Why Yemen Will Never Be One Country
The deep history of a land that has never been a single polity, the Saudi compulsion to control it, and the Iranian hand now closing on the Red Sea.
I. A Land Older Than Arabia’s Idea of Itself
Before there was a Saudi Arabia, before there was an Islam, there was Yemen — and it was rich. The Sabaeans built the Marib dam around the eighth century BCE and irrigated a desert into an agricultural empire. The Himyarites who followed them ruled the southern peninsula for centuries. The Romans called this corner Arabia Felix, Happy Arabia, because it was green where the rest of the peninsula was stone. Frankincense and myrrh moved north out of these valleys along caravan roads that made Mecca itself a waystation on somebody else’s trade. That is the first fact worth holding: Yemen is not a periphery of Arabia. Arabia is, in a sense, a periphery of Yemen. The Gulf states are modern constructions built on hydrocarbons discovered in the 1930s. Yemen is a three-thousand-year-old civilisational zone that fell out of the money.
But it was never one polity. The Marib dam collapsed around 570 CE and the irrigation civilisation with it. In 897 CE, a descendant of the Prophet named al-Hadi ila’l-Haqq Yahya arrived in Sa’dah in the far northern mountains, invited in to arbitrate a tribal dispute, and founded a Zaydi imamate that would rule the northern highlands in one form or another for the next thousand years. Zaydism is a branch of Shia Islam, but a peculiar one — in matters of law and daily practice it sits closer to the Sunni Shafi’i school than to Iranian Twelver Shi’ism. It was a highland aristocracy of sayyids, claiming descent from the Prophet, ruling over a tribal confederacy that never fully submitted to anyone.
The coast was always a different country. The Tihama plain along the Red Sea, the port of Aden, the Hadhramaut wadis stretching east toward Oman — these were Shafi’i Sunni, mercantile, oriented outward to the Indian Ocean. Hadhrami merchants built trading dynasties in Singapore, Java, Malabar and East Africa. The highland imam’s writ and the coastal merchant’s world were two separate civilisations occupying one map.
Empire then made the division legal. The Ottomans occupied the north in the sixteenth century, left, and came back in 1849. The British seized Aden in 1839 as a coaling station on the road to India and gradually assembled a protectorate out of more than twenty sultanates, emirates and sheikhdoms in the hinterland. In 1904, London and Constantinople drew a line between their spheres — the so-called Violet Line — and that arbitrary imperial boundary became, in time, the border between North and South Yemen. It is roughly the line the country is fighting over today.
II. Two States, One Flag, No Nation
The twentieth century gave Yemen two entirely different political experiments. In the north, the Zaydi imamate survived Ottoman withdrawal and ran a deliberately sealed, medieval theocracy until 1962, when Nasserist officers overthrew it and declared the Yemen Arab Republic. What followed was the peninsula’s first great proxy war: Egypt poured tens of thousands of troops in to defend the republic, Saudi Arabia armed the royalist tribes to restore the imam, and eight years of mountain warfare bled Nasser so badly that historians call it Egypt’s Vietnam. Note the pattern, because it recurs: Riyadh’s instinct, from the very first moment, was to fund the northern highland tribes against a central Yemeni state.
In the south, the British were driven out of Aden in 1967 by Marxist guerrillas who established the People’s Democratic Republic of Yemen — the only avowedly communist state the Arab world ever produced. Soviet-aligned, secular, with a nationalised economy and a women’s code more liberal than anything in the Gulf. In 1986, the PDRY’s ruling party tore itself apart in ten days of street fighting in Aden that killed thousands, including much of its own leadership. The south’s tendency toward internal fracture predates the current war by four decades.
Unification in May 1990 was not a national awakening. It was an elite merger executed under duress: the Soviet Union was collapsing and Aden’s patron was disappearing; oil had been found near the shared border and neither side could exploit it alone. Ali Abdullah Saleh, the northern strongman, and Ali Salim al-Beidh, the southern socialist, shook hands over a deal neither side’s institutions could absorb. Two armies were merged on paper and kept separate in practice. It lasted four years.
In 1994 the south attempted secession, and Saleh crushed it — significantly, by mobilising northern tribal levies and Islamist fighters returning from Afghanistan against the “godless” southerners. The unified state that emerged was not a union. It was a conquest, followed by three decades of northern elites asset-stripping southern land, ports and oil. Every southerner under fifty has grown up inside that grievance.
Saleh himself described governing Yemen as dancing on the heads of snakes. He meant it as a boast. It was actually a diagnosis: he never built a state, he built a patronage machine that rented the loyalty of tribes, generals and clerics with oil revenue. When the oil began running out in the late 2000s, the machine had nothing left to distribute — and the snakes stopped dancing.
III. Why Unity Keeps Failing
The conventional reading is that Yemen is broken by foreign intervention. That is half true and analytically lazy. Yemen is broken because the thing being held together was assembled from parts that have never shared a single political logic.
Geography. The highlands, the Tihama coastal plain, Aden’s port hinterland, and the Hadhramaut east are four distinct economies. Mountains make insurgency cheap and central taxation expensive. Any force that takes Sanaa still does not control the wadis.
