One hundred seventeen dead in four days, 1,790 families fleeing their homes, fifteen government airstrikes, and one strait just twenty-nine kilometers wide. Yemen’s war is back, and this time it speaks the language of global logistics rather than the struggle for Sanaa.
At around 5:30 a.m. local time on September 3, Ansar Allah assault groups advanced simultaneously along several sectors west of Taiz. Tube artillery, ballistic missiles, and attack drones struck ahead of the infantry. The axes of advance were Maqbanah and Jabal Habashi, districts virtually unknown outside Yemen.
By the evening of September 6, medical sources on both sides of the front had counted 117 dead: 54 government troops and 63 fighters from the movement. A day earlier, Agence France-Presse had reported a different toll: 129 people, including 66 government soldiers, 62 Houthis, and one civilian. The United Nations Office for the Coordination of Humanitarian Affairs reported that roughly 1,790 families had fled their homes over just two days, September 3 and 4, while some camps for displaced people had emptied completely. Their residents had been forced to flee for a second time, and some for a third.
The numbers differ. The direction of the offensive does not.
Beyond Maqbanah, the terrain descends toward the Red Sea coast. There lies the port of Mocha, while roughly sixty kilometers farther south is Bab el-Mandeb, where the Red Sea narrows to twenty-nine kilometers and splits into two navigable channels. About eight percent of the world’s oil and, by various estimates, as much as one-tenth of global maritime cargo traffic passes through this funnel.
Hans Grundberg, the United Nations Secretary-General’s Special Envoy for Yemen, told the Security Council on August 13 that four years of relative calm could be lost within a few weeks of escalation. At the time, the warning sounded like obligatory diplomatic alarm. Exactly three weeks later, it ceased to be a figure of speech.
The Truce That Died Twice
Formally, the truce brokered by the United Nations on April 2, 2022, expired that fall. In practice, it endured for four years. The major fronts froze, and the war was reduced to mortar exchanges and minefields.
Taiz did not kill that equilibrium.
The broader regional war did. It began on February 28, 2026, with U.S. and Israeli strikes on Iran. What followed was a chain of events in which every link appeared local, while their cumulative effect proved catastrophic.
In early July, government aircraft struck Sanaa International Airport in what Aden officially described as an attempt to prevent an Iranian plane from landing. According to available reports, the aircraft was carrying a delegation of Houthi leaders returning from the funeral of Iran’s supreme leader, Ali Khamenei. The plane landed at another airfield. On July 12, the Houthis attacked Abha International Airport in southern Saudi Arabia with drones and missiles.
On July 20, the movement declared a naval blockade of the kingdom. Its wording was uncompromising: any vessel loading or unloading cargo at Saudi ports would be prohibited. Two days later, Houthi military spokesman Yahya Saree announced ballistic missile, cruise missile, and drone attacks against two Saudi tankers, Encelia and Layla. The Saudi Press Agency confirmed that Encelia had been hit, adding that the crew was unharmed. Saree also claimed that about ten ships had turned away before reaching Saudi ports. I have seen no independent confirmation of that figure.
By September 2026, this had produced an extraordinary configuration: a nonstate armed movement had declared a maritime blockade against one of the world’s largest oil exporters and had partially enforced it.
The ground offensive around Taiz was not the beginning of the escalation. It was its fourth act.
Al-Qadha: Three Syllables That Could Black Out an Entire Province
Taiz has been half-encircled since 2015. The government holds the city, but supplies it through a single genuinely functional artery: the road running west toward the coast and Mocha.
It is precisely along this artery that the Houthis achieved the only operational success that truly matters. According to the London-based Asharq Al-Awsat on September 7, the breakthrough occurred around Al-Qadha on the Jabal Habashi front. That same day, government forces were not mounting a counteroffensive but reorganizing their defenses, stabilizing the line and preparing to dislodge the enemy from the captured heights.
The importance of this sector has little to do with square kilometers. Ibrahim Jalal, a senior researcher with the Cross-Border Conflict Evidence, Policy and Trends project, described the logic precisely: sustained progress here would rupture the Taiz-coast defensive network, isolate Mocha and Al-Khawkhah from reinforcements and supplies, and create a land approach from which pressure could be exerted on the southern Red Sea coast and, ultimately, on the Bab el-Mandeb axis.
