Northern Ireland

Chapter 2
A Brief History of the Census in Ireland/Northern Ireland
Ian White
Office for National Statistics

Registrar General Northern Ireland Annual Report 2011

Ian White works for the Office for National Statistics (ONS). He has worked at ONS (and its predecessors,
the Office for Population censuses and Surveys, and the General Register Office) since 1970, and has been
involved in every UK census (in one way or another) since then (five censuses in all). He is the last remaining
active member of the ONS staff who worked at Somerset House, the home of the very first General Register
Office (GRO) Census in England and Wales in 1841. Ian has a wealth of experience on census matters. He
headed the work on UK legislation for the 2001 and 2011 censuses and has been involved in working within
Eurostat and the United Nations Economic Commission for Europe on international census matters.
Ian has written a book on the history of the census in the United Kingdom (census and Sensitivity) to be
published by Palgrave Macmillan next year. This chapter contains material extracted from the book and
presents a look at the history of the census throughout Ireland before 1921 and in Northern Ireland since
1926.
Ian was supported in his writing of this chapter by Mr Uel McMath from NISRAs Census Office. Uel has
worked on the last four censuses in Northern Ireland since 1981 with an occasional inter-census sabbatical.
Uel managed the Field Operations elements of both the 2001 and 2011 censuses and holds a wealth of
experience of the recent censuses and also of the legislative basis of the census in Northern Ireland.

36

Many inquiries of a similar nature, might no doubt, have been pursued with an advantage to a correct
knowledge of the condition of the people. We felt, in fact, that a census ought to be a social survey, not a
bare enumeration, but we were restrained by the apprehension that jealousy and prejudice might be excited
if we made our inquiries too searching and too minute People are slow to see that questions relating to
themselves and their households can have any bearing on the general good, and forget that, in accounts of
large numbers, the individual is wholly lost sight of in the average, but that averages can only be obtained by
an accurate knowledge of all that pertains to the individual.
Report of the Commissioners appointed to take the Census of Ireland for the year 18411

2.1 The all-Irish censuses:

1813-1911

1813
2.1.1
The 1800 Population Act that provided
for the first census in Great Britain obtained Royal
Assent on 31 December, just one day before
passing of the Act of Union. This, together with a
very much different local administration in Ireland
meant that the 1801 Census did not extend across
all parts of the new United Kingdom. However, to
make amends for this, a Bill to take a census in
Ireland was presented to Parliament on 31 January
1806 by James Fitzgerald, the MP for Ennis, County
Clare.
2.1.2
He noted, thanking the Speaker for
allowing time to introduce such a measure, that a
census in Ireland would help to

assimilate the laws of and regulations of

Ireland to those of this country, as far as it is
just and expedient,

2.1.3
Fitzgerald echoed arguments that Charles
Abbott MP had made in the House of Commons in
support of the 1800 Bill:

It must indeed be obvious to every man

conversant with the subject that to ascertain the
population of a country is a grand desideratum
in political economy, whether considered with
regard to its physical force, its agricultural
produce, or its financial capacities.

2.1.4
He observed, however, that in Ireland
there was no mechanism equivalent to English poor
law administration, for enumerating the population.
Instead he proposed that the Bill should give the
responsibility to

the magistrates of the City of Dublin with

the advice of their recorder, and the magistrates
of the other Irish counties aided by the advice
of the assistant barristers, at a general quarter
session, to appoint one or more householder,
according to necessity, for each barony, parish
or such other district as may be deemed
expedient, for the purposes of collecting a
list of the inhabitants of such districts, the
expense arising from such appointments to be
defrayed by the assessment of the grand juries
of the several counties in the same manner as
presentment for roads etc, are now provided.

