bauer sues in a German court. Petronac's lawyers argue that the decree was a governmental act, immunizing the breach. Bauer's lawyers argue that the underlying contract was commercial and the immunity waiver in the contract's arbitration clause applies. A German court will have to decide whether Petronac is sufficiently separate from the state, whether the commercial activity exception applies, and whether the arbitration clause constitutes an effective waiver. Each of those questions has a real answer, but none of them is fast or cheap to obtain.
What people consistently get wrong
The most common misconception is that a signed arbitration clause solves the problem. It doesn't. Not even close.
An arbitration clause can waive immunity from jurisdiction, meaning the SOE agrees to participate in arbitration proceedings. But immunity has a second, distinct layer: immunity from execution, which protects assets from being seized to satisfy an award. Winning an arbitration against a state-owned airline is meaningfully different from collecting on that win. The airline's aircraft, parked at foreign airports, may be protected from seizure under separate treaty frameworks or domestic immunity statutes, even if the arbitration clause was airtight.
The prolonged enforcement battle that followed Argentina's nationalization of its state oil company YPF illustrated this gap with unusual clarity. Creditors who held favorable judgments found that actually attaching Argentine state assets in foreign courts was a years-long project requiring creative lawyering across multiple jurisdictions. A court judgment against a recalcitrant sovereign is, in the worst cases, less a legal instrument than a very expensive piece of paper. Winning on paper and winning in practice are not the same thing.
Some jurisdictions carve out exceptions for commercial assets specifically. Under the FSIA, property used for commercial activity in the United States can be attached to satisfy a judgment, even if other assets are protected. But the plaintiff must identify those assets, which requires its own investigation, and the definition of "used for commercial activity" is litigated constantly.
What this means for anyone on the other side of an SOE contract
The doctrine shapes behavior before a dispute ever arises. Sophisticated counterparties to SOE contracts now routinely negotiate explicit, broad immunity waivers covering both jurisdiction and execution. They seek to have disputes resolved in jurisdictions with mature enforcement frameworks. They try to identify, in advance, what assets the SOE holds in neutral third countries. Some require letters of credit or performance bonds precisely because they know that a court judgment against a sovereign entity may be more decorative than useful.
Ask yourself this: if the counterparty to your contract can pass a law to escape it, what exactly is your contract worth?
None of the standard precautions are foolproof. A state that decides, for political reasons, that it will not honor a commercial obligation has tools available to it that a private debtor does not. It can pass a law. It can invoke national security. It can simply refuse and wait out the reputational damage, calculating that it needs the contract more than the contractor needs it.
Sovereign immunity did not begin as a mechanism to protect bad actors. It began as a principle of international order, the recognition that states must be able to govern without foreign courts second-guessing every act of administration. The restrictive theory was a reasonable attempt to separate governing from trading. The problem is that states, particularly those that own large commercial enterprises, have never fully accepted that distinction. They want the credibility of private commerce and the protection of sovereign status simultaneously, which is rather like a bank robber claiming the rights of a depositor. Courts in creditor countries have spent half a century trying to stop them from having both at once. The outcome, case by case, remains genuinely uncertain, and the contractor in the West African arbitration chamber has probably already called his lawyer about the cost of the next appeal.