Mocha, meanwhile, has already been under fire for weeks. On August 9, a missile and drone attack on the port killed four soldiers and three civilians and wounded fifteen people. Air defenses shot down eleven drones. On September 3, government forces reported that fourteen ballistic missiles had been fired at areas along the western coast and that an explosives-laden boat heading toward the port had been destroyed. A second such vessel was later reported destroyed near Al-Khawkhah.
Maritime kamikaze drones near Mocha are an easy detail to miss in the broader battlefield picture. They matter more than many kilometers of front line. They are the same class of weapon that brought container shipping to a halt in 2024.
Twenty-Nine Kilometers Worth More Than Sanaa
A conventional civil war is fought for the capital. Yemen’s is fought for the shipping lane.
The Houthis have controlled Sanaa since September 2014, and eleven years of war have failed to dislodge them militarily. Over the same period, the strategic value of the coastal strip has increased many times over, and the numbers show just how much.
From November 2023 through January 2025, researchers calculated that the Houthis attacked more than one hundred commercial vessels. The consequences for the Suez Canal were devastating. Canal revenue plunged from a record $9.4 billion in fiscal year 2022–23 to roughly $5 billion in 2024, a 47 percent decline that cost Egypt about $4.4 billion in foreign-currency revenue in a single year.
A recovery followed, although estimates differ depending on methodology. Suez Canal Authority Chairman Osama Rabie said on September 6 that ship traffic in the second half of 2026 had increased by 40 percent and dollar revenues by 50 percent. He projected revenues of $5.8 billion to $6 billion by the end of December, compared with approximately $4.1 billion a year earlier. Independent industry estimates calculated on a calendar-year basis and incorporating the full second half of 2025 put 2025 revenue at around $7 billion and had forecast more than $8.5 billion for 2026. The discrepancy stems from Cairo calculating by fiscal year while analysts use the calendar year. Both methodologies agree on one point.
The canal has not returned to its prewar operating pattern.
The figure I consider most revealing came from Rabie himself: today, between 14,000 and 15,000 ships transit the Suez Canal, compared with roughly 26,000 in 2023. Nearly half of maritime traffic is still routing around the Cape of Good Hope, adding seven to fourteen days to voyages between Asia and Europe and thousands of miles of additional fuel consumption.
The Houthis produced this outcome without sinking much of anything. It was enough to make the passage sufficiently dangerous for insurers to rewrite their rates and shipowners to rewrite their schedules. In the twenty-first century, controlling a strait does not require a navy. Mobile missile launchers, drones, explosive boats, and the right point on the map can be enough.
They already have that point. The push toward Mocha is an attempt to widen it and move it closer to the water.
The Mines of August: How It Happened Forty-Two Years Ago
The idea of cheaply disrupting the Red Sea was not invented in Saada.
Between July 9 and August 15, 1984, around twenty commercial vessels flying Soviet, Japanese, Chinese, Polish, Greek, Turkish, and North Korean flags struck mines in the Gulf of Suez and the southern Red Sea, including Bab el-Mandeb itself. A CIA memorandum dated August 28 of that year and declassified in 2010 recorded nineteen damaged vessels and cautiously pointed to a Libyan ship that had been operating in the area in July, while stressing that the evidence was inconclusive. The pattern of damage indicated modern mines with relatively small warheads. An underground group calling itself Al-Jihad claimed to have laid 190 mines.
Not a single ship sank.
The response was vastly more expensive. The United States launched Operation Intense Look, deploying three ships, eight helicopters, more than 1,500 personnel, and the first combat employment in history of Helicopter Mine Countermeasures Squadron 14. Britain launched Operation Harling. French, Italian, Dutch, and Soviet forces joined the minesweeping effort. At its peak, twenty-six ships from six foreign countries were operating in the Gulf of Suez and the Red Sea. Insurance premiums rose immediately, even though traffic through Suez continued almost normally.
One vessel carrying mines against six navies working through December. The cost asymmetry on which the entire Houthi strategy now rests was demonstrated in these same waters when Abdul-Malik al-Houthi was five years old.
There is one difference from 1984, and it makes today more dangerous. Back then, no actor permanently controlled the coast. A shipping lane could be mined once, covertly, through intermediaries. Today, the shoreline is controlled by an armed movement with its own intelligence and production capabilities, able to regenerate the threat indefinitely. An advance toward Mocha is an attempt to extend that coastline by several dozen additional kilometers.
The Pincers: Hormuz in the East, Bab el-Mandeb in the West
I have followed the Strait of Hormuz story since March, and the central lesson of these months is simple: a strait does not have to be completely closed to crash a market.
