* Census Dates

Got general Family History research questions - this is the place
avatar
arthurk
Famous
Posts: 199
Joined: 31 Jan 2015 20:24
Family Historian: V6.2

Census Dates

Post by arthurk » 05 Aug 2017 13:05

In the course of this topic Adding census dates day and month (15153) on the General Usage board, the question has arisen of which date to actually record for a census. This is something I've been wondering about recently, and since it's not just about using FH, I thought I'd start a topic here.

The first family history text book that I bought was Terrick Fitzhugh's Dictionary of Genealogy (3rd edition). His article on the census includes this:
The Census of 1851 was taken on 31 March for the night of 30/31 March...
The Census of 1861 was taken on 8 April for the night of 7/8 April.
This is consistent with the headings above the Names box on the English 1841 and 1851 censuses, which refer to those who "abode the night of ..." (or "the preceding night"): the past tense implies that the information was recorded after the night in question.

A similar concept is stated more explicitly in the instructions to enumerators in the Scottish 1871 census:
You should be very careful that no person alive* at midnight, dwelling in the house or lodgings on the night of April 2nd, is omitted from the Schedule; and that no inmate who was then absent is inserted, except those travelling or out at work during that night, and who return home on Monday morning, April 3rd, all of whom must be entered in the Schedule. (Section II 3d relating to duties on Monday 3 April) (found at http://www.rootsweb.ancestry.com/~sctfc ... enumInst71)
(I couldn't find what the asterisk referred to, but the rest of the instruction seems clear.)

To me at least it seems logical that a census relating to a particular Sunday/Monday night (as UK censuses are) should be taken on the Monday, recording what was actually the case (except for reasons of work or travel - Scotland 1871), and it should therefore be dated as the Monday rather than the Sunday, and this is how I've always recorded them. Surely on the Sunday, the best you could have would be an expectation of who would be there, and that they wouldn't die before midnight.

However, others apparently think otherwise. The National Archives guide gives Sunday dates (http://www.nationalarchives.gov.uk/help ... uses-taken), as does that from the Society of Genealogists (http://www.sog.org.uk/learn/help-gettin ... uide-four/).

What do others think? Is there any hope of consensus?

User avatar
LornaCraig
Megastar
Posts: 1536
Joined: 11 Jan 2005 17:36
Family Historian: V6.2
Location: Oxfordshire, UK

Re: Census Dates

Post by LornaCraig » 05 Aug 2017 13:37

At least in the UK the census information was recorded fairly promptly. I have found several cases of USA and Canadian censuses where, because of the delay in recording information, it is unreliable. In some cases people simply forgot who had been where on the night in question, and in some cases they simply told the emumerator who was 'under the roof' on the day the enumerator visited.

For example, I have a case of one man who appeared in the Canadian census of 1901 twice. The cesus was supposed to be 'as at' 31 March, but the date of the enumeration was also recorded. On 19 April he was recorded in Brandon, Manitoba, but by 4 May he had returned to his family in Assiniboia Eas, NWT (later within Saskatchewan) and he was recorded there. So I have created two census events for him with different dates in 1901, but neither of them is the official census date!
Lorna

avatar
brianlummis
Famous
Posts: 174
Joined: 18 Dec 2014 11:06
Family Historian: V6.2
Location: Suffolk, England
Contact:

Re: Census Dates

Post by brianlummis » 05 Aug 2017 16:28

My understanding is that the date of the Census was laid down under the various Census Acts and they were all Sundays from 1841. According to Edward Higgs book A Clearer Sense of the Census published in 1996 by the Public Record Office (as it was known prior to becoming the National Archives), householders were issued with schedules to complete on the day of the Census as to who was in residence that night and the Enumerator was tasked with collecting them the next day and often filling in any missing information. They were later transcribed into the enumerators books which is what we see today.

I can find no evidence from official sources that the dates were anything other than Sunday although any delay in the schedules being collected did lead to those type of errors where people appeared twice. Bearing in mind the numbers involved and in the early days the illiteracy of many who were asked to complete the schedules it is not surprising that there are a few anomalies.

Brian

User avatar
AdrianBruce
Megastar
Posts: 837
Joined: 09 Aug 2003 21:02
Family Historian: V6.2
Location: South Cheshire
Contact:

Re: Census Dates

Post by AdrianBruce » 05 Aug 2017 19:22

Frankly, I thought that there was a consensus, and that it was for the Sunday.