2.1.5
Fitzgerald was concerned to give the
House the fullest opportunity of examining the
Bill, and to that end he proposed that he should
not move the second reading until after the Easter
recess. However, having been given the opportunity
to do so, the House clearly thought that the
desideratum was not strong enough, since the Bill
failed to get through its next Parliamentary stage
and was dropped. And it was only during the debate
on the next Population Bill for Great Britain in 1810
that the proposal for a census in Ireland was again
raised.
2.1.6
This led to a second Bill introduced on
29 January 1812 by Sir John Newport MP, who
expressed his surprise that

a measure should have been so long

deferred which would enable Parliament to
ascertain the number of persons for whom it
is to legislate. Twice since the Union has the
population of Great Britain been calculated, but
in this respect, as well as in others, the natives
of Ireland have been totally neglected.
37

Registrar General Northern Ireland Annual Report 2011

2.1.7
He considered it an absurdity that all Bills
did not automatically cover the whole of the United
Kingdom and that it was left to the individual
introducing them to determine what parts of the
kingdom should be included. He noted that Ireland
was more often than not excluded.
2.1.8
Conducted under the supervision of
William Shaw Mason, the 1813 Census attempted
to follow the mode of the 1811 Census in Great
Britain, but with an administrative system rather
similar to that set out in the aborted 1806 Bill.
The Enumerators were drawn from the barony
constables and parish officers, which led to some
hostility from the local population. The resulting
conflict, combined with poor administration, led to
the submission of very incomplete returns. Mason
valiantly extended the Census until 1815 at which
point he finally gave up the attempt to collect any
further returns. No results were published at the
time, but in the Report of the subsequent census
of 1821, Mason wrote of the 1813 enumeration

At the expiration of two years employed

in trying to accomplish the object of the
Legislature, it was found on examining the
returns, that out of the forty Counties and
Counties of Cities into which Ireland is divided,
ten only furnished complete returns; in four,
no steps whatsoever were taken in pursuance
with the Act; and those of the remaining
twenty six were inaccurate or defective. The
Act therefore may be considered to have been
wholly inoperative as to its main objective,
that of ascertaining the number of souls by
actual enumeration. By the aid of comparative
calculations founded on previous inquiries and
on the partial results of the Act, the amount of
population in 1813 has been conjectured to be
5,937,856.2

Whitelaws 1798 Census of Dublin

2.1.9
It was of particular disappointment to
Mason that the City of Dublin was one of the
districts for which no returns were collected, not
least because this was clearly the most populous,
but also because of the earlier prodigious, and
38

far more successful, effort to enumerate the

city by the Reverend James Whitelaw in 1798.
In a footnote to the 1813 Census figure quoted
in the 1821 Report (and which was estimated
on the basis of some rather spurious statistical
calculations, and is generally regarded as
unreliable), Mason added:

Although in a recapitulation of this general

nature, the attempts of ascertaining the
population of particular districts are not
properly admissible, yet it would be scarcely
justice to the memory of a most intelligent
and persevering character, the late Rev. James
Whitelaw, to suffer his account of the population
of the city of Dublin to pass wholly unnoticed.
The peculiar circumstances of that city, during
the rebellion of 1798, led him to undertake
an account of the population by actual
enumeration. At that period every householder
was obliged to affix on the outside of his door, a
list of the names of every person then residing
in the house. The numbers were then collected
by Mr Whitelaw, and published by him, together
with a comparative statement of the numbers,
taken in 1803, by the conservators of the
peace, after the insurrection in that year. The
totals, in both cases, were

In 1798

Houses
Inhabitants
16,401 172,091

In 1804

Houses
Inhabitants
15,958 169,528

Decrease in
six years

Houses
443

Inhabitants
2,563

2.1.10 Indeed, Whitelaw (though he little realised

it at first) had set himself a daunting task, but one
only made possible, ironically, by the disturbed
state of the country resulting from the United
Irishmen Rebellion of 1798. As a security measure
at that time the Lord Mayor of Dublin, Thomas
Fleming, ordered that a list of inhabitants be fixed
to the door of every house in the city. Thus the task
of census-taking seemed to Whitelaw at first to be
a fairly straight-forward matter of merely collecting
the number of names.

2.1.11 Fuelled by the lack of any reliable data

on the population of the country, Whitelaw had
been motivated to contemplate such an innovative
exercise in the first place for much the same
reasons that had driven John Rickman to propose
a census in England, in his 1796 pamphlet on
Thoughts on the utility and facility of a general
enumeration of the people of the British Empire.
(Rickman would, of course, go on to oversee the
first four censuses in Great Britain.

