The floating borders – how maritime law decides who owns the sea

An Arleigh Burke-class guided-missile destroyer is transiting the Spratly Islands. Two People’s Liberation Army Navy frigates shadow it at close range, broadcasting warnings over marine radio. The American crew logs the encounter as routine. The Chinese crews log it as an illegal incursion into sovereign waters.

Both governments are invoking the same treaty.

The United States describes the operation as a freedom-of-navigation exercise conducted under UNCLOS — the UN Convention on the Law of the Sea — which its lawyers cite by article number. China, which ratified UNCLOS in 1996, declares the transit a violation of the same treaty. In July 2012, 34 Republican senators pre-committed to vote against UNCLOS ratification — enough to deny the two-thirds Senate supermajority the treaty requires. No floor vote was scheduled. None has been since.

One country is enforcing a treaty it hasn’t joined. The other joined the treaty and declared its rulings inapplicable when they went against it. The treaty itself — 320 articles, nine annexes, its own international courts, three and a half decades of jurisprudence — can tell you precisely which of them is wrong. It cannot make either of them stop.

That’s the situation.

Lines on Water

UNCLOS emerged from nine years of negotiation. The Third UN Conference on the Law of the Sea ran from 1973 to 1982, with roughly 160 nations at the table, and produced 320 articles and nine annexes in service of one ambition: to write down, precisely, who owns what in the ocean.

Before UNCLOS, the answer was a muddle. The 1958 Geneva Conventions had codified some maritime norms but left the central question unresolved — how far out does a coastal state’s claim extend? Three nautical miles? Twelve? Two hundred? Different states asserted different answers. Fishing vessels were impounded. Oil rigs contested. Fishermen arrested in waters their families had worked for generations. UNCLOS resolved this through zones, each with a distinct legal character.

The innermost is the territorial sea, running twelve nautical miles from a coastal state’s baseline. Within it, the state exercises full sovereignty — same legal status as its land, extending to the airspace above and seabed below. Foreign warships may transit under the doctrine of innocent passage, which means they move through without stopping, without conducting anything the coastal state might characterize as threatening. But the zone belongs to the coastal state.

Beyond twelve miles and out to twenty-four runs the contiguous zone — an enforcement reach without a sovereignty claim. A government can pursue and punish violations of its customs, immigration, fiscal, and sanitary laws, but only for violations that originated within its own territory. Extended arm, not extended territory.

Then the zone that drives the conflicts. Out to 200 nautical miles from the baseline, the coastal state holds an Exclusive Economic Zone: the absolute right to fish, drill, mine, and generate energy within those waters. No sovereignty over the water column itself — foreign ships pass freely, foreign aircraft fly overhead — but a total monopoly on resource extraction. The EEZ concept didn’t exist in the 1958 Geneva Conventions. UNCLOS invented it, and its scale is extraordinary.

The Maldives has a total land area of roughly 298 square kilometers — smaller than most mid-sized cities. Its EEZ covers approximately 923,322 square kilometers of ocean. That jurisdictional footprint is more than three thousand times its land area, and it exists entirely because the Maldives has a coastline. This is why islands matter. This is why even rocks matter.

Beyond 200 miles, where the seabed is a geological extension of a state’s continental landmass, UNCLOS permits claims out to 350 nautical miles. These require scientific submissions to the Commission on the Limits of the Continental Shelf, which reviews them and issues recommendations. More than 60 countries had made CLCS submissions as of 2024. In the Arctic, Russia, Canada, and Denmark have each filed overlapping claims asserting rights to the same submarine ridge. Russia revised its submission in 2015; the CLCS issued recommendations in February 2023 that Russia accepted — but the CLCS validates science, it doesn’t draw borders. Where two states have each submitted scientifically credible claims to the same area, the boundary between them falls to bilateral negotiation, not the CLCS process. Russia and Norway managed this: a 2010 delimitation treaty settled a forty-year Barents Sea dispute. Russia, Canada, and Denmark have not resolved theirs.

Add up every EEZ asserted by every coastal state and the resulting coverage approaches roughly 42 percent of the world’s ocean surface. The ocean is fully addressed. Often, more than one state is claiming the same address.

The High Seas and the Common Heritage Principle

In 1967, Arvid Pardo, Malta's ambassador to the United Nations, gave a speech to the UN General Assembly proposing something genuinely radical: the seabed beyond national jurisdiction should be declared the "Common Heritage of Mankind." Not owned by any state, managed collectively, revenues shared among all nations. UNCLOS Part XI enacted this, creating the International Seabed Authority to govern deep-sea mining in international waters and directing a share of its revenues to developing nations.

