[steveBK123]

yes, thats exactly what I am arguing. that is - pointing out weird numeric improbabilities in birth dates is not damning in terms of strictly indicating pension fraud. they are unreliable as recently as 60 years ago, and certainly were even more unreliable 100+ years ago.

[vintermann]

But what's the scale of this problem? If only 1 in 100 have curiously round birthdates, that's not a problem. If 1 out of 2 have curiously round birthdates, it will still be suspicious if 100% of supercentenarians had it.

I don't quite understand why it would be a problem for birth certificates. Surely the date of birth noted on them doesn't have to be the date they're first printed?

[steveBK123]

The point of my story was that, yes, in this poor European country as recently as 1960.. people were regularly registering children weeks late, and the government would only issue birth cert where date=today. Which is why all of my aunts & uncles have a real & official birthdate which don't match.

This was happening in a place & time where there were telephones & buses. It was simply inconvenient to get to the city immediately, so people went when they had the next opportunity, and presented opportunity to "choose a birthdate" by when they appeared.

This is only 1 particular example in 1 place of weirdness of official birth dates.

Imagine areas of the world a little further back when travel would have been by foot or horse. Given that this was happening even in a somewhat developed place & time, all sorts of stuff could be happening elsewhere for random, benign, non-pension-fraud related reasons.

[Retric]

But your example doesn’t fit the pattern of days divisible by 5 so it’s irrelevant.

The 5th of January is equally likely to be on any day of the week including weekends. Simply delaying + office days etc don’t get you that pattern it only shows up if people are filling in arbitrary dates.

Further if people are filling in arbitrarily dates +/- N years or even months that’s going to dramatically increase the number of officially extremely old deaths. Aka far more people die at 100 than 107. So randomly add or subtract a few years and on official paperwork you might lose one person who actually died at 107, but then add 2 people who actually died at 100 but officially died at 107.

[tsimionescu]

> I don't quite understand why it would be a problem for birth certificates. Surely the date of birth noted on them doesn't have to be the date they're first printed?

I don't think people cared (or care even today) what is the date written on the birth certificate. And if the state really cares, they won't accept some date they're being told: using the current date is safer.

[vintermann]

Why would the state trust their citizens so little? Where I come from births were registered by (Lutheran) priests for hundreds of years, and they had no problem accepting and noting the birth date the parents provided. Why would they not trust it, why would parents lie?

[narag]

A friend named his daughter Andrea and registered her... before she was born. There was some benefits that expired at the end of the year.

At that time, there was no easy way to know the fetus' sex, so he chose female in a hunch and used a name that is female in Spanish, but male in other languages, so he could later allege an error.

She was born in January, a couple of weeks later than the certificate.