Certainly, one can argue what single date should be applied to a cross-midnight event and to some extent the decision is arbitrary - but my own gut feeling (and experience when completing it) is that the task is done, in 99% of the cases, on the Sunday, and therefore there is a good reason for applying the Sunday date. Nobody, I'm certain, ever held on to their blank form until the next day just to be certain.

Where people are in transit over the midnight, again it is an arbitrary decision which single place is the operative one according to the rules.

So, in the nicest possible way, because there are always going to be anomalies of people appearing in two places, the whole thing is a bit arbitrary and there's probably no sense in attempting to apply logic to it....

So for me, the answer is Sunday but I'm not claiming that there's any logic behind it.
Adrian

User avatar
BillH
Megastar
Posts: 1405
Joined: 31 May 2010 03:40
Family Historian: V6.2
Location: Washington State, USA

Re: Census Dates

Post by BillH » 05 Aug 2017 22:55

In the 1800's in the US populations were pretty sparse when you left the east coast and distances are pretty vast. Enumerators often had to travel hundreds of miles to get from one populated area to another. This could take many days by horseback in some cases. This led to many people not being enumerated quickly.

In 1850 the official enumeration date was June 1 1850. I have a family that appears in three censuses that year.

2 Oct 1850 in Clark County, Missouri
29 Oct 1850 in Yuba County, California
30 Oct 1850 in Butte County, California

Note that all three were enumerated at least 4 months after the official date of the census. The family was from Missouri and moved to California in 1850.

So I have quite a few situations where I have found someone in two different places in the same census and both are correct.

avatar
arthurk
Famous
Posts: 199
Joined: 31 Jan 2015 20:24
Family Historian: V6.2

Re: Census Dates

Post by arthurk » 06 Aug 2017 10:26

I'm aware that in the US people were rarely enumerated on the official census day, but thankfully (for me) that's an issue I haven't had to deal with much.

What's bothering me is that for many years I've been following the Fitzhugh approach and recording the Monday dates, while at the same time realising that others (probably the majority) record the Sunday dates. Now, having recently come across the lists of dates provided by authoritative bodies such as TNA and the SoG, which give the Sundays, I'm concerned that I might actually have been doing it wrong.

Or, while at present I believe that the Monday dates are probably more correct, if others see my research there's a chance that they'll judge it to be sloppy (which I don't think it is) unless I explain my usage.

The mention of the Census Acts above might be pertinent - I wonder if they include a recommendation of how to cite the date?

avatar
TimTreeby
Famous
Posts: 150
Joined: 12 Sep 2003 14:56
Family Historian: V6.2
Location: Ogwell, Devon
Contact:

Re: Census Dates

Post by TimTreeby » 06 Aug 2017 11:58

Extract from 1850 Census Act to enable the 1851 Census to take place.
1851Census.PNG
1851Census.PNG (22.45 KiB) Viewed 6065 times
I would suggest that the Sunday date is correct, as that is the Date of which the information is required, not that of when it was collected.
Although that does mean I have my 1841 Census date wrong but all others correct.

Copy of form that Households would of filled in for 1851 Census.
1851Census HS.PNG
1851Census HS.PNG (116.36 KiB) Viewed 6065 times

User avatar
Valkrider
Megastar
Posts: 1064
Joined: 04 Jun 2012 19:03
Family Historian: V6.2
Location: Spain
Contact:

Re: Census Dates

Post by Valkrider » 06 Aug 2017 12:20

Arthur

I would correct the dates to the proper one if it was my data.

avatar
arthurk
Famous
Posts: 199
Joined: 31 Jan 2015 20:24
Family Historian: V6.2

Re: Census Dates

Post by arthurk » 06 Aug 2017 12:48

Thank you, Tim, that's all very helpful.

I see that the householder's form mentions only the night of Sunday 30th March, as does the extract from the legislation. Monday 31st is the day for the enumerator to visit and record the information.

That, in fact, is more or less the way I've handled US censuses: the list is of people living at an address on a certain date, even though the information may have been collected some considerable time later. And it's the "certain date" that I've recorded, though on a good day I've made a note of when it was enumerated as well.