Reagan refused to sign UNCLOS in 1982. Part XI was the stated reason — the ISA's revenue-sharing mechanism, which his administration characterized as a wealth transfer from industrialized nations to the developing world. The characterization was accurate. The 1994 Implementation Agreement then renegotiated Part XI almost entirely: redistribution mechanisms gutted, technology-transfer requirements removed, the US guaranteed a permanent seat at the ISA's decision-making council. The treaty that emerged was nearly unrecognizable as the document Reagan had refused. The Senate still hasn't ratified it.

Where the Lines Fight Each Other

The South China Sea covers approximately three million square kilometers. Six governments claim pieces of it: China, Taiwan, Vietnam, the Philippines, Malaysia, and Brunei. Extend 200-nautical-mile EEZs from each of their coastlines and islands and the circles overlap catastrophically.

China’s claim is distinctive and, from a legal standpoint, deliberately obscure. A 1947 Republic of China government map published on December 1 of that year showed an eleven-dash U-shaped line enclosing most of the South China Sea. The People’s Republic inherited the claim, dropped two of the dashes in 1952 as a gesture of solidarity with North Vietnam — abandoning Chinese claims to the Gulf of Tonkin — and has maintained the nine-dash line ever since. The line sweeps around approximately 90 percent of the South China Sea, running directly through every other claimant’s EEZ assertions. And China has, for decades, refused to specify what the line legally means — sovereignty? historic rights? sovereign rights? The ambiguity is not a diplomatic failure. It is the strategy: keeping the claim legally operative across multiple frameworks while making adjudication nearly impossible.

In January 2013, the Philippines forced the question. Manila filed for compulsory arbitration under UNCLOS Annex VII — the provision allowing a party to initiate proceedings against another party without the other’s consent. China refused to participate, declared the tribunal without jurisdiction, and called the proceedings “political theatre.” The panel convened anyway.

On July 12, 2016, the Permanent Court of Arbitration issued its award. The Philippines had spent three years building the case; the ruling runs to 501 pages. The findings were specific and comprehensive: China’s nine-dash line has no legal basis under UNCLOS. Mischief Reef and Second Thomas Shoal lie within the Philippines’ EEZ, not within any Chinese maritime entitlement. None of the features China occupies in the Spratlys qualify as islands under Article 121 of UNCLOS — capable of generating their own 200-mile EEZ. Most are rocks or low-tide elevations with far more limited consequences.

China’s Ministry of Foreign Affairs issued a statement the same day. The award, it declared, was “null and void and has no binding force.”

Read that statement carefully, because what it doesn’t contain matters as much as what it does. There is no sentence engaging with the tribunal’s legal reasoning. No argument that the methodology was flawed. No alternative interpretation of Article 121. No rebuttal at all — just a flat declaration that because the outcome was inconvenient, the process hadn’t happened. China then continued construction on the features the tribunal had ruled it was illegally occupying. Satellite imagery tracked by the Center for Strategic and International Studies documented military-grade airstrips, port facilities, and radar installations across multiple Spratly features — construction ongoing.

The lesson isn’t that China is uniquely lawless — the US walked away from an ICJ ruling in the Nicaragua case when it went the wrong way. The lesson is structural: UNCLOS’s dispute resolution system produced a binding, procedurally impeccable award, and the state with the power to disregard it did, and the treaty contains nothing to make that cost more than a press release from Manila.

The tribunal’s authority extends precisely as far as the sovereign chooses to let it run.

What the Tribunal Actually Found

UNCLOS Article 121 classifies maritime features into three tiers. An island — land above water at high tide, capable of sustaining human habitation or economic life of its own — generates a full 200-nautical-mile EEZ. A rock — above water at high tide but incapable of sustaining habitation or independent economic life — generates only a 12-mile territorial sea. A low-tide elevation — submerged at high tide — generates nothing.

The 2016 tribunal applied these categories to every feature China had claimed, occupied, or built on. Scarborough Shoal: a rock. Most Spratly features: rocks or low-tide elevations. What China had positioned as potentially generating continent-scale EEZ rights turned out, under legal scrutiny, to generate almost nothing beyond 12 miles. The ruling stripped the legal basis from enormous potential maritime claims. That China continues to occupy those features regardless is precisely the gap between legal clarity and legal consequence.

The Bedroom Problem

Okinotorishima sits roughly 1,740 kilometers south of Tokyo in the Philippine Sea. It exists in the public record almost entirely because of what Japan has spent to keep it above water.