Hormuz is in a condition that maritime intelligence company Windward aptly described not as reopening but as a monitored pause: coordination with Iranian armed forces is required for every transit. A Congressional Research Service report issued in early August noted that intermittent Iranian attacks on shipping and retaliatory U.S. strikes had seriously disrupted traffic through the strait for much of the previous five months.
The market reacted accordingly. On March 12, Brent crude broke above $100 a barrel. The June 17 U.S.-Iran memorandum on the passage of commercial vessels collapsed almost immediately. On August 10, after Iranian Foreign Ministry spokesman Esmail Baghaei said there were no conditions for reopening Hormuz as long as the U.S. naval blockade remained in place, WTI jumped roughly five percent to $82.13, while Brent rose to $87.72. In its scenarios, Goldman Sachs allowed for an average Brent price above $100 throughout 2026 and as high as $120 in the third quarter if the closure became prolonged.
Now superimpose the geography of the Arabian Peninsula on that picture.
Hormuz is the eastern exit. Bab el-Mandeb is the western one. When the eastern outlet operates under monitored conditions, Riyadh shifts export pressure westward: through the East-West Pipeline to Yanbu, then into the Red Sea, and onward through Bab el-Mandeb. That is precisely the western route the Houthis declared closed on July 20.
The symmetry is too precise to dismiss as coincidence. There is insufficient evidence to claim that every Houthi move is operationally coordinated with Tehran. The movement has its own agency and its own grievances against Riyadh. Yet the strategic convergence of interests is beyond dispute. On August 20, Islamic Revolutionary Guard Corps representative Hossein Mohebi told Defa Press that Saudi Arabia would certainly be unable either to contain Yemen and Ansar Allah or to remain safe from their strikes.
For Iran, whose network of allies in Lebanon and Iraq has suffered severe losses over the past year and a half, the Yemeni lever remains the cheapest and most effective. Building its own naval infrastructure three thousand kilometers from Bandar Abbas would cost Tehran incomparably more.
Aircraft Have Returned to Yemen’s Skies, and That Is More Dangerous Than It Looks
On September 4, government air forces carried out at least fifteen strikes on Houthi positions, troop concentrations, and equipment in Taiz and Hodeidah.
Fifteen sorties in the Middle East of 2026 amount to a statistical rounding error. The significance lies not in the number but in the return of the instrument itself: for the first time in the current cycle, government aviation is directly targeting a ground offensive.
From there, the escalation ladder has only one next rung.
Aden’s own air force consists of a handful of operational aircraft and suffers from a chronic shortage of munitions. Any escalation of the air campaign inevitably runs into dependence on the Royal Saudi Air Force. That is why the decision to launch airstrikes should be read as a signal: commanders regarded the breakthrough at Al-Qadha as a strategic threat, not a tactical nuisance.
The Houthis understand that ladder just as well. It is reasonable to assume they are deliberately testing it.
Riyadh’s Red Line Does Not Run Through the Mountains of Taiz
Saudi Arabia entered Yemen in March 2015 expecting a short campaign. Instead, it got eight years of war, the attacks on Abqaiq, a collapse in reputation, and an armed movement that air power never succeeded in destroying. After 2022, the kingdom methodically constructed an exit strategy: the truce, normalization with Tehran under Chinese mediation in March 2023, and quiet talks with the Houthis through Oman.
Riyadh has considerable tolerance for a few kilometers of shifting front lines in the mountains of Taiz. It has none for Houthi military infrastructure directly on the water.
The reason extends beyond oil exports, and it is rarely stated openly. The entire showcase of Vision 2030 sits on the Red Sea coast: NEOM and its Trojena ski resort, scheduled to host the Asian Winter Games in 2029, the Red Sea Global resort developments, ports, and logistics zones. The kingdom is preparing to host World Expo 2030 in Riyadh and the FIFA World Cup in 2034.
An investment program on that scale is incompatible with a sea where insurers impose war-risk premiums. A missile strike on a tanker in the Red Sea hits Saudi Arabia’s balance sheet twice: first through exports, then through the cost of attracting capital to its coastal megaprojects.
That defines the threshold. As long as the fighting remains over mountain heights, Riyadh will restrict itself to intelligence, ammunition, fuel, and money. The appearance of Houthi missile crews with a direct line of sight to the shipping lane changes the calculation instantly.
A Unity That Did Not Exist Nine Months Ago
This is where the story begins that Western situation reports rarely examine, even though the current fighting makes little sense without it.
