n.V1-5.01| THE WORD WITH NO CLOCK

01| The anti-poverty industry has two halves that share a name and share almost nothing else. One responds to events — war, earthquake, drought, epidemic, an outflow of people across a border. The other works on chronic poverty over decades. Only the first is called an emergency, and that word is where the trouble starts, because an emergency implies a clock.

02| The clock is stated openly. The crisis half's declared horizon is weeks to a few years. Its financing runs on annual appeals. Its working vocabulary is surge, response, scale-up. Every structural choice — how it is funded, how it coordinates, what it counts as success — follows from the assumption that it is temporary.

03| Development Initiatives counts a country as being in protracted crisis once it has had five or more consecutive years of UN-coordinated appeals. In 2021 there were thirty-six such countries, up from thirty-four in 2020, with a further twenty in recurrent crisis. That is where the people are:

"Most people in need of humanitarian assistance live in countries experiencing protracted crisis: the 36 countries experiencing protracted crisis in 2021 accounted for 74% (227.3 million) of total people in need."— Development Initiatives, Global Humanitarian Assistance Report 2022

04| And five years is the threshold, not the average. Dadaab opened in Kenya between October 1991 and June 1992. Kakuma opened in 1992. Both are still running, still under UNHCR management, still holding substantial populations. Kenya has announced closure repeatedly and not closed them.

05| Set thirty years against a stated horizon of weeks to a few years. That is not a rounding error. It is one to two orders of magnitude. The standard operation is a long-running public service, financed by an annual emergency appeal, in a country where the emergency framing has been in place for a decade or more. The appeal cycle is annual. The operation is generational.

06| The machinery itself is real and works. Eleven sectoral clusters with designated UN leads. A Humanitarian Coordinator in each crisis country chairing a team that must produce a needs overview and a response plan every year. Sphere sets minimum litres of water per person per day. The IPC decides when a food situation counts as famine, and that classification unlocks money. WFP moves over four million tonnes of food a year. UNHAS flies to more than 350 destinations. UNICEF distributes over two billion vaccine doses annually, including in active conflict. When surveyed, affected populations criticise how assistance is delivered and do not reject the assistance.

07| Now look at the reform history, because it explains what the machinery has and has not been built to do. The cluster system came after uncoordinated response to the 2004 tsunami produced duplication and gaps in front of the global press. The Grand Bargain came after the 2016 World Humanitarian Summit. Anticipatory action frameworks came after the slow responses of the 2010s. Every one of those reforms addressed a failure to start fast enough or coordinate well enough. Not one of them addressed a failure to stop.

08| That is not an oversight. It is a record of what the system has been publicly shamed for. Nobody has ever been shamed on television for staying.

09| There are three ways an operation ends, and only one is designed. The designed one is deactivation, and the IASC guidance sets two criteria:

"(i) The humanitarian situation improves, significantly reducing humanitarian needs and associated response and coordination gaps and/or (ii) national structures acquire sufficient capacity to coordinate and meet residual humanitarian needs in line with humanitarian principles."— IASC, Guidance on Cluster Transition and Deactivation

10| Read those carefully. Both are conditions about the state of the world. Neither is a deadline, a maximum duration, or a scheduled review that forces anyone to decide anything. Nothing in the text obliges anyone to build the national capacity the second criterion requires. The capacity either appears or it does not, and until it appears the cluster stays activated. An operation can satisfy no exit criterion for twenty consecutive years without breaking a single rule.

11| The second way out is not ending, which is the default described above.

12| The third is collapse, and it is the most common way a large operation actually stops. Humanitarian financing is extraordinarily concentrated — the United States, Germany and the EU institutions together provide more than sixty percent of it in most years, the top ten donors more than eighty-five. When that pool moved in 2025, the effect was immediate: layoffs, country offices closed, programmes suspended mid-cycle. No other donor can substitute at that scale on short notice.

13| Underfunding is the normal condition even without a shock. The 2024 appeal asked for around 49 billion dollars across 72 plans and received about 21 billion — forty-three percent. The industry consistently raises less than half of what it says it needs, and the shortfall is not spread evenly. Ukraine's 2022 appeal was funded above one hundred percent. Sudan, holding the largest displacement crisis in the world, sat near thirty percent in 2024.

14| So the realistic alternative to planned handover is not permanence. It is an unplanned stop — a programme that ran for a decade with no transition design, then halted because a donor changed policy or attention moved elsewhere. No receiving institution, because none was prepared. No residual capacity, because the guidance that would have required building it was never triggered. The population experiences the end as an event rather than a transfer.

15| This reframes what overstaying means, and the reframing is the point of the chapter. An operation that runs fifteen years and stops in ninety days has not overstayed and handed over late. It never planned to leave, and then left involuntarily. Those are different failures and they need different names.