Registration of births, deaths and marriages in Ireland started in 1864 except for non-Catholic marriages which start in 1845. There are over 15.5 million records available on-line and they are a major resource for family history research. A separate blog provides the details of the records that are available and how to access them. However, a question often asked is whether the early records are fully comprehensive, i.e. were all relevant events captured by the registration process or are there records missing; and if so, what is the scale of missing records ? This blog reviews the evidence.
A historical and practical context is useful in understanding the factors that might have caused gaps in the civil records.
The Registration System. The registration process was built around the infrastructure that was already in place to administer the Poor Law. Officers were appointed , and a Registrar’s office was built in each Poor Law Union associated with, but separate to, the workhouse. The law specified the size of the registrar’s office building and the facilities to be provided including a strong-room for storage of the records. The core registration staff appointed included 163 Superintendent Registrars and 770 District Registrars. They were respectively responsible for areas called Superintendent Registrar’s Districts (which were effectively the same as Poor Law unions) and their sub-divisions called Registrar’s Districts. All of these staff ultimately reported to a Registrar General who was based in Dublin. His office collected the information and compiled annual reports which are available on-line.
Within each Poor Law Union the registration staff were appointed by the local Poor Law boards of guardians. These bodies administered the Poor Law in their area, including management of the local workhouse. Their members were land-owners and other local dignitaries. To guide the Boards in their appointments, the law specified minimum qualification for registrars: they should be over 23 years of age, be local residents and not bankrupt! The clerks of the Poor Law Unions, by virtue of their office, had an entitlement to be appointed to the more senior role of Superintendent Registrar. They could fulfil this role as an additional part of their job as clerks. Similarly, the local Medical Officers of Dispensary Districts had an entitlement to be appointed to the more junior role of District Registrar. The vast majority of the roles were filled in this way. Only 32 Registrars were appointed by the Guardians, and most of these were local medical doctors or public officials. In addition to establishment of the registration infrastructure in advance of the 1864 start date, public notices and other notifications of the registration process were posted locally and churches and other organisations notified. In summary, the infrastructure for the registration process was set up on time and everything would appear to have been in place, and appropriate, to administer the new law.
Although registrars were ready to receive the information, in practice the onus for registration was on the public. There was no fee to register but non-compliance involved fines which were significant for the average person at the time. The process was:
- Birth. A birth had to be reported within 21 days by the parent(s); or (if they are unable for some reason) the nurse; or the occupiers of the house where the birth occurred. Penalty for failure to register was £1. This may have had the effect of causing parents to report a birth date which is later than the actual date so as avoid the fine. A check of the baptismal date may be useful. A change of family history relevance was made in 1881. Until this date mothers of illegitimate children could indicate the name of the father in the registration of birth. From 1881 this could only be done with the agreement, and signature, of the father.
- Marriage. ‘The husband’ must obtain a blank certificate from the local registrar, bring it to the clergyman for completion and signing at the ceremony; and return it to the Registrar within 3 days. Clergy were supplied with blank certificates for convenience, but the husband was obliged to return the completed certificate. This process caused problems – see below. Penalty for failure to register was £10.
- Death: A death had to be reported (within 7 days) by a person present at the death or ‘last illness’; or the occupier of the house where it took place; or ‘Any person present at, or having a knowledge of, the circumstances…’. Penalty for failure was £1. As most registrars were medical professionals, the cause of death was also recorded where possible. To ensure conformity in defining the cause of death, each registrar was supplied with a copy of ‘Dr. Farr’s Nosology‘, a guide also used by registration staff in England and Wales. The full title, and a sample entry, are shown at right. The cause of death was defined according to medical knowledge of the time. A web search of the terms used may provide useful information on the current understanding of the cause. A significant change in classification of causes was made in 1881 and is detailed in the Registrar’s Report for that year.
Despite all of these preparations, not all events were registered in the early years. This is demonstrated in the analysis reported below, and also acknowledged by the Registrar General himself. In his Annual Report for 1864 he states “… judging from the number of events recorded in 1864, I consider that many births, deaths, and marriages have not been registered. In several instances prosecutions have been resorted to with some advantage; but notwithstanding the exertions made …. numerous cases of neglect to register births and deaths have been reported to me, but frequently at too late a period … to allow of the parties being proceeded against at the Petty Sessions“.
On marriages, the Registrar particularly notes that “registration of marriages…. is not at all satisfactory in some districts, owing to the clergy who have celebrated the marriages having declined to sign the necessary certificates”. The reason for this non-cooperation is not clear, but any of the issues discussed below could have been a cause. In his 1870 report the Registrar noted that ” I regret to state that there appears but little improvement..”. By 1881 the Catholic clergy seem to have become more cooperative. In his Report of that year, the Registrar General states that arrangements for recording of Catholic marriages were “still of a cumbrous and unsatisfactory character. While the Roman Catholic clergy….. and the Registrars do their best to administer the Act in their localities, yet so much depends upon the manner in which the husband discharges his duty as conveyor of the certificate of marriage to the Registrar”. As a result, he suggests “many marriages are imperfectly registered, and … a considerable number even escape registration altogether through the ignorance or carelessness of the parties married“. He advocates a change in the legislation to address this.
