Using census records
to research family trees
A page from the 1901 St Helier census, including the grandfather and great-grandparents of Jerripedia editor Mike Bisson
One of the most valuable resources for family history researchers is the set of Jersey Censuses which can be accessed by Ancestry subscribers and are also found on a number of other subscription websites.
A formal census was first carried out in 1841, and subsequently every ten years. Census statistics are released soon after each census is held, but the original pages filled in by the enumerators are only released 100 years after a census was held. So, it is currently possible to research the censuses in 1841, 1851, 1861, 1871, 1881, 1891, 1901 and 1911.}}
A family's census return can provide very valuable information to help trace your ancestors and build up a family tree, and before the availability in Jerripedia of church records of baptisms, marriages and burials, and parish registrations from 1842 of births, marriages and deaths they were often the only source available to those without access to original records, who were only able to search online for family information.
However, census returns alone can produce a very misleading family tree and reliance on these records is responsible for many of the glaring inaccuracies in online trees, particularly those in Ancestry.
What is included in the census pages?
Each page from the censuses covers a number of adjacent properties, listing the families living in each of them on the day the census was taken.
The exact dates for each census were as follows.
- 1841 - 6 June 1841 - Details of the census
- 1851 - 30 March 1851 - Details of the census
- 1861 - 7 April 1861 - Details of the census
- 1871 - 2 April 1871 - Details of the census
- 1881 - 3 April 1881 - Details of the census
- 1891 - 5 April 1891 - Details of the census
- 1901 - 31 March 1901 - Details of the census
- 1911 - 2 April 1911 - Details of the census
The head of each household was required to complete a form (or, if he or she could not write, have it completed for them by the district enumerator) declaring all the people living at the property - defined as sleeping there on the night of the census. During the 20th century it was very common for householders to fail to declare tenants, lodgers and others not part of their family, but this happened less in the 19th century, when the occupants of a property would probably be known to the enumerator.
Important information about what censuses contained and how they were conducted can be found by following the links above for each census. Differing details are given from census to census, which may be relevant to other years. Here is a general guide to what census records contain
The return would show the names of those living in the houshold. Usually only one personal name was recorded, invariably the name by which the individual was known. This would not necessarily be their first forename, and in many cases it would not be the name which was recorded at the time of birth and/or baptism. People given French personal names at birth frequently changed to an anglicised version of the name as the 19th century progressed, particularly in St Helier, which did not cling to the use of French and Jerriais to the same extent as the country parishes.
Many Pierres became Peter, Jeannes became Jane, Jeans became John, etc. It was quite common for the change to take effect at the time of a marriage, which means that children would be known by a French personal name while living with their parents, but adopt the Anglicised version when they married, left home and became heads of their own houshold. This can complicate the process of tracking ancestors backwards through successive censuses.
The problem also arises when researching birth, baptism and marriage records. It is not unusual to find a man using one personal name at the time of his marriage, and perhaps for the registration of the birth or baptism of the earliest of his children, and then to use another version of the name when subsequent children arrived.
Family names would rarely change, because they have legal status, but spelling errors are by no means uncommon.
Wives are invariably recorded in censuses with their husband's surname, although in other official records it is more usual for them to be listed under their maiden name, and described as wife of ....
This has led to many online family trees based solely on census returns, showing a wife with the same surname as her husband and children. Usually it can be assumed that the woman had a different maiden name, but there are, of course, a small, but significant number of marriages between couples with the same surname, sometimes cousins, sometimes not closely related, if at all.
The age of each household member on the day of the census was recorded. This was shown as a number of years, or months for children born within a year of the census date. The exception is the 1841 census, for which the age of all adults was rounded down to the nearest multiple of five.
This can be one reason why an individual's date of birth appears to vary from one census to the next. It might be thought that someone shown as 30 years old in the 1851 census, would be recorded as 40 in the 1861 census, and 50, 60, 70 etc in subsequent censuses. If only research into family records was that simple! It was very common for mistakes to be made in recording ages, or for people to lie about their age.
There may have been many different reasons for this, the most common being that as ladies grew older, some tended to claim that they were younger than their actual age. Others who lied about their age when they married may have progressively reverted to their real age in subsequent censuses.
There are numerous examples of people whose date of birth, calculated from the age shown at different censuses, appears to move backwards and forwards. This gives the family historian little option but to choose an average, or identify the year of birth from other records.
Registrations of births after 1842, shown on a birth certificate or in parish registers, are considered the most accurate indication of birth dates, because they were invariably recorded within a fortnight of a baby being born. Baptism records after 1842 usually show birth and baptism dates, and for those children born before 1842 , it can usually be assumed that they were baptised within a few weeks of their birth, although there are exceptions.
The census return indicates who is head of a household. Usually this will be a man, but widows will be recorded as head, as will wives of husbands who are absent from home on census night, perhaps working on a ship, in hospital, or with some other reason for not being with their family. The relationship of others to the head of household is indicated in the return.
Sometimes this helps identify a wife's maiden name. If the head of household's widowed mother-in-law is living with the family, her name as shown in the return will probably be the same as her daughter's maiden name. Not always, however, because some widows reverted to their own maiden name after their husband's death, and others may have given their surname from a second or subsequent marriage. A woman described as the head of household's sister-in-law could be married or widowed, and using her married name, or may even be the wife or widow of the head of household's brother. Care must be taken in making assumptions from names given in censuses because these are classic means by which errors can creep into family trees based exclusively on census returns.
The occupation of each member of the houshold is given in the census return, and usually, for the head of household, there will be added information if he owns his own business and employs others. Employees may show in the houshold return, or may be living elsewhere. These details provide a valuable insight into the social standing of a family, as do listings for live-in servants.
Place of birth
The parish of birth of each houshold member is shown. This is usually reliable, but it is far from unknown for a declaration to be made that all children of a couple were born in one parish, whereas one or more was actually born when the family moved to another parish. Those born outside the island, in England or France, are usually recorded as such, and sometimes their place of birth is helpfully narrowed down to a particular town or village. Beware of spelling errors for birthplaces in Brittany and Normandy!
When early children are shown as born outside the island and later children as born in a Jersey parish, that is usually a strong indication that the couple married before arriving in Jersey and their time of arrival can easily be estimated.
Searching census records
The Ancestry search engine allows a search to be made for an individual name, or the pages covering individual districts within a parish can be browsed. This can be very helpful if a search for a name does not produce the expected result. This is not uncommon because the standard of transcription accuracy is poor and a simple spelling error in a transcribed name can make the record impossible to find.
Frequently a family will appear at the same address in two censuses, but be missing in an intervening year. A search for the page relating to that address may reveal that the family was at the same location, but has simply not been found by the search engine.