At high tide, two areas remain exposed: one roughly the size of a twin bed, the other of a small bedroom. Combined exposed surface: approximately ten square meters. The Japanese government has spent more than $600 million on the atoll — $280 million encasing the coral in concrete, another $50 million installing a titanium debris net to hold it against the waves and the slow rise of sea levels that would otherwise finish it.

The reason is a single provision of UNCLOS.

Article 121(3): “Rocks which cannot sustain human habitation or economic life of their own shall have no exclusive economic zone or continental shelf.” Japan classifies Okinotorishima as an island. China, South Korea, and Taiwan classify it as a rock. If Japan is right, it holds an EEZ of approximately 400,000 square kilometers — larger than Japan’s own land area, larger than Germany. If the critics are right, Japan gets a 12-mile territorial sea and nothing beyond.

Japan has done what it can to manufacture the conditions for island status. Research facilities operate on the atoll. Tokyo argues their presence constitutes human habitation under Article 121. China’s argument is literal: the provision says “economic life of their own,” and a government-funded weather station staffed by rotating researchers on government salaries is not an economy, and certainly not an independent one. The phrase “of their own” — self-sustaining? government-subsidized but permanent? any human presence at all? — is exactly the question UNCLOS fails to answer with the precision it elsewhere excels at.

No ITLOS panel and no UNCLOS arbitration tribunal has formally ruled on Okinotorishima’s classification. In 2012, the CLCS declined to make recommendations on Japan’s extended continental shelf submissions in the area, citing objections from China and South Korea and noting it would wait until the dispute was resolved. In May 2025, as a Chinese research vessel operated near the atoll, China’s Ministry of Foreign Affairs reiterated its position: Okinotorishima is a reef, not an island, and Japan’s claimed EEZ around it has no legal basis under UNCLOS. The dispute has not moved toward resolution. The rocks remain contested, and Japan keeps the concrete coming.

Stepped back from, the accounting is almost funny. A sovereign nation has spent roughly the annual GDP of a small Pacific island state preserving two rocks above the waterline because a treaty drafted in 1982 did not define with sufficient precision what separates a rock from an island.

But Okinotorishima is a different failure mode from the South China Sea. It’s not a powerful state bulldozing a ruling it dislikes — no ruling exists to bulldoze. It’s UNCLOS’s architecture generating its own bizarre incentive structures through the gap between what the treaty intended and what it actually said. The ambition created the stakes. The definitional imprecision created the incentive to spend half a billion dollars gaming them.

The State That Runs the Show Without Joining It

Reagan’s refusal to sign UNCLOS in 1982 rested almost entirely on Part XI. The deep-seabed mining regime — specifically the International Seabed Authority’s mandate to share extraction revenues with developing nations and require technology transfers — struck his administration as a redistribution mechanism hostile to American industry. The deep ocean was being declared common property, its riches taxed and shared. Reagan declined.

Twelve years later, the 1994 Implementation Agreement renegotiated Part XI from scratch. Revenue redistribution: gutted. Technology-transfer requirements: removed. The US: guaranteed a permanent seat at the ISA’s decision-making council. The treaty that emerged was, on the substance of Reagan’s specific objections, almost unrecognizable as the document he’d refused.

Every administration from Clinton through Biden supported ratification. The military establishment was, if anything, even more uniform: every Joint Chiefs Chairman who testified on the subject, every Secretary of Defense who addressed it, recommended ratification — because non-party status creates persistent diplomatic liability every time the Navy runs a freedom-of-navigation operation against a state that can point out the US isn’t even a member of the treaty it’s enforcing.

The Senate blocked it each time the question approached a vote.

The 2012 case is the most precisely documented. By July 16 of that year, 34 Republican senators had pre-committed to vote against ratification — coordinated by Jim DeMint of South Carolina and Jim Inhofe of Oklahoma, with Portman and Ayotte among the last to commit. With 34 pre-committed noes, scheduling a floor vote would have been scheduling a public defeat. No vote was scheduled.

The stated objections: sovereignty concerns, ISA governance, and UNCLOS Article 82, which requires royalty payments on extended continental shelf resource production to flow to the ISA — starting at 1 percent in the sixth year of production and rising by 1 percent annually to a maximum of 7 percent. Senators from oil-producing states found the provision particularly objectionable.