Sect and hierarchy. Roughly a third of Yemenis are Zaydi, concentrated in the northern highlands; the majority are Shafi’i Sunni, concentrated everywhere else. The thousand-year sayyid claim to rule by descent is not a theological curiosity — it is a live political claim that the Houthi movement has openly revived, and it is unacceptable to most of the country’s population.
Armed pluralism. Tribal confederations in Yemen are not folklore; they are functioning security providers with their own arsenals. The state has never held a monopoly on violence, not once in the modern era.
The southern question. The 2013–14 National Dialogue produced a six-region federal blueprint that collapsed precisely because it satisfied no one: it would have left the Houthi heartland landlocked and resource-poor, and it split the south into two units, killing any path to southern statehood. Both losers rejected it. Within months the Houthis took Sanaa.
And then there is the most recent proof of concept. From December 2025, the UAE-backed Southern Transitional Council seized much of the non-Houthi south and announced it was consolidating toward an independence referendum by 2028 — before a Saudi-backed counter-offensive in January 2026 reversed its gains almost as fast, expelled its ministers from the governing council, and forced the group’s apparent dissolution. The episode strained relations between Riyadh and Abu Dhabi, which had been backing opposing factions within the same internationally recognised government, and the UAE announced the withdrawal of its remaining forces.
The separatists have since relaunched under a new name and are back in the field as a “Southern Resistance” that fights the Houthis and ambushes government convoys in the same week. That is the shape of Yemeni politics in one paragraph. Not two sides. A lattice of armed factions whose alliances shift by district and by month.
Worth naming an important nuance the headlines flatten: the current war is not formally a war of partition — the Houthis in the north are not seeking a separate state. They want all of Yemen. It is the south that wants out. Unity fails not because everyone wants secession, but because the one faction strong enough to unify the country by force is the one the rest of the country will not live under.
IV. The Saudi Compulsion
Why does Riyadh care so much about the poorest country in the Arab world? Start with the border. Saudi Arabia shares roughly 1,800 kilometres of frontier with Yemen — its longest land border by far, running through mountains it cannot seal. And that border is itself a wound: Asir, Jizan and Najran, three provinces now firmly Saudi, were taken from Yemen in the 1934 war and formalised in the Treaty of Taif. Yemeni irredentism over those provinces is not fringe. Every Saudi strategic planner knows it.
Then demography. Yemen holds something on the order of 40 million people — comparable to or exceeding the Saudi citizen population, in a country with a warrior culture, mass conscription potential, and none of the Gulf’s dependence on imported labour. A Yemen that was unified, competently governed and economically functional would be the natural demographic and military centre of gravity of the Arabian Peninsula.
Riyadh’s policy for ninety years has been shaped by a simple preference: a Yemen weak enough to be manageable, divided enough to be purchasable, but not so collapsed that it exports chaos northward. That is an extraordinarily narrow target, and Saudi Arabia has missed it in both directions repeatedly.
Third, and increasingly decisive: the sea. Saudi Arabia’s oil exports historically ran through the Strait of Hormuz. Its Red Sea terminals at Yanbu and Jizan, fed by the east–west pipeline, exist precisely as the hedge against a Hormuz closure. But a Red Sea route is only a hedge if the Bab al-Mandab — the strait at Yemen’s southwestern tip, where Africa and Arabia nearly touch — remains open. Whoever holds Yemen’s Red Sea coast holds the emergency exit.
In 2026, that stopped being theoretical. With Hormuz closed since the war on Iran began in late February, Saudi Arabia was routing more than half its daily oil production through Red Sea terminals — until July, when Red Sea exports fell sharply after the Houthis declared a blockade on Saudi shipping. Both doors closed at once. That is the nightmare Saudi planners have war-gamed for decades, and it arrived in a single summer.
The response has been to reach outside the Gulf entirely. In August, Saudi Arabia, Turkey and Pakistan signed a NATO-style collective defence pact in Mecca, declaring that an attack on one is an attack on all — a new layer of protection beyond Riyadh’s longstanding reliance on Washington, amid growing regional doubt about whether the United States can still be counted on to defend its Gulf allies.
Those doubts were not misplaced. On 10 September, Mohammed bin Salman reportedly called Donald Trump twice urging American strikes on the Houthis, and was refused; US officials said they did not intend to intervene directly, though over a hundred American advisers were already in the kingdom providing targeting and geospatial intelligence for the Saudi air campaign. Intelligence, yes. Ownership, no.
V. Iran’s Cheap War
Here the analysis has to be precise, because both sides of the propaganda war distort it. The Houthis are not an Iranian invention. The movement emerged in 1990s Sa’dah as a Zaydi revivalist reaction to Salafi missionary expansion into their heartland and to the marginalisation of the highlands under Saleh. It fought six wars against the Yemeni state between 2004 and 2010 and received little Iranian support during that period. It is indigenous, and it would exist without Tehran.
