How Old Are You?

How old are you? How many hundreds of times have you had to answer this question? Through our lives it has governed when we received childhood vaccinations, when we started (and left) school, when we could first drink, get a driving licence and marry. Then, later, it dictates when we can receive our pension and possibly a free television licence. We are asked our birthdate whenever we interact with government or the health service, whenever we want to borrow money or open a savings account or get a credit card. If necessary, we can prove our age by producing a birth certificate. So we are really familiar with our birth date and have little difficulty when asked our age every ten years when the census comes around.

Simpler times

But now transport yourself back to 1851. You are a man or woman in late middle age, married with children and grandchildren and faced with the census form. How old are you? The situation 174 years ago was very different. You have had very little interaction with authority throughout your life. The little schooling you received (if any) was not dependent on your age and you started work when strong and responsible enough to hold down a job. Save money? Chance would be a fine thing! A pension? This would not be a reality for another 58 years; you expect to work until you are no longer able, like every generation before you. How old are you? You have never really had any cause to think of this. In truth you don't really know and don't really care.

When your grandchildren were born a few years ago you remember your son having to register the births and he got a birth certificate which provided a record of the date, but when you were born, the only record was the baptism register in the parish where you were born, but since moving to Manchester as a child, you have never been back there and certainly are not going to make the long journey just to get a copy of the register entry. You can't ask your parents if they can remember; they have been in the churchyard for many years. You remember that they recorded these matters in a family bible, but this passed down to your older brother and disappeared when he died a few years ago. You'll just have to guess.

Perhaps you think back and remember that the Peterloo massacre happened just before you got married and you know that was in 1819 and you would have been about 20 years old then, so perhaps you were born around 1800 so you are about 49 to 50 years old. The form does not have space to accommodate "About 50" so you simply enter "50". What does it really matter? There is nobody who will dispute this.

Where is the evidence for this?

This proposition may seem fanciful, but consider the chart below.

The data shows ages recorded in the 1851 census and is taken from the data transcribed from the Unfilmed returns for Manchester. The curves show for each age the number of persons recorded with that age; men are shown in blue, women in orange. The curves show a population which is substantially under the age of 40 but with a distribution which tails off progressively with relatively few people over the age of 80, as might be expected.

But what is immediately obvious is that far from showing a smooth distribution, there are marked peaks at ages 30, 40, 50 and 60 with a smaller one at 70 and a "ripple" at 80. Ages of 10 and 20 do not show the same peaks. This would support the proposition that older people, born before the start of civil birth registration, were uncertain about their dates of birth and tended to gravitate towards the nearest multiple of ten which they felt was appropriate to their impression of their age.

It is noticeable that the same peaks apply to both men and women but also that the peaks for women's ages are distinctly more marked than those for men. This suggests that women were even less confident of their true age than men.

Is there another explanation?

It has been suggested that the clustering of ages around the multiples of ten is a hangover from the requirement in previous censuses, in particular the census of 1841, that adult ages should be rounded down to the nearest 5 years. If this was the case, we would expect to see similar peaks at 35, 45 55 and so on, yet there is no significant peak at any of these ages. In any event, is it likely that the head of household, when filling out the household form, would remember, much less hark back to, a practice from ten years previously, when filling out a form which explicitly asks for their age? Indeed, the evidence is that in 1841 many householders ignored (or failed to understand) the rounding down rule and some enumerators carried these precise ages forward when copying into their enumeration books. Likewise, there is no evidence that enumerators rounded ages when copying the household forms into their enumeration books. Ages which are multiples of ten appear on pages where other adult ages do not. There seems little reason to attribute the observation to hangovers from the previous practice.

What does it matter?

This is an important matter for family historians. The 1851 is usually our jumping-off point for the period before civil registration when we will be dependent upon baptism registers to identify the earlier generations. The census provides an age (and by implication a year of birth/baptism) and a birthplace. We will use these to search the baptism registers for the birth place looking for our ancestor's baptism in the years indicated in the census. If he was aged 50 then we will expect to find a baptism between (broadly) April 1800 and May/June 1801. But what if we don't?

Let's say we find a baptism in 1797 with the right name. Is this our ancestor? Strict interpretation of the age in the census would say no, but if we accept the premise that by 1851 he only knew he was "about 50" then it is possible he may be. It is also possible, with a common forename, that there may be several children baptised around the right time. The one which became your ancestor may be any of these. It is beyond the scope of this article to discuss how this problem might be resolved, but clearly it exists and will have to be addressed.

We can perhaps take this line of argument further. How reliable are any of the ages of older adults? How many thought "About 48" or "about 75"? These randomly inaccurate ages will not leave evidence in the form of peaks in the distribution.

The message to the researcher is to be VERY cautious about the acceptance of ages in the 1851 census. Indeed, be cautious about ages in subsequent censuses. Our "about 50" year old, if he survives to be enumerated in 1861, 1871 or, who knows, 1881 will be no wiser as to his true age (and perhaps less so) than he was in 1851. The ages of children under 14 may be more reliable (they should have a birth certificate and in any event, parents may be fairly certain as to their year of birth) but should be treated with caution nevertheless.

Then, of course, the place of birth may not be right - but that's another story!