Core 1 asks you to explain why people move and what happens where they arrive. Most textbook examples are decades old. This one is being counted while you read it, which makes it the best kind of case study and the most dangerous kind to quote carelessly.
What happened, in order
Yemen has been at war since 2015. What changed this month is the front line. On 3 September 2026 Houthi forces launched an offensive against the internationally recognised government in Ta’iz Governorate, then pushed along the Red Sea plain toward the port of Mokha and the approaches to the Bab al-Mandeb strait. Government forces, backed by Saudi air support, counterattacked in western Ta’iz. The Mokha to Ta’iz road, the main supply route into the city, was cut.
The geography of the offensive explains the geography of the flight. People did not leave randomly. They left the districts the fighting crossed, and they moved to the nearest places that were not on fire.
Where they went
Read this table as a distribution, not a list. Almost everyone stayed inside Yemen, and most stayed inside or next to the governorate they fled.
| Governorate | Displaced people, 21 Sep 2026 | What that tells you |
|---|---|---|
| Ta’iz | 73,494 | Where the fighting is. Also where most of the displaced are hosted, which is why the camps are so close to the front. |
| Lahj | 33,828 | The next governorate along the coast. Short distance moves dominate. |
| Aden | 8,946 | The government seat and the largest urban destination. |
| Al Hodeidah | 8,244 | North along the same Red Sea plain. |
| Abyan | 3,324 | East of Aden. |
| Hadramawt | 1,008 | Far east. Only those with money or family reach this far. |
| Ma’rib | 594 | Inland, and already hosting displaced people from earlier phases of the war. |
At district level the concentration is sharper still. Jabal Habashi in Ta’iz is hosting more than 23,000 people, Al Madaribah Wa Al Aarah in Lahj nearly 21,500, Al Ma’afer in Ta’iz more than 16,500 and Ash Shamayatayn more than 12,300. Othman Belbeisi of IOM noted that many of these families have already moved more than once
. Repeat displacement is the detail that turns a number into a level of desperation.
Two counts of the same crisis
The camp management authority in Ta’iz told Al Jazeera that 65,845 people were displaced in that governorate between 10 August and 15 September, and that about 95 percent came from five districts: Jabal Habashi, al-Maafer, Maqbana, al-Wazi’iyah and Mawza. IOM’s figure for Ta’iz on 16 September was above 63,000.
There is a third number. On 18 September, Reuters reported IOM saying at least 112,000 people had been displaced inside Yemen in the previous two weeks. Three days later IOM’s own release quoted 118,086 for that same date.
| Source | Figure | What it actually counts |
|---|---|---|
| IOM DTM, 21 Sep | 129,438 | Individuals in 21,573 households, seven governorates, the full escalation to date. |
| IOM via Reuters, 18 Sep | At least 112,000 | A rounded minimum given at a Geneva press briefing, framed as the last two weeks. |
| IOM DTM, 18 Sep | 118,086 | The dataset figure for the same day, published later in the 21 September release. |
| Ta’iz camp authority | 65,845 in Ta’iz | One governorate, counted by local officials from 10 August, a longer window than IOM’s. |
None of these is a lie. They differ because they cover different start dates, different definitions of a displaced person, and different areas an enumerator could actually reach. A press briefing rounds down and hedges with “at least”; a dataset publishes to the individual; a local authority counts its own governorate on its own timeline. IOM itself states that it cannot report on some governorates because of access limits. When a resource booklet hands you a figure with a source attached, this is the kind of question the examiner wants you asking.
The route that reversed
This is the part a geographer should not miss. The Eastern Route normally runs the other way: Ethiopians and Somalis cross from Djibouti and Somalia into Yemen, then head overland toward Saudi Arabia looking for work. In March 2026 alone IOM tracked 17,027 migrants entering Yemen, 74 percent of them departing from Djibouti, and 86 percent of them Ethiopian.
Since early September, boats have been carrying people the other way. By the evening of 20 September, 3,106 people had crossed from Yemen to Djibouti, landing mostly around Obock, an arid district with little water and summer temperatures that routinely pass 45°C. A corridor built by smugglers to carry labour migrants north is now carrying refugees south. Same sea, same boats, opposite direction, different legal status on arrival.
Most of those crossing are Yemeni families, though IOM says a small number of Ethiopian migrants are now travelling in that direction too, giving up on the journey they had started. Authorities in Somaliland and Puntland report increased arrivals from Yemen as well, still in the hundreds.
What this does to the receiving country
Djibouti is the other half of the story and the half students usually forget. Sandrine Desamours of UNHCR told reporters in Geneva on 18 September that the government of Djibouti is preparing for up to 10,000 refugees, and that health care and education services will take the strain.