Figure 1 The title page of Whitelaws report

2.1.12 Referring to the population controversy

that had been raging in England throughout the
latter half of the 18th century, Whitelaw wrote in
his report to Philip, Earl of Hardwicke, the Lord
Lieutenant General and Governor General of Ireland

Struck with this strange diversity of opinion,

anxious to ascertain the truth, and influenced,
perhaps, by a laudable ambition of being the
first to offer to the public what is often wished
for in vain, an accurate well-arranged census of
a considerable capital, I availed myself of the
favourable opportunity offered by the unhappy
situation of this city at the commencement of
the late rebellion; and with the sanction of the
Government, but at my own private expense
and toil, began a census of the inhabitants of
the city of Dublin, early in the month of May,
1798.3

2.1.13 He initially anticipated no real difficulty,

believing that he would have little more to do
than transcribe carefully the list of the inhabitants
posted on the doors. Indeed, in more prosperous
areas there was generally some individual member
of the family competent enough to do the task,
and, there, Whitelaw found the lists to be correct
(though he did not explain how he verified this). But
it was clear that among poorer families that formed
by far the majority of the population, the case was
very different. There, he reported

The lists on the doors were . frequently

illegible, and generally short of the actual
number by a third or even a half.

2.1.14 Thus he resorted to a more intrusive

enquiry

In order effectually to obviate the difficulty, my

assistants and I, , explored, in the
burning month of the summer, every room of
those habitations, from the cellar to the garret,
and on the spot ascertained their population.

2.1.15 He expected opposition but experienced

none, and found that

everyone was ready to cooperate.

2.1.16 Whitelaw put this down to a combination

of a fear and dread that the authorities would take
punitive action against anyone showing any form
of disaffection, and a belief by the inhabitants
that he was employed by the Government to help
prepare some sort of system for the relief of
their conditions. He noted that, in the course of
the whole survey, only one Enumerator had been
assaulted, when

in attempting to remonstrate with a butcher

of Ormond Market on the incorrectness of his
list, the butcher flung at him a quantity of blood
and offal.

39

Registrar General Northern Ireland Annual Report 2011

2.1.17 The method for carrying out the

enumeration that Whitelaw described proved to be
a model that subsequent census takers could only
aspire to, in particular, the lengths he took in the
training and supervision of his field staff. He was
initially very much concerned about the ability of
his Enumerators to carry out the task:

in 1815 leading to a rather more successful

attempt in 1821 to carry out the census throughout
the whole of the island at the same time. The
Act provided a more stable administration for
the census than its predecessor, and overall
responsibility was transferred from the Grand Juries
to the Bench of Magistrates.

I was, at first, much embarrassed by the

inexperience of my assistants. I employed them
therefore in taking surveys of the streets that I
had already surveyed myself, until I discovered
that they had attained a sufficient degree of
accuracy. I never relied on their returns with
implicit confidence, but made them frequently
act as checks on each other. Two or more of
them frequently surveyed the same street
in succession, without any communication
with each other, and if any material variation
occurred, I investigated it myself on the spot.

2.1.22 The Enumerators were mostly appointed

from the pool of tax collectors, who, like the Poor
Law officials in England, would be familiar with
the area and with its inhabitants, and for the
first time in the United Kingdom the Enumerator
was instructed to write down the details about
each individual in a record book. The information
collected included the name, relationship to head
of household, age, occupation, number of acres
for landholders, and any observations. The census
was taken on 28 May 1821, as it was elsewhere in
the United Kingdom.

2.1.18 In addition, Whitelaw himself reenumerated the hardest-to-count populations living

in the poorest streets, as he expected these to
produce the most error. Thus he was able to report

2.1.23 Much of the improvement in this census

was due to the fact that the tabulations were
carried out centrally, leading to uniformity of the
returns of families. However, the resulting total
population figure of 6.8 million may still have
been somewhat of an undercount, since not only
were the Enumerators inexperienced (though they
may well have been familiar with their area), but a
general lack of public confidence in the Government
made householders reluctant to disclose to the
authorities information that they believed might
then be used as the basis of further oppression,
and rendering them liable to tax duty and/or
service in the militia.

in the poorer parts of the city, there are few

streets that have not been twice, and some,
even three times surveyed.