So to be consistent, I guess I probably ought to change my UK census Monday dates to the preceding Sundays, but I'll probably give it a bit more thought, and more opinions would be welcome.

What it boils down to could be the question "what is a census?" If it's the act of enumerating/recording, then maybe the Monday dates are correct. If on the other hand it's a record of household information, then you could argue that it's not a fact or event at all, and its only place in a family history program should be as a source. Hmmmm :?

avatar
arthurk
Famous
Posts: 199
Joined: 31 Jan 2015 20:24
Family Historian: V6.2

Re: Census Dates

Post by arthurk » 06 Aug 2017 13:12

Valkrider wrote:Arthur

I would correct the dates to the proper one if it was my data.
Yes, but the whole point of my question is, which is the proper one?

(I may now be a bit nearer the answer - see my previous post. This one from Valkrider appeared while I was writing that, so sorry it looked as though I was ignoring it.)

User avatar
tatewise
Megastar
Posts: 16462
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: Census Dates

Post by tatewise » 06 Aug 2017 13:25

I know it is not very official, but most of the Wikipedia entries for each UK Census year says the "night of the" Sunday date.
Also most references use the Sunday date.
Most genealogists appear to accept the Sunday date as the nominal reference point.
However, it is understood that the details sometimes don't always apply to exactly that date, and that can be handled in Notes or possibly the Citation Entry Date.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
Valkrider
Megastar
Posts: 1064
Joined: 04 Jun 2012 19:03
Family Historian: V6.2
Location: Spain
Contact:

Re: Census Dates

Post by Valkrider » 06 Aug 2017 13:29

My take on 'the night of' we need to consider our ancestors lives certainly in the early 1800's when a large proportion were Ag Lab's would have been in bed no later than 10pm and probably even earlier and so the Sunday should be used. With migration from agriculture to cities and an appropriate change of lifestyle then by the 1911 census came round then maybe they were staying up later?

User avatar
AdrianBruce
Megastar
Posts: 837
Joined: 09 Aug 2003 21:02
Family Historian: V6.2
Location: South Cheshire
Contact:

Re: Census Dates

Post by AdrianBruce » 06 Aug 2017 16:18

arthurk wrote:... If on the other hand it's a record of household information, then you could argue that it's not a fact or event at all, and its only place in a family history program should be as a source. Hmmmm :?
Yes and no. At least, the way I do it. Sort of.

It's a record of the situation as it is on the night of the Sunday and therefore the information should be, it seems to me, dated by the Sunday - slightly arbitrary what is meant by "the night of the Sunday", as I've said, but since we usually just apply a single date, then the Sunday date seems the most logical as it's the one that is mentioned (thanks for the image of the instructions above). Therefore I end up with occupation and residence facts (say) dated by the Sunday. Except that I don't - I only put the month and year in for these, as anything else seems over pedantic. Rather like Mr Spock saying, "I calculate impact in 13.45" seconds, Captain", when it's taken him nearly 2 seconds to say it. So up to this point then the census is just a source, as it is for births, names, relationships, etc. Indeed, if the same residence / occupation fact applies over a 10y period, I amalgamate it with the previous fact in a date range month-year to month-year.

For ease and completeness of recording, I also record a census event with the Sunday date. That's really done just for ease of tracking - it's not historically significant is it? In fact if I'm writing reports (hand-written or software generated) I never mention the census except as a source.

I should point out that, partly for historical reasons, I don't use Ancestral Sources.
Adrian

User avatar
BillH
Megastar
Posts: 1405
Joined: 31 May 2010 03:40
Family Historian: V6.2
Location: Washington State, USA

Re: Census Dates

Post by BillH » 06 Aug 2017 19:03

I know a lot of folks do assume residence based on census facts. I don't create residence facts based on census records because I have quite a few examples of folks enumerated as being at a particular house when the census was enumerated who didn't actually reside there. They were only visiting or working there. So for me it is a bit misleading to have residence facts based on the census. I do create the census fact which implies only that someone was there on that date, not living there.

For me it does seem like a fact. The person was enumerated on that date at that location which is a fact. The census form is the source. Kind of like a marriage and a marriage certificate. The source implies the fact.

Just my two cents.