The US has maintained this position without visible embarrassment for four decades. The mechanism is customary international law: rules so widely and consistently followed by the international community become binding even on states that never formally accepted them. On this basis, the US argues UNCLOS’s navigational provisions don’t reflect treaty obligations so much as pre-existing customary law binding on all states. It is convenient: the US invokes customary law for the UNCLOS provisions that benefit it and declines to participate in the institutional machinery — ITLOS, compulsory arbitration — that might produce rulings against it.

In practice: the US claims its own 200-nautical-mile EEZ, one of the world’s largest. It runs freedom-of-navigation operations invoking UNCLOS principles against Chinese, Iranian, and other states’ maritime claims it considers excessive. And in late 2023, the State Department issued an extended continental shelf claim asserting resource rights in the Arctic and Pacific beyond 200 nautical miles — rights whose entire legal basis derives from the existence of UNCLOS.

The mirror image of the South China Sea is difficult to miss. China ratified UNCLOS and declared its rulings inapplicable when they went against it. The US declined to ratify and declares its provisions applicable when they benefit it. Both states use the treaty as a vocabulary of legitimacy — invoking its authority when useful, ignoring its obligations when costly. Neither is constrained by it when constraint would actually cost something.

The treaty functions not as law in any domestic sense but as a shared reference point for arguments about who is behaving correctly — one that every party cites selectively and no party is compelled to honor.

What a FONOP Actually Is

Freedom-of-navigation operations are routinely described as military provocations — warships sent to antagonize a rival. That framing misses the legal mechanism. A FONOP is a deliberate assertion, conducted through navigation, that a stated restriction on that navigation has no legal basis. The US Department of Defense publishes an annual Freedom of Navigation Report listing which states' claims were challenged, which specific assertions were at issue, and on what legal grounds the US considers them excessive.

China appears on the list regularly, as do India and numerous US allies. The FONOP program is fundamentally about building a legal record — a consistent, documented practice of navigation supporting the US position that UNCLOS freedom-of-navigation provisions constitute customary international law binding on all states, not just UNCLOS parties. If the US never challenges a claim, it risks implicitly acknowledging it. The warship is not primarily a threat. It is a legal filing delivered at twelve knots.

The Most Sophisticated Body of Nothing

UNCLOS has more institutional enforcement machinery than almost any other instrument of international law. The International Tribunal for the Law of the Sea had 33 cases on its docket as of early 2025 — seized fishing vessels, maritime boundaries, prompt release of detained ships. For states roughly matched in power, or where noncompliance costs more than compliance, the system works. ITLOS rulings have mattered. Countries have complied.

UNCLOS Annex VII provides compulsory arbitration that any party can trigger against any other, even without consent. The Philippines used it. The tribunal produced the most elaborate award the system has ever generated: three years of proceedings, 501 pages, every contested feature individually classified and adjudicated.

The state with the power to ignore it did.

And the US, simultaneously, spent four decades blocking ratification of the same treaty while claiming its EEZ, conducting FONOPs under its navigational provisions, and asserting resource rights in the Arctic and Pacific that derive their legal basis entirely from the treaty’s existence. Two different relationships to the same document, producing the same structural outcome: UNCLOS as vocabulary, not constraint.

The enforcement gap is not a design flaw a better treaty could fix. Stronger sanctions, expanded ITLOS jurisdiction, mandatory compliance review — none of these address what is actually the problem: sovereignty means something. A state is the final authority within its domain, and no treaty overrides that without the state’s consent. When a state withdraws consent, it withdraws it. International law works when compliance is more advantageous than defiance — when the costs in reputation, economics, and strategic positioning outweigh the costs of compliance. When that calculation flips, the treaty cannot force it back.

UNCLOS is distinctive only in how completely it makes this visible.

Most international legal instruments leave enough ambiguity for states to construct plausible cover for noncompliance — a competing interpretation, a reservation, some alternative norm. UNCLOS, with its 320 articles and nautical-mile precision, stripped most of that away. When the 2016 tribunal ruled, China could not plausibly argue the ruling was unclear. It could only say it didn’t count. That is not a legal argument. It is a power claim dressed in legal language — and UNCLOS’s precision makes the difference between the two unmistakable.

Whether that clarity is worth something — whether it’s better than a regime of vague customary norms where every violation is deniable and no dispute ever reaches a tribunal — is a genuinely open question. At least UNCLOS tells you, precisely, who is doing what to whom and what treaty provision they are violating in doing it.

The destroyer is still in the Spratly Islands chain. The frigates are still shadowing it. The 2016 ruling sits in the PCA’s record, all 501 pages. The artificial islands are still there, occupied and expanding. Both sides are logging the same encounter as either lawful or unlawful, depending on which stack of documents you consult.