2.1.19 Whitelaw was particularly concerned with

determining the density of population, not only to
highlight the extent of deprivation in the poorer
districts, but also to produce an algorithm which
could be applied to calculate the populations of
other European capital cities.
2.1.20 His figures revealed, for example, that
the parishes within the old walled part of the city
were much more densely populated than the newer
suburbs, ranging from an horrendously overcrowded
439 persons per acre in St Michaels (opposite
Christchurch) to a more comfortable 87 per acre in
St Thomass (around the Sackville Street/Gardiner
Street area now known as OConnell Street).
1821 and 1831: improving the count?
2.1.21 Following the failure of Masons first
enumeration, another Act had followed immediately
40

2.1.24 However, by the time the 1821 Census

was taken, relations between the Government and
the public seemed easier than they had been for
some years before, or were to be for some years
after, and the Catholic clergy ameliorated some
of the publics concerns, encouraging their church
members to give accurate information. Not without
some success it would seem since, in the Report
of the 1841 Census1, the Census Commissioners,
while they thought that the 1821 was indeed likely
to have produced an undercount, did not regard
it as defective enough to make the difference

between it and the 1831 Census figure (14.2 per

cent) an inaccurate representation of population
growth over the decade. Nevertheless, much of the
information collected in the 1821 Census remained
unpublished.

according to the Report of the 1841 Census, was

taken

2.1.25 The 1831 Census itself was an enigmatic

enquiry. Very little commentary from it exists, and
we can only make conjectures from the limited
results that were published and the few comments
made in the reports of subsequent Irish censuses.
2.1.26 The census was undertaken by George
Hatchell, designated, according to the only report
published as The Officer appointed by the Chief
Secretary of the Lord Lieutenant for digesting and
arranging the Population Returns of Ireland. The
Report, however, gives neither an introduction
nor any details of enumeration procedures
adopted - not even the date of which the census
was nominally supposed to have been carried
out4. It is likely, however, that a procedure similar
to that employed in 1821 was followed, where
Enumerators, recruited from the pool of local tax
collectors, recorded information by means of doorto-door visits. Certainly there would have been little
time to develop a significantly different approach
since the relevant legislation for it was only passed
on 30 March that year.
2.1.27 The questions asked were along the
lines of those included in the 1831 British Census,
and omitted the question on age. Interestingly,
however, one surviving return, for the parish of
Dunboe in County Londonderry records information
on religion, separately identifying Church of
Ireland, Roman Catholic, Presbyterian and
Other, though there is no reference to any such
information in Hatchells Report. It is possible
therefore, that such information was recorded by
the Enumerator as part of his observations and
was, thus, not universally collected and would not
therefore have been officially reported.
2.1.28 It is generally thought that, whereas the
1821 Census under-estimated the population, the
1831 figure of 7.8 million was an over-count arising
from the fact that, in the event, the enumeration,

at different dates, at different times

extending over a considerable period. It
is understood, too, that the enumerators
considered that they would be paid - and
in many cases were paid - in proportion to
the numbers they enumerated, the obvious
tendency of which would be to augment the total
numbers.1

2.1.29 Certainly the figures were, at the time,

considered unsatisfactory since they were subject
to a further examination and correction in 1834 by
the Commissioners of Public Instruction for their
1835 Report.
1841: a pioneering effort
2.1.30 In 1841, while use was being made
of the recently created Registration Service in
England and Wales, in Ireland a new network of the
officers and men of the Dublin Metropolitan and
Constabulary police force were employed as the
field force for this and subsequent censuses.
2.1.31 The 1841 Census in Ireland was, in
comparison both with the two preceding censuses
and with the corresponding enumeration in Great
Britain, a remarkable achievement, not only in the
method of the collection of data but in the range
of questions covered. In addition to the topics
included in England and Wales, the Irish census
form enquired into members of the family who
were absent and who had died since the previous
census, and a separate form collected information
to assess the quality of the housing stock and
some agricultural statistics.
2.1.32 The authority for the census was the
Census (Ireland) Act, 1840 passed the day after
the corresponding legislation for Great Britain on
10 August 1840. At first Parliament had agreed
that the census should be taken