To give an indication of the extent of non-registration, the registrar (in his 1864 Report) provided a comparison of the rate of births, deaths and marriages in 11 other European countries and regions (ranging from Scotland to Spain) and compares these with the rate recorded in Ireland (see above). The average rate of births in these countries is 1 for every 28 people in the country (ranging from 1 in 23 in part of Germany to 1 in 38 in France) whereas in Ireland it was 1 in 42. Because of the wide range, it is not possible to determine the extent of missing births, but it does suggest that the register is not complete. In a later report the Registrar modifies his estimation of the low level of birth registrations. In this case he calculates that the number of births should be directly related to the number of married women. The 1861 Irish census showed a much lower proportion of married women of child-bearing age than in England. This he suggested was the cause of the lower than expected level of births in Ireland.
The extent of missing death records is noted in the Registrar General’s 1881 report. A comparison with the returns of the 1871 census showed that “the number of deaths registered in the City· of Dublin fell short of the number of Dublin citizens buried in the principal Dublin cemeteries by about 9 per cent“. This level of non-registration, he suggests, could be extrapolated to the whole country. Registration of deaths was greatly improved in 1879 when clerks of cemeteries etc. were obliged to make a return of persons buried in their grounds.
What are the factors that might have caused this non-compliance?
Public Commitment: Registration was introduced in 1864, which was around 15 years after the end of a Famine that had devastated the country and whose economic and social impacts were still being experienced. The population had been reduced by 2 million through death and emigration, and the British government’s laissez faire policy of non-intervention was widely, and correctly, perceived to be a major contributor. It would be understandable if there was public cynicism that a government which had shown no concern for their deaths, was now demanding that its details be recorded. The 1860s was also the early period of growth of Fenianism, the Irish Republican Brotherhood, and several of the other organised political movements which would ultimately lead to Irish independence. Full cooperation with British administrative requirements may have been resisted by some.
Literacy: A further challenge for the registration process was illiteracy. The 1861 census shows that 42% of the population were illiterate. This ranged from 34% in Leinster Province (the East and SE counties) to 63% in Connaught (the mid -West counties). This caused difficulty in promoting awareness of the new registration process, and also in collection of details.
The Registrar’s Staff: Cooperation with the process required public willingness to submit information, and also some zeal on the part of the local registrars to locally promote the need to register, and to encourage submission. It may be that not all of the registration staff were diligent in their duties. We have no evidence on this issue.
Given these challenges, it would be extra-ordinary if every event (i.e. birth, marriage or death) was captured by this new process.
Analysis of Missing records
The rate of registration did increase over the following decades as is documented in subsequent Annual Reports, and also suggested by our analysis reported below. To estimate the extent of missing records, I made a comparison of Catholic baptisms and birth registrations; and of Catholic weddings and their registration in a random sample of parishes. The rationale for this, and the process used, are detailed below:
Baptisms versus births: Every Catholic child baptised in 1864 should be registered as a birth, so a comparison of baptisms and registered births should indicate the level of non-registration.
I extracted 10 baptisms from the 1864 registers of 21 randomly selected Catholic parishes available on the National Library registers website. The parish selection was made from a map on this site and a wide geographical spread was ensured. Obviously only parishes for which there were registers for 1864 were included. Early January baptisms were omitted to allow for those who may have been born in 1863. The records chosen were usually sequential records within the register, but if the writing in a particular record was unclear, it was omitted in favour of a more legible record. I then checked the civil birth records on www.irishgenealogy.ie to see which of the baptised children had been registered. In the few parishes(see table) where there were high numbers of missing births, I extracted a further 10 records to validate the finding.
This study (see table at right) shows that the average number of unregistered births per parish is 2 (i.e. 20%) in the 21 parishes examined. However, the proportion ranges from 80% missing (in Inver, Co. Donegal) to zero (3 parishes had none missing). It may well be the case that some of the apparently missing events are in the register but in some unfindable form (see notes on name variations below) but that is very unlikely to be the reason for all missing records.
Catholic church marriages versus civil marriages: For marriages an identical process was followed. A random set of 20 Catholic registers were selected (again from the map on the NLI registers website). The exceptions to random selection were Inver, Sneem and Gortnahoe which had the highest levels of missing birth registrations (see Table above). These were selected to assess if marriage registration was also poor. It was not. The same process of selection of records was applied. I then checked the civil marriage records on www.irishgenealogy.ie to see which of the marriages had been registered. This study (see table below) shows that the average number of unregistered marriages per parish is 1.7 (17%) over the 20 parishes examined. However, once again there are regional variations. In one parish (Oristown, Co. Meath) none of the marriages were registered, and in St Nicholas, Galway only five of the 10 marriages were registered.