Call it failure if you like. But that requires a coherent idea of what success would look like — a 320-article treaty with its own courts and the power to compel states that have decided to stop complying. No such treaty exists. This is as close as international law has gotten.

Gen AI Disclaimer

Some contents of this page were generated and/or edited with the help of a Generative AI.

Media

Guohua Song – Pexels

Key Sources and References

United Nations Convention on the Law of the Sea (UNCLOS), United Nations, opened for signature December 10, 1982, entered into force November 16, 1994. Full text via the United Nations Division for Ocean Affairs and the Law of the Sea: https://www.un.org/depts/los/convention_agreements/texts/unclos/unclos_e.pdf

Commission on the Limits of the Continental Shelf (CLCS), United Nations. Submissions and recommendations record: https://www.un.org/Depts/los/clcs_new/clcs_home.htm

Permanent Court of Arbitration, In the Matter of the South China Sea Arbitration (Republic of the Philippines v. People’s Republic of China), PCA Case No. 2013-19, Award, July 12, 2016. Official record: https://pca-cpa.org/en/cases/7/

Ministry of Foreign Affairs of the People’s Republic of China, “Statement of the Government of the People’s Republic of China on China’s Territorial Sovereignty and Maritime Rights and Interests in the South China Sea,” July 12, 2016: https://www.fmprc.gov.cn/nanhai/eng/snhwtlcwj_1/201607/t20160712_8527297.htm

Ministry of Foreign Affairs of the People’s Republic of China, spokesperson Mao Ning statement on Okinotorishima, May 27, 2025. Regular press conference transcript: https://www.mfa.gov.cn/eng/xw/fyrbt/202505/t20250527_11634354.html

International Tribunal for the Law of the Sea (ITLOS), List of Cases, Hamburg, as of early 2025: https://www.itlos.org/en/main/cases/list-of-cases/

Wright, Thomas J., “Outlaw of the Sea: The Senate Republicans’ UNCLOS Blunder,” Brookings Institution, August 7, 2012: https://www.brookings.edu/articles/outlaw-of-the-sea-the-senate-republicans-unclos-blunder/

Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States of America), International Court of Justice, Judgment of 27 June 1986: https://www.icj-cij.org/case/70

US Department of Defense, Annual Freedom of Navigation (FON) Report. Published annually. Reports and program documentation available via the Under Secretary of Defense for Policy at: https://policy.defense.gov/OUSDP-Offices/FON/

US Department of State, “Continental Shelf and Maritime Boundaries; Notice of Limits,” Federal Register Vol. 88, No. 244, p. 88470, December 21, 2023 (Public Notice 12244). Official government record of the December 19, 2023 announcement of U.S. extended continental shelf outer limits: https://www.federalregister.gov/documents/2023/12/21/2023-28159/continental-shelf-and-maritime-boundaries-notice-of-limits

Treaty between the Kingdom of Norway and the Russian Federation concerning Maritime Delimitation and Cooperation in the Barents Sea and the Arctic Ocean, September 15, 2010 (entered into force July 7, 2011). Treaty text via United Nations Division for Ocean Affairs and the Law of the Sea: https://www.un.org/depts/los/LEGISLATIONANDTREATIES/PDFFILES/TREATIES/NOR-RUS2010.PDF

Pardo, Arvid, Statement to the First Committee of the United Nations General Assembly, October–November 1967. Full text via the United Nations Division for Ocean Affairs and the Law of the Sea: https://www.un.org/depts/los/convention_agreements/texts/pardo_ga1967.pdf

Herbert Smith Freehills Kramer (formerly Herbert Smith Freehills), “A Rock or an Island? The Significance of Okinotorishima and Its Status Under the International Law of the Sea,” July 2012: https://www.hsfkramer.com/notes/arbitration/2012-07/a-rock-or-an-island-the-significance-of-okinotorishima-and-its-status-under-the-international-law-of-the-sea (access may require firm registration)

Center for Strategic and International Studies, Asia Maritime Transparency Initiative (CSIS AMTI), “China’s Big Three Near Completion,” March 27, 2017 (updated June 29, 2017). Documents military-grade airstrips, port facilities, and radar installations at Fiery Cross Reef, Subi Reef, and Mischief Reef: https://amti.csis.org/chinas-big-three-near-completion/

Dreyer, June Teufel, “The Curious Case of Okinotori: Reef, Rock, or Island?” Pacific Forum CSIS, PacNet #59, July 18, 2016: https://pacforum.org/publications/pacnet-59-the-curious-case-of-okinotori-reef-rock-or-island/