on one or more consecutive days in the

month of July 1841,

with the actual date(s) to be determined by the

Lord Lieutenant; a date of 1 July eventually being
41

Registrar General Northern Ireland Annual Report 2011

fixed. But following some adverse reaction to

such a comparatively late date, the decision
was subsequently taken, at the suggestion of
the Statistical Society of London, to bring the
enumeration forward to 6 June so that it should be
simultaneous with the census in Great Britain. A
further Act was passed, accordingly, to amend the
earlier legislation.
2.1.33 The Act was considerably less specific
in detail than the corresponding Act for Great
Britain; but this led, ironically, to the more detailed
question content. It required Enumerators to

take an account, in writing, of the

number of persons dwelling therein, and of the
sex, age, and occupation of all such persons,
distinguishing the persons born in the place or
parish, and county in which they shall be then
dwelling; and shall also take an account of the
number of inhabited houses and of uninhabited
houses, and of houses then building, within
such districts respectively; and shall also
take an account of all such further particulars
as by such instructions they may be directed to
inquire into; such particulars and instructions
having no reference to the religion of any person
or persons.

2.1.34 This gave a relatively free hand to

the Irish authorities to include any additional
questions as they wished. Though the collection of
information on religion was specifically prohibited,
other information considered to be of value could
be collected on the authorisation of the Lord
Lieutenant, who delegated this responsibility
to the Chief Secretary, who in turn passed the
task of carrying out the census to a group of
Commissioners, namely William Tighe Hamilton,
Henry J Brownrigg, William Robert Wills Wilde and
Thomas Aiskew Larcom. Each was chosen because
of the responsibilities they held in other areas of
Irish public administration.
2.1.35 Hamilton, the Chief Commissioner, was a
civil servant in the Chief Secretarys office, holding
the purse strings and providing authority for the
collection of the unspecified information. Brownrigg
42

was Inspector General of the Constabulary, and, as

such, controlled the personnel necessary for the
enumeration. Wilde (father of Oscar) played a key
role as a medical and cultural commentator in the
1841 and subsequent Irish censuses (for 1841
he was employed particularly to analyse the data
on deaths which would be collected at the census
for the first time). And Larcom, a career soldier,
was working at the Ordnance Survey of Ireland
and responsible for providing statistical, social,
geological and topographical guides for Ireland.
2.1.36 In addition to paving the way for a much
larger census, with more questions in order to
enquire into the condition of the people, Larcom,
and his fellow Commissioners also introduced
a number of other important innovations to the
Census, none of which had been carried out, either
in Ireland or in Great Britain, before.
2.1.37 The Enumerators (mainly constables)
were instructed first to complete a survey of the
houses in their district, recoding information for
each on;
whether it was built or being built;
whether the walls were constructed from mud,
stone or brick, and whether the roof was slated
or thatched;
whether it was a private dwelling, public
building, manufactory, hotel, public house,
lodging house, shop, school etc;
whether inhabited or uninhabited; and
the number of stories (including basement)
rooms (exclusive of closets) and front windows.
2.1.38 This enabled a four-fold classification
of dwellings derived from the number of rooms
and windows, and durability (as assessed by the
material used in the constructions of their walls
and roofs), to be produced which would, the
Commissioners hoped

throw some useful light upon the general

condition of the community, as there can be no
more obvious indication of the advances and
condition of a people than improvement in the
comfort of their residence.1

2.1.39 For farms, information was also collected

on:
the number of acres (distinguishing between
arable, pasture and waste); and
numbers of livestock (horses, cattle, sheep,
goats, pigs and poultry)
and for all properties, the Enumerators were
required to record:
the number of families in the house; and
the name of the head of each family.
2.1.40 A definition of families was given for the
first time, being

either one individual living independently in

a house, or part of a house, on his or her own
means of support, or several individuals related
to each other, as parents and children, brothers
or sisters etc, with the addition of servants or
visitors, living together in the same house, or
part of a house, upon one common means of
support. A lodger, or family of lodgers, who do
not board with the family of the house, should
be treated as a distinct family. Where such do
board with the family of the house, they should
be treated as one with it.