However, in general there was reasonable compliance and in 14 of the 20 parishes either all marriages were registered (4 parishes) or only 1 was missing (10 parishes). The full results are in the table at left.
Clearly in some areas there is significant non- registration of births, whereas in others there is almost complete registration (admittedly based on small samples). The Registrar, in his 1864 report notes that some priests were unwilling to sign the registration form. This is a likely factor in the case of Oristown, for instance. The local parish priest, Matthew Kelsh, died in 1867 and from then the registration of marriages improves, which suggests that he was one of the priests noted by the Registrar General as refusing to sign the certificate.
I carried out a further analysis by checking whether church marriages in 10 Catholic parishes in 1880 were registered. Ten marriages in each of the 10 parishes were checked but only 2 marriages from these 100 marriages were not registered.
Other approaches. Another means of assessing the extent of missing marriage records is to compare church marriages on Roots Ireland with the number registered. For this exercise, I chose some relatively uncommon names and made a comparison between the numbers of church marriages and registered marriages. The rationale for choosing less common names was simply to make the comparison process easier by ensuring a lower number of records. For instance, searches of Roots Ireland in 1865 show:
- There are 5 Catholic marriages of persons called Daniel Ryan: 3 are registered in Civil records.
- There are 7 marriages of persons called Sarah Curran: 5 are registered
- There are 6 marriages of persons called Mary Mulholland 3 are registered
- There are 12 marriages of persons called Comber/Comer 4 of these are registered.
It should be noted that the indexing of civil records is not perfect and some of these marriages may be registered but are not being found in searches. Nevertheless, this does give an indication of the extent of missing records. It is also generally consistent with the above studies.
What can we tell from all of this? Some births and marriages are definitely missing from the early period of registration. The extent of missing records is regionally diverse. However, the cause of this variation is unclear. One might imagine that rural regions in which people live in remote communities might have lower levels of registration. However, Lettermullin in the west of County Galway, which is as remote a place as exists in Ireland, had no missing births (See ‘Missing Birth’ Table above) while St. Mary’s in Cork city had 5 missing births. Other factors are clearly at play. Some parish registers note those travelling through as ‘vagi’ or strangers by other notations. It might be imagined that these transitory people would be less inclined to register. In a few incidences, it was noted that the missing registration was for births or marriages marked thus. However, it was not a general rule.
Another feature of this exercise was the disparity in name spellings found between church and civil records. This is the bane of genealogists. The extent of illiteracy in 1864 is indicated above so many people would not be registering their own names. In practice, baptismal and birth registration was an exercise in transcription of a spoken name into a written form by two different people, i.e. the priest and the registrar. The priest would have had more local connection and cultural sympathy with his parishioners. He is more likely to render a name according to local and national practice. The registrar, on the other hand, was probably not a local person. By virtue of their medical education, and the fact that they have a public appointment, they are less likely to have local cultural sympathy. They are therefore more likely (for very understandable and innocent reasons) to render local names as synonyms of English names. These are generalisations, I freely admit. An illiterate person registering an event is not in a position to correct whatever spelling is entered in the record, and thus we get two very different spellings for the same person within the same area and within the same month.
Some of the significant disparities in spelling detected just in this exercise are listed below. The simpler variations (Maloney – Moloney; Loague – Logue etc.) are omitted as they are frequent. However, other variations might be difficult to find in any search. The most significant was a child baptised as McVeigh in Ballymacnab parish, Armagh in 1864 but registered in the civil records as McCreagh. Other name variations between birth and baptismal names include:
Downing – Downey, Byrne – Burns, Adgley – Adley, Mullany – Mulvany, Neylan-Nealon; Gonagle – McGonigle; McGready-McGruddy, Dulap-Dunlop, Whelehan – Houlihan, Bresnihan – Brosnan, Currane – Cournane, Connellan – Conlon, Camphill – Campbell, McEnulla – McNullar, Deegan-Digan, Lynam – Lynane; Doherty – Daughuty, Magrilan – McGrillan; Hagarty – Hegarty;
A few renditions of first names might also be mentioned: Manus (a relatively common name in Northern counties) is registered as Manasses (an unrelated biblical name) in several records; and Maurice as Morris.
Finally, there are the births in which no name is recorded at all. Many births are reported only as ‘Unknown’. These are generally births reported by the major hospitals. John Grenham deals with this issue in a blog and reports that no fewer than 193,493 births are recorded between 1864-1919 with no forename.