The anti-Houthi camp’s principal weakness has always been fragmentation. In the winter of 2025, that weakness nearly ended with the country breaking into three pieces.
On December 2, 2025, the Southern Transitional Council under Aidarus al-Zoubaidi launched an offensive and, within a month, seized Hadramawt, Al Mahrah, and nearly the entire territory of the former South Yemen, by some estimates roughly 52 percent of the country. Saudi Arabia accused the United Arab Emirates of backing the separatists. On December 30, the coalition struck the port of Mukalla, saying it was targeting an Emirati arms shipment unloaded from two vessels that had arrived from Fujairah with their transponders switched off.
That same day, Rashad al-Alimi, chairman of the Presidential Leadership Council, terminated the defense agreement with Abu Dhabi and gave Emirati forces twenty-four hours to leave. The UAE announced that it was withdrawing its troops “of its own volition.” On January 9, 2026, the Southern Transitional Council dissolved itself, and al-Zoubaidi left the country. On January 15, Prime Minister Salem Saleh bin Braik resigned, and former foreign minister Shaya al-Zindani took over the government.
This leads to a conclusion that changes the assessment of the entire current campaign.
The consolidation of the government camp, now described as a new strategic reality, has existed for exactly eight months. It was produced not by a Yemeni political process but by Saudi coercive arbitration over a fellow coalition partner. The Houthis are attacking a structure held together by an external brace, and one objective of the operation is to test whether that brace will hold.
The first four days produced an unexpected answer.
It held.
Hays: The Counterattack the Houthis Did Not Expect
On September 5, government and allied forces announced the complete recapture of the Hays district in southern Hodeidah. The district center had already been retaken in 2018, but its rural periphery had remained under Houthi control for years. By September 7, government units had shifted from containment to pursuit and were advancing north toward the Al-Jarrahi district.
The significance of this axis deserves separate explanation because it turns the Houthi offensive into an operational trap. Capturing Al-Jarrahi would open the road toward Houthi-held Hodeidah and sever the communications route linking the province with Ibb and Tihama. Hodeidah is the principal port serving the Houthi-controlled part of the country and the main supply channel for the north.
At the same time, on September 5, government forces advanced on the Hayfan front southeast of Taiz, placing Houthi supply routes into northwestern Lahij and southern Taiz under threat. Field sources reported dozens of prisoners, many of them minors, a detail that says more about the state of the movement’s mobilization pool than any official statement.
The balance after four days looks like this. The Houthis achieved a tactical breakthrough near Al-Qadha and threatened the Taiz-Mocha road. In exchange, they opened a front near Hodeidah, where their enemy is advancing for the first time in years.
That was not a favorable trade.
Why the Movement Wants War Now
The first level of the answer is domestic. It was articulated by a man whose position can hardly be called neutral, though that does not make his argument any weaker. Yasin Saeed Noman, Yemen’s ambassador to the United Kingdom and one of the oldest figures in southern Yemeni politics, described the offensive as an attempt to divert attention from repression, hunger, unpaid public-sector salaries, and the collapse of basic services in areas controlled by the movement. He also called for Maqbanah to be turned into a graveyard for this military project.
The formulation is harsh and rhetorical, but the mechanism is accurately described. An armed movement built around permanent mobilization explains economic collapse through blockade and aggression. Peace brings the agenda back to salaries and the quality of governance, questions far more dangerous to a revolutionary structure than enemy missiles.
The second level is military. During four years of relative calm, the movement made a technological leap: mass deployment of drones, improved ballistic missile accuracy, and the development of antiship weapons and explosive surface vessels.
The third level is regional. Yemeni Information Minister Muammar al-Eryani claimed in early September that the Houthis had built a transnational network for exchanging weapons, technology, expertise, and personnel training, including contacts with Somalia’s al-Shabaab, and were expanding operations on both sides of the Gulf of Aden. I cannot independently verify that assertion. I therefore classify it as a substantiated allegation rather than an established fact.
The Arithmetic That Does Not Add Up
Over four days, reported death tolls ranging from 50 to 129 circulated publicly. The discrepancy is not necessarily the product of bad faith. It reflects the structure of wartime reporting itself.
Al Arabiya reported at least one hundred people killed on both sides on Thursday alone. Official Yemeni media, covering the same period, spoke of fifty Houthis killed and wounded, including two field commanders. But “killed and wounded” is not comparable with “dead.” Agence France-Presse counted 129 by September 5. By September 6, medical sources on both sides converged on 117.