- a description which we easily recognise as the

basis for the definition of a household today.
2.1.41 The information recorded on the housing
schedule relating to families and heads of families
was then used by the Enumerator as a check on
the information collected from the household form,
which (as in Great Britain) was to be completed by
the head of family for the first time. Blank copies
of these were supposed to be distributed to each
family in each house in the week before the census
by the Enumerator, who was instructed to ensure
that the house number used in the housing form
was written on the household form.
2.1.42 The latter contained three tables for
each head of family to complete. The first asked for
information relating to each member of the family
on:

age, sex, and marital status;

relationship to head of family;
year of marriage(s);
occupation;
school attendance and literacy (recording either
Read, Read and write or Cannot read); and
birthplace.
2.1.43 For those living on farms further
information was collected on the number of days
employment and the level of wages.
2.1.44 The occupation data collected allowed
Larcom to derive an occupational classification
somewhat more developed than those adopted
by John Rickman and William Farr in Great Britain.
Firstly, families were classified according to their
pursuits in a way that distinguished between
those dependent on agriculture and those on
manufacture and trade, and which provided a
reasonable comparison with the figures presented
for Great Britain (and for Ireland from the previous
census). Secondly, they were classified according
to their means thus creating three groups which
could be said to represent upper, middle and lower
social classes. Thirdly, people were classified
according to how they ministered to the wants of
others.
2.1.45 The second table sought to record
name, age, sex, relationship, occupation, and place
of residence for those family members absent
on census night, which, thereby, anticipated the
collection of information on all residents in Great
Britain by some 90 years. The Commissioners
Report, using such information, together with that
on occupation, devoted much space to commenting
on migration, both long-term and short-term, and
both internal and external. In commenting on the
phenomenon of economic migration, for example,
they wrote

It is gratifying to see that a considerable

amount of productive labour is afforded by the
counties in which the manufacture of linen and
cotton prevails. Antrim, Down, Armagh and
43

Registrar General Northern Ireland Annual Report 2011

Londonderry by the prevalence of bleachers,

indicate their peculiar proportions. In almost
every county a considerable number of male
strangers appear to be weavers, and of females,
spinners. Shoemakers appear to migrate more
than tailors, but both are numerous. Artificers
in building would seem to be everywhere in
demand, and form a moveable class.
2.1.46 And in commenting on the reasons
for including a question on absent residents
specifically, the Commissioners noted (describing a
sort of 19th century e-Borders)

We had taken measures to obviate certain

inconveniences which we had reason to
apprehend would arise from the lateness of
the season at which the census was taken. Of
these the principle arose from the fact, familiar
to everyone acquainted with Ireland, that in
the month of June the agricultural population
is in a state of considerable movement; the
labourers resorting, in search of work, to
neighbouring counties, which require more
labour than the resident population can supply,
and many proceeding to England or Scotland
for the purpose of reaping the harvest. The
numbers thus migrating to Great Britain having
been variously stated, we required from the
police at every port, an enumeration of all deck
passengers who embarked on board the various
packets during the summer of 1841. For this
purpose a competent officer was stationed
at each packet office, and each person as he
received his ticket for embarking was asked
from what country he came.

2.1.47 The third table was to record

information on all people who had married, and
those who had died in the family residence, during
the previous decade. Name, age, sex, relationship
to the current head of family, occupation, and
cause and date of death, were required, and the
responses provided a mass of statistical data to
enable William Wilde to report in some depth, in
a second volume of the Census Report, analyses
of mortality in Ireland in a manner that Farr was
only able to emulate in England and Wales with the
44

benefit of civil registration after the establishment

of the General Register Office. Wildes report
certainly overshadowed much of the remainder of
the published census output.
2.1.48 The amount of information collected and
analysed in the 1841 Irish Census was thus by far
the greatest ever attempted anywhere by a census
in the British Isles but it might have been even
larger but for the Commissioners concerns about
not making the Census too intrusive. They noted in
their Report

Many inquiries of a similar nature, might no

doubt, have been pursued with an advantage
to a correct knowledge of the condition of the
people. We felt, in fact, that a census ought
to be a social survey, not a bare enumeration,
but we were restrained by the apprehension
that jealousy and prejudice might be excited if
we made our inquires too searching and too
minute.

