In Colombia this week, the president put the number of missing at around 195. Civilian databases put it near 4,000. Both figures were published on the same day, by people acting in good faith.

This happens after every large disaster, and understanding why makes the coverage far easier to read.

Counting the dead

A death toll sounds like a simple count. It is a count of a specific thing, and different bodies count different things.

Some figures include only bodies recovered and formally identified. Some include bodies recovered but not yet identified. Some include people confirmed dead by hospitals, which captures those who died after being rescued. Some later include deaths indirectly caused — from untreated illness when hospitals were damaged, or from conditions in temporary shelter.

A national emergency agency, a city government, a police force and a news organisation compiling its own tally will therefore produce different numbers at the same moment, all of them defensible.

Why the missing are harder still

Missing-person counts are the least reliable figure published after a disaster, and they routinely differ by an order of magnitude.

Official lists generally require a formal report, often with identity documents and a verified last known location. That filters out duplicates and rumour. It also excludes anyone whose family has not reached an official — which after a disaster means anyone without phone signal, transport, documents, or the confidence to approach the authorities.

Civilian databases accept submissions directly. They are faster and they reach people the official system misses. They also double-count — the same person reported by three relatives under three spellings — and they rarely remove names once someone turns up safe.

Both are useful. Neither is the number.

Most of the missing are alive

This is the part most worth knowing, because it is genuinely reassuring and almost never stated.

In large disasters, the substantial majority of people reported missing turn out to be alive. They evacuated to a relative’s house, they were admitted to a hospital under a misspelled name, their phone was lost or without power, or they simply had no way to contact anyone for several days.

Missing lists shrink dramatically as communications are restored. A reader who treats a high missing figure as a forecast of the eventual death toll will almost always be wrong.

Why officials are cautious

Governments are often accused of understating tolls, and sometimes that accusation is correct. But there are also reasons for caution that have nothing to do with concealment.

An official figure carries legal weight — for death certificates, insurance, inheritance and compensation. Declaring someone dead who is alive causes serious harm. And a figure that jumps up and down as it is corrected damages trust in everything else the authorities say.

The trade-off is real: accuracy costs time, and speed costs accuracy. There is no option that avoids both.

How to read the numbers

Three habits help. Note who is publishing a figure and what they count. Expect death tolls to rise for days and missing counts to fall. And treat any single number in the first week as provisional, whoever issued it.

The figure that eventually matters most is usually published months later and gets almost no coverage: the excess mortality estimate, which captures deaths the immediate count never reached.

---

Prompted by the discrepancy in missing-person figures reported after the Colombia earthquake of 10 August 2026.