Put that figure in proportion. Djibouti has a population of roughly one million and imports most of its food. Ten thousand arrivals there is a demographic event in a way that the same number arriving in Aden is not. This is what “impact on host communities” means in the syllabus: not a general statement about pressure, but a ratio between arrivals and the capacity of the place receiving them.
129,438 people displaced inside Yemen. Up to 10,000 expected in Djibouti. The internal flow is more than twelve times larger, yet the external one is the one that gets international coverage and a refugee framework. Explain that gap and you are doing geography rather than reporting it.
The baseline this lands on
Everything above is the new layer. Underneath it sits a country where, before any of this started, the UN counted 22.3 million people needing humanitarian assistance in 2026, including 5.2 million already displaced, and 18.3 million people acutely food insecure. The 2026 response plan asked for 2.16 billion US dollars on 18 March. By May it was 12.7 percent funded.
“129,438 people displaced” is a September 2026 figure for this escalation only. It is not Yemen’s displaced population, which is above five million. Mixing the two is the most common way students lose marks on a crisis like this one.
Where it connects in this course
Forced migration with a named push factor, counted flows, identifiable destinations and pressure on host communities. The whole sub topic in one event.
Food insecurity at 18.3 million people, water systems already failing, and an appeal funded in single figures for most of the year.
District, governorate, national, regional. The answer changes at every level, and the district level is where the crowding actually happens.
Who controls a road decides who eats. The Mokha to Ta’iz route is a supply line before it is a line on a map.
Bab al-Mandeb again. The same chokepoint logic as the Hormuz story, with displacement as the human cost of contesting it.
A number that changes every three days. Practise citing a figure with its date and its agency, then updating it without rewriting your argument.
For Monday
- IOM recorded 118,086 displaced on 18 September and 129,438 on 21 September. Explain what a rising count tells you about the conflict, and what it tells you about the counting system. These are two different answers.
- Ta’iz is both the governorate people are fleeing and the governorate hosting the most displaced people. Explain how both can be true, using named districts.
- Only 1,008 people reached Hadramawt and 594 reached Ma’rib. What does distance decay explain here, and what does it fail to explain?
- The local camp authority and IOM give different totals for Ta’iz. Set out three legitimate reasons for the gap, then say which figure you would cite in an exam answer and why.
- A migration corridor that normally carries Ethiopians into Yemen is now carrying people out of Yemen. Explain what changes about a person’s legal status and their rights when the direction reverses.
- Djibouti is planning for up to 10,000 refugees; Yemen has displaced more than twelve times that internally. Explain why the smaller flow may be the harder one for the receiving country to absorb.
- Reuters reported at least 112,000 displaced on 18 September; IOM’s own dataset gives 118,086 for the same day. Explain why a press briefing and a dataset would differ, and why “at least” is doing real work in that sentence.
- Yemen already had 5.2 million displaced people before September. Explain why adding 129,438 to a system that is 12.7 percent funded is not a proportional increase in suffering.
- Compare this with the Iran and Hormuz story on this site. Both involve a maritime chokepoint. Identify what each one shows that the other does not.
Sources
- ReutersYemen conflict displaces 112,000 people inside the country — 18 September 2026. The Geneva briefing: the rounded IOM figure, the UNHCR planning number for Djibouti, and the arrivals reported in Somaliland and Puntland.
- IOMNew IOM data: nearly 130,000 displaced in Yemen — 21 September 2026. The governorate and district breakdown, the three day rate of change, and the Djibouti crossing figure.
- IOM YemenYemen displaced top 104,000 as IOM calls for urgent action — 16 September 2026. The earlier point in the series, useful for showing the rate rather than the total.
- Al JazeeraTa’iz camps overflow with Yemenis fleeing fighting — 16 September 2026. The local authority count and the five districts. Read it next to the IOM release and note what differs.
- Al JazeeraSaudi Houthi fighting in Yemen escalates — 8 September 2026. Background on the fronts and the Bab al-Mandeb dimension.
- CNNThe next phase of the Houthi campaign — 21 September 2026. Analyst assessment of the Ta’iz fronts and the supply routes.
- OCHAAid agencies appeal for 2.16 billion dollars for 2026 — 18 March 2026. The baseline figures: people in need, existing displacement, food insecurity.
- IOM DTMDisplacement Tracking Matrix, Yemen — The live data source behind every IOM figure here, including the access limitations note. Go here for an updated number before you cite this page.
Current as of 22 September 2026. IOM is publishing new totals every few days, so every figure on this page should be checked against DTM before you use it. Expect the numbers to be higher, and expect some of the early district figures to be revised.