What Iran did was recognise, after 2014, the highest-return investment in the modern Middle East. Support has increased steadily since, as part of Tehran’s regional alliance network. For the price of smuggled missile components, drone technology, training cadres and targeting intelligence — a rounding error in Iranian military spending — Tehran acquired the ability to threaten Saudi oil infrastructure, close a global shipping chokepoint, and bleed the wealthiest Arab state indefinitely.
No Iranian conscript dies. No Iranian territory is exposed. It is the highest leverage ratio in contemporary geopolitics. And in 2026 the relationship crossed a threshold from sponsorship to something closer to joint command.
In July, Reuters reported that Iran had instructed the Houthis to be ready to close the Bab al-Mandab if the United States struck Iranian power infrastructure. On 13 July, a Mahan Air flight carrying IRGC commanders and weapons attempted to land at Sanaa airport; Saudi and Yemeni government forces bombed the runway to stop it, and the Houthis declared the ceasefire dead. A week later they announced a naval blockade of Saudi Arabia and began striking Saudi tankers.
The September offensive across the Tihama was reportedly guided and armed by the IRGC and led by Quds Force officer Abdolreza Shahlaei, with hundreds of IRGC personnel reported in country. Yemeni government forces claim to have recovered documents from a captured Houthi commander’s laptop detailing a four-stage IRGC plan to take the Bab al-Mandab.
The Houthis themselves have dropped the pretence of local grievance. Their deputy information minister said the movement had joined the war to support Iran, and described coordination with Iran, Lebanon and Iraq.
Understand what this means structurally. Iran’s strategy in this war has been to hold the world’s chokepoints. Having closed Hormuz, it has now, through a proxy, taken the other end of the Arabian trade system. Two straits, one hand.
VI. September 2026: The Door Closes
The military facts of this month are stark. Fighting that began on 3 September in Taiz became the deadliest in Yemen since the 2022 truce. The Houthis rolled up the entire Yemeni Red Sea coast — Hays, Al Khawkhah, the Hanish Islands, the historic port of Mocha, then Dhubab and Perim Island at the mouth of the Bab al-Mandab, with government forces confirming their withdrawal.
These are the first changes in frontline control since 2022. Houthi positions now sit roughly 130 kilometres from Camp Lemonnier, the American base in Djibouti, raising real questions about that base’s future.
Some 125,000 Yemenis have been displaced, and more than two thousand refugees crossed to Djibouti in a single day, some rescued at sea after their boats ran out of fuel. Analysts read the timing plainly: the Houthis are exploiting Saudi weakness while Hormuz is shut.
Meanwhile the Mecca pact is being tested and found soft. Pakistan’s defence minister warned that Houthi attacks on Saudi Arabia could activate the treaty, then Islamabad clarified that no military action had been discussed but that it would act when the time came. Some analysts read the absence of Turkish or Pakistani action as a violation of the agreement; others note that the pact requires parliamentary approval in those countries, while Saudi Arabia can act by royal decree.
A mutual defence treaty that no signatory wants to trigger is a press release with letterhead.
VII. Why This Matters From Lagos
Africans should stop reading this as a distant Arab quarrel. The Bab al-Mandab is an African strait as much as an Arabian one — Djibouti, Eritrea and Somalia sit on the far shore, and roughly a tenth of global seaborne trade passes through it. Every container rerouted around the Cape adds cost to goods landing in Apapa and Tema. Egypt’s Suez revenue, already savaged, degrades further, with knock-on effects across a continental economy where Cairo is a major player.
There is a security dimension too. Yemeni government forces claim al-Shabaab fighters were found among Houthi casualties in the west; Houthi–al-Shabaab cooperation has been documented since at least 2023. A Red Sea corridor in which an Iranian-armed force on one shore transacts with a jihadist insurgency on the other is not a Middle Eastern story. It is a Horn of Africa story, and eventually a Sahelian one, because weapons and tactics migrate along the same routes that ideas do.
And the strategic lesson is one African states should take seriously: a weak state astride a chokepoint does not stay sovereign. It becomes real estate. Yemen’s coastline is currently being contested by Iranian officers, Saudi jets, Emirati proxies, American targeting software and Israeli offers of assistance — and almost none of those actors are asking Yemenis what they want.
Any nation that controls critical geography without controlling itself will find that others volunteer to do the controlling.
Conclusion: Partition Is Not the Failure, It Is the Description
Yemen will not be unified by this war, and probably not by the next one. The Houthis can take territory but cannot govern a Shafi’i majority that will not accept sayyid rule. The Presidential Leadership Council cannot govern because it is a committee of rival patrons held together by Saudi money. The southerners want out and are too fractured to leave. Saudi Arabia cannot conquer Yemen and cannot abandon it. Iran does not want Yemen unified — a permanently bleeding Yemen is precisely the asset.
The honest analytical position is that the unitary Yemeni state created in 1990 is already dead, and has been since 2014. What remains is a contest over which armed authorities will control which fragments, and which foreign patron will pay for each.
The international insistence on Yemen’s territorial integrity is not a policy. It is a refusal to write a new map, sustained by the fear that once one Arab border is redrawn, others will follow.
Arabia Felix is not happy. It is the pivot on which the Red Sea now turns, and it is being turned by hands that are not Yemeni.