User avatar
tatewise
Megastar
Posts: 16462
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: Census Dates

Post by tatewise » 08 Aug 2017 09:47

An idea you might like to consider that clarifies when the Census applies is to use an Interpreted Date Phrase.
e.g.
InterpretedDatePhrase.png
InterpretedDatePhrase.png (21.81 KiB) Viewed 5913 times
That Date format satisfies such as the Lookup Missing Census Facts Plugin, but is not compatible with Ancestral Sources.

The Search and Replace Plugin can be used to convert all the existing Census Dates (N.B. month must be in full).
e.g.
SearchAndReplaceCensusDate.png
SearchAndReplaceCensusDate.png (21.84 KiB) Viewed 5913 times
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

avatar
arthurk
Famous
Posts: 199
Joined: 31 Jan 2015 20:24
Family Historian: V6.2

Re: Census Dates

Post by arthurk » 08 Aug 2017 10:55

Thanks for all the further thoughts.

Adrian - your thinking is very similar to mine regarding applying census dates to occupations, residences etc.

Bill - thanks for the reminder that an address in a census isn't necessarily the usual residence.

Mike - if (or as seems likely, when) I make the change, the Search and Replace plugin would be my weapon of choice.

However, I've played around a bit with date phrases (eg 7/8 April 1861), and don't really like the way they're presented, with the phrase always in inverted commas, and in the facts tab, a question mark at the end of the interpreted date. If you miss out the interpreted date, the facts tab shows no date at all (not ideal), and I suspect it would also play havoc with sorting facts.

For my current thinking, an ideal solution would be to allow a date such as 7/8 April 1861 to appear as entered without inverted commas, and if the database required an interpreted date, make it the 7th.

Putting "midnight" in might seem to be a way of clarifying matters, but when it's given as a precise point in time I'm always left wondering whether it means the start or the end of the day, so I'd want to expand it to "at midnight on 7/8 April". I've never seen anyone refer to a census in that way, possibly because it's getting rather lengthy and, dare I say, just as pedantic as my insistence on using Mondays rather than Sundays.

So in the long run, I suspect I'll join the mainstream and stick to Sundays.

User avatar
tatewise
Megastar
Posts: 16462
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: Census Dates

Post by tatewise » 08 Aug 2017 13:03

Arthur, if you do decide to adjust your Census dates to the Sundays, then follow the advice in Adding census dates day and month (15153) and use the Ancestral Sources 'Census Date Check' technique that is far more efficient than the Search and Replace Plugin in this special case.

A date such as 7/8 April 1861 is not allowed because it is not valid GEDCOM.
The best you could do is a date range Between 7 April 1861 and 8 April 1861.
But such dates would not be compatible with Ancestral Sources or Lookup Missing Census Facts Plugin.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

avatar
arthurk
Famous
Posts: 199
Joined: 31 Jan 2015 20:24
Family Historian: V6.2

Re: Census Dates

Post by arthurk » 08 Aug 2017 15:02

tatewise wrote:Arthur, if you do decide to adjust your Census dates to the Sundays, then follow the advice in Adding census dates day and month (15153) and use the Ancestral Sources 'Census Date Check' technique that is far more efficient than the Search and Replace Plugin in this special case.
If I used Ancestral Sources I'd certainly consider that, but as it's a one-off job, I'll probably stick to the plug-in.

In fact it was easy enough to run a census date check (and find some errors :o ) by running a fact query: eg include those with a Census fact, then exclude unless year equals 1851.
A date such as 7/8 April 1861 is not allowed because it is not valid GEDCOM.
The best you could do is a date range Between 7 April 1861 and 8 April 1861.
But such dates would not be compatible with Ancestral Sources or Lookup Missing Census Facts Plugin.
I hadn't thought about GEDCOM validity, so thanks for pointing that out. I'd already considered using Between dates, but they didn't seem right for an event that took place on a single day. From FH Help:
A range allows you to specify a date, albeit imprecisely (for example, "John Smith died sometime between 1941 and 1943"). Do not confuse this with specifying a period (see the Period tab).

If you know the year, but not the month or day (or the year and month but not the day), you do not have to use a range. It is easier to use the Date tab, and omit the missing details.
So I suppose you could dodge the issue and just use a month - except in 1901 when it was the night of 31 March-1 April...