My assessment is that the medical-source tally of 54 government troops and 63 Houthis represents the most defensible lower bound because it was compiled from hospitals on both sides of the front and does not depend on military press offices. Anything above that requires confirmation. Lower numbers reflect the informational incentives of the parties rather than the condition of the hospitals.
The scale itself is not in doubt.
Western Yemen has not seen ground fighting this intense since 2022.
Hunger Arrives Before the Army
The war is returning to a country whose margin of resilience has been exhausted.
According to United Nations estimates, more than 21 million Yemenis require humanitarian assistance in 2026. Roughly 18.3 million face acute food insecurity, while 2.2 million children under five suffer from acute malnutrition. Around 4.8 million people are already internally displaced, many of them repeatedly. The country imports approximately ninety percent of its food.
Funding, meanwhile, has collapsed. The 2025 humanitarian response plan was only 29 percent funded. By August 2026, according to United Nations Emergency Relief Coordinator Tom Fletcher, the appeal had received barely 20 percent of the required funds. Fletcher described the consequences without diplomatic euphemism: when funding falls, food rations are cut and clinics close.
Then the mechanism Yemen has perfected begins operating almost automatically.
First come the battles. Then the roads. Then the ports.
On September 5, a missile strike hit the Hayjah al-Abd road linking Taiz with southern districts. According to the state-run Saba news agency, four civilians were killed and nine wounded, with passenger buses among the vehicles hit. On September 6, sniper fire in residential areas of Al-Qadha and Rahbah wounded a woman and three children.
A road does not need to be physically destroyed. It is enough to make people stop using it, and the region is paralyzed.
A soft blockade costs less than an artillery barrage and works faster.
Three Scenarios and Four Testable Forecasts
The first scenario is limited stabilization. The Houthis retain some of the heights, government forces retake the rest, and the front freezes. Both sides declare victory. I estimate the probability at roughly fifty percent.
The second scenario is a prolonged battle for the western coast. Mocha becomes a zone of permanent military danger, while Riyadh increases intelligence and logistical support without formally reentering the war. Roughly one chance in three.
The third scenario is regionalization. The Houthis strike Saudi energy infrastructure, the kingdom resumes an air campaign over Sanaa and Saada, and the Yemeni front merges with the Iranian one. The remaining probability belongs to this scenario, along with nearly all of the catastrophic consequences.
To make this analysis something that can be tested rather than merely read, I will record four forecasts.
The Taiz-Mocha road will not be permanently severed before November 1, 2026. The attackers lack the logistics required to hold Al-Qadha under air and artillery strikes from two directions. The Royal Saudi Air Force will not conduct openly acknowledged strikes on Sanaa before the end of October 2026 unless the Houthis hit a Saudi oil facility. The port of Hodeidah will not change hands in 2026 despite the success of government forces near Hays. The Suez Canal will not return to its 2023 traffic levels this year. Rabie’s forecast of $5.8 billion to $6 billion in revenue will prove to be a ceiling, not a floor.
If I am wrong, I will acknowledge it publicly.
The Greatest Mistake Is to Keep Treating Yemen as a Periphery
For decades, world politics treated Yemen as a humanitarian footnote: a poor country, an endless civil war, tribes, hunger, and another United Nations appeal for money.
Geography does not distinguish between rich and poor.
It distinguishes between nodes.
An armed movement from the poorest country in the Arab world forced half of the global container fleet to sail around Africa and removed billions of dollars from Egypt’s revenues. The economic power of a state and its ability to create global disruption have finally become completely detached from one another. An aircraft carrier is no longer required. Medium-range missiles, drones, explosive boats, and the right sixty kilometers of land are enough.
The paradox of September 2026 is that no one formally wants a major war. Riyadh spent four years constructing an exit from Yemen only to discover that the exit leads back inside. Aden needs economic reconstruction, not new fronts. Tehran wants to preserve its leverage, not lose it. Millions of Yemenis want to bring a decade of catastrophe to an end.
Wars of this kind begin differently.
Each side takes the next step that appears rational from its own perspective, and the sum of those rational steps is called escalation.
On September 5, Agence France-Presse photographers in Hays, sixty-eight kilometers from Taiz, photographed fighters being driven to the front in the beds of pickup trucks, rifles in hand, rocket-propelled grenade launchers slung across their backs, dust hanging in the air.
From the bed of that pickup to the shipping lane carrying a container from Shanghai to Rotterdam is less than one hundred kilometers in a straight line.
The global economy still has not understood just how close that is.