Deaths today by age: an age-by-age memorial index

On the first day, the age in a notice is doing something quite specific. It is not statistics and it is not biography. It is identification. In a same-day report, the age is often the only thing separating the person being written about from everyone else who shares their name.

That is a heavy load for a number that, on day one, is usually the least verified thing on the page.

Nobody is named on this page and no death is reported. The subject is the field, not anybody it has ever described.

What to take away

• The commonest discrepancy in any fresh report is exactly one year, and it usually has nothing to do with anyone being wrong about the person.
• An age-sorted view of one day's reported deaths looks like data and behaves like noise.
• Among corrections requested on memorial notices, an age or a date being wrong is one of the most common, and one of the most keenly felt.
• Prefer a full date of birth over an age; the date can regenerate the age, and the age cannot regenerate the date.

Where a first-day age comes from

Three routes, in rough order of frequency:

The family supplies it, through a funeral home or in a statement, from memory, in the hours after a death. Almost always right. Occasionally out by a year, for reasons nobody involved would notice at the time.

Someone calculates it from a birth year that was already in a file: a profile, a database entry, an old cutting. This inherits every error in the source and adds one of its own, because subtracting years gets the answer wrong for anyone who had not yet reached their birthday that year.

Someone reuses a published figure, which is the same as the previous route with an extra copy in between. If the earlier figure was stale, it stays stale.

None of these is a document. The document, the civil registration of the death, exists on a different schedule, is obtained from the registering jurisdiction on its own terms, as the federal guide to getting a certified death certificate sets out for one country, and in most places is not available to a publication on the day. That is why a same-day age is a reported figure rather than a confirmed one.

The one-year problem

The commonest discrepancy in any fresh report is exactly one year, and it usually has nothing to do with anyone being wrong about the person.

Two publications describing the same death can print ages a year apart because one had a full birth date and the other had only a birth year. Someone who died shortly before their birthday is a year younger than year-minus-year arithmetic suggests. Nobody made an error of fact; they made different arithmetic from different inputs.

Because of this, a one-year difference between early reports is weak evidence of anything. What it does tell you is that at least one of the outlets is working from a year rather than a date, which is useful to know before you copy either figure.

Why sorting a single day by age tells you nothing

An age-sorted view of one day's reported deaths looks like data and behaves like noise.

What is reported on any given day depends on when families chose to announce, which institutions issued statements that morning, which deaths happened somewhere with fast reporting, and what else was competing for space. A day is a tiny, self-selecting sample shaped by editorial and administrative timing. Its age distribution changes completely from one day to the next for reasons entirely unconnected to how people die.

Grouped over a year the same field starts to say something about coverage (still not about mortality, but about who gets written about), and that is taken up in age across a full year. Over a single day it does not even support that. For anything resembling mortality figures, national statistical offices publish them from complete registration, and those returns are gathered internationally in the WHO Mortality Database.

What a same-day age is genuinely good for is the narrow thing it was there for: telling a reader whether the person in the report is the person they were thinking of.

The detail families ask to have fixed

Among corrections requested on memorial notices, an age or a date being wrong is one of the most common, and one of the most keenly felt. It seems trivial from the outside: a year either way, in a piece that got everything else right. It does not feel trivial to the people who knew exactly how old he was, and who are now looking at a public record that does not.

If you are the person who spotted it, the request is easy to make well. Quote the sentence as published, give the correct figure, and give the birth date if you have it, since a date settles the arithmetic in a way an age cannot. Send it to the publication, and to the funeral home if the notice was placed through them, because they hold what was originally submitted.

If you are the person who wrote it, fix it and say that you did. An age correction is small enough that the temptation is to change the number quietly. The dated note underneath costs nothing and is the difference between a record and a moving target.

Handling the field carefully on day one

• Prefer a full date of birth over an age; the date can regenerate the age, and the age cannot regenerate the date.
• Say where the figure came from if it was not supplied by the family: "according to" is doing real work in a first-day report.
• If you have only a birth year, either say so or omit the age. A calculated age presented as a fact is a guess with the uncertainty stripped off.
• Never use the age alone to decide that two records refer to the same person.
• Leave it out rather than estimate it. A missing age is a gap; an invented one is an error that gets copied.

An age is also the smallest possible summary of a life, and a notice that leans on it is usually a notice that has not yet been able to say anything else. That is entirely reasonable in the first hours. It is worth remembering that the number is a placeholder for the things that will be said later, and not a measure of anything.

Where this sits

The parent index for same-day reports is deaths today. The same field read across a whole year is covered in deaths by year, and the intermediate unit in deaths by month. For how a correction to a published notice is asked for and handled, see deaths this week.

Common questions

Two reports give ages a year apart. Has somebody made a mistake?

Usually not. One had a full birth date and the other only a birth year, and the person had not yet reached their birthday. It is arithmetic, not error.

Should a first-day report give an age at all?

It is useful for identification, which is what it is there for. Where only a birth year is known, say so or leave it out.

How do I get an age corrected in a notice?

Quote the sentence as published, give the correct figure and the birth date if you have it, and send it to the publication and to the funeral home that placed the notice.

Why does a corrected age need a dated note?

Because the number has already been copied. A quiet change leaves every copy standing with nothing to show it was ever questioned.

Can I match two records on age alone?

No. Ages collide constantly, and a first-day age is one of the least verified fields on the page.