User avatar
tatewise
Megastar
Posts: 16462
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: Census Dates

Post by tatewise » 08 Aug 2017 15:28

The advantage of Ancestral Sources is that it looks at various fields such as Place and Date with a summary of ALL Census Events in one dialogue with suggestions of which Date is appropriate depending on Country, or you can override with whatever you prefer.

The snag with Search and Replace is that you need to know what Dates you may have used for each Census and set the Search box appropriately, which may not be as easy as it sounds. Then repeat for each Cenus year, so the process needs repeating about ten times or more.

It takes next to no time to download and install AS, and if you really don't want to use it (but why not?) then it can be just as quickly uninstalled.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

avatar
uktony
Gold
Posts: 29
Joined: 20 Jan 2010 14:12
Family Historian: V6.2
Location: Australia
Contact:

Re: Census Dates

Post by uktony » 02 Jun 2019 07:12

Following this excellent thread, I wonder if it is possible for AS to add an Enumeration Date to each AS Census entry page, somehow linked to the census date. This would for the user to input.

I further wonder if the individual's age could then be linked to this second date.

Looking at some of the earler US entries ( especially1840/1850) it is obvious that

a) the ED can be some weeks or sometimes months after the census date. (Reading this thread gives me a better understanding of why!)

b) It is clear that age entries will, in all likelihood, be at the ED, rather than the census date. I even have several examples of children born after the census date who appear on the census return. It makes the fact tab look odd that they are recorded with negative ages.

It's probably not an easy thing to do, as there are undoubtedly consequences elsewhere within FH.

Thoughts and views on alternatives anyone?

User avatar
tatewise
Megastar
Posts: 16462
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: Census Dates

Post by tatewise » 02 Jun 2019 10:54

Tony, there is no easy solution to that scenario.
Even if such a Date were added to AS the problem is where to save it in the standard FH/GEDCOM data structure.

If you chose to enter such Census records with the Enumeration Date instead of its Census Date that would work in FH, but tools like the AS Census Date Check and the Lookup Missing Census Events Plugin would treat that Enumeration Date as incorrect.

Assuming the occurrences you quote are relatively rare, then add a Note to explain the negative ages.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

avatar
smoothster
Silver
Posts: 5
Joined: 06 Dec 2015 17:03
Family Historian: V6
Location: UK

Re: Census Dates

Post by smoothster » 08 Jun 2019 08:31

Just to complicate matters further: what about people travelling on overnight sleeper trains or more recently, overnight flights?

User avatar
mjashby
Superstar
Posts: 428
Joined: 23 Oct 2004 10:45
Family Historian: V6.2
Location: Yorkshire

Re: Census Dates

Post by mjashby » 08 Jun 2019 13:57

Not at all complicated. See the last example: https://census.ukdataservice.ac.uk/use- ... uses/forms and https://census.ukdataservice.ac.uk/medi ... mm-est.pdf

Of course, every Census has a different set of forms and demands, so there is limited uniformity of you are looking at historic Censuses. It depends what the specific aims of the individual Census was as all of the information was solely gathered for the purpose of collating statistical data to inform government planning. We are simply fortunate that the UK authorities chose to retain the analysis documents, which underpin the statistics, unlike the Australia goverment, which chose to destroy all of their Census documentation once the statistics had been extracted.

User avatar
tatewise
Megastar
Posts: 16462
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: Census Dates

Post by tatewise » 08 Jun 2019 18:06

I suspect smoothster was referring to the uktony scenario of USA census dates being significantly different from the enumeration dates. That problem does not arise for Census records in the UK.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
AdrianBruce
Megastar
Posts: 837
Joined: 09 Aug 2003 21:02
Family Historian: V6.2
Location: South Cheshire
Contact:

Re: Census Dates

Post by AdrianBruce » 09 Jun 2019 11:25

smoothster wrote:... what about people travelling on overnight sleeper trains or more recently, overnight flights?
You'd need to look at the actual instructions but as I remember it, there are instructions that are supposed to cover that sort of scenario, so that in the UK census, everyone should appear only once. Nevertheless, that sort of scenario is complex enough to run the risk of people not following instructions and I remember reading about a soldier who was enumerated with his original depot, travelled overnight / late on the Sunday to a new garrison, and got enumerated there as well.

People working overnight, if I recall correctly, are supposed to be enumerated at their residence, to add another possible variant.
Adrian

Post Reply