2.1.49 And in observing the publics lack of

understanding of the nature and purpose of the
census, which is no less prevalent today, the
Commissioners added:

People are slow to see that questions relating

to themselves and their households can
have any bearing on the general good, and
forget that, in accounts of large numbers, the
individual is wholly lost sight of in the average,
but that averages can only be obtained by an
accurate knowledge of all that pertains to the
individual.

2.1.50 The Census produced by far the most

accurate count of population in Ireland to that date.
1851-1871: a period of population decline
2.1.51 Three years after the 1841 Census, the
Irish Marriages Act 1844, created the General
Register Office in Ireland and the post of Registrar
General. William Donnelly was the first appointee
and held office until 1876, and was responsible
for conducting the next three censuses. However,
the Registrars thereby appointed to register
non-Catholic marriages and to solemnise civil

marriages, were not given the same enumeration

responsibilities as their counterparts in England
and Wales. In Ireland, therefore, the machinery of
the field operation was again left principally to the
police.
2.1.52 The Census (Ireland) Act 1850 took effect
in late July that year, and prescribed a census in
1851 to be taken on 30 March (the same day as
in Great Britain) that would be much the same
as the previous enumeration. And, again, it left
the decision whether or not to include additional
questions, in effect, to the discretion of Donnelly,
as Chief Commissioner, and William Wilde, who
had been retained as Assistant Commissioner.

forms were produced to facilitate better coverage,

specifically, in workhouses, hospitals, schools
and colleges, barracks, prisons, and at ports. The
process for completing the schedules was still left,
however, to the Enumerator, suggesting that either
at least this element of the enumeration process
was not unsuccessful in 1841, or that there was
no viable alternative.
2.1.57 The Commissioners were pleased to be
able to report to the Lord Lieutenant in June 1856
that

2.1.53 It is very probable that ill health, at that

time, prevented Larcom from taking up the post
of Registrar General and Chief Commissioner, but
he later went on to become Under Secretary for
Ireland and, in that role, was influential in ensuring
that the civil registration was extended to births
and deaths in an act passed in 1863.
2.1.54 More new questions were, indeed,
added to the schedule by Donnelly and Wilde. The
household form added both a column to identify
persons who were deaf and dumb or blind, and an
instruction to the education question that

The word Irish is to be added in this column

to the name of each person who speaks Irish,
but who cannot speak English; and the words
Irish and English to the names of those
who can speak both the Irish and English
languages.

2.1.55 The latter thus had anticipated, by twenty

years, the question on Gaelic speakers in the
Scottish census, and by thirty years the equivalent
enquiry in Wales. The 1850 Act continued,
however, to prohibit, specifically, any enquiry into
religion.
2.1.56 A new separate form was introduced to
collect information on persons who were

afflicted with insanity or idiocy,

and, reflecting the difficulties experienced in 1841

in enumerating institutions, a variety of other

owing to the zeal and diligence of the

Superintendents, and the good feeling of all
classes, we received within a reasonable time
from the 30th March, the returns from every
Constabulary district in Ireland complete, with
only very few exceptions, in which the parties at
first refused to fill in the forms. In these cases
a remonstrance from us had the desired effect
of procuring the returns, so that in no instance
had we to resort to the compulsory powers in
order to obtain the necessary information.5

2.1.58 Nevertheless, despite a seemingly

universal coverage, Donnelly and Wilde had to
report a much decreased population of some
6,552,385. This was almost 20 per cent fewer
than in 1841 a decline attributed to migration
and famine during the decade, following on from
the continued widespread failure of the potato
crop, particularly in 1841 and 1844, and, far more
catastrophically, in the years after 1845. Only in
the major urban areas of Dublin, Belfast, Galway,
Limerick, Waterford, Cork, Kilkenny and Drogheda
were increases in population recorded.
2.1.59 The 1861 Census, carried out on 7 April
that year (again, as elsewhere in the UK) counted
even fewer souls, just 5,798,967, but Donnelly and
Wilde were able to report, more encouragingly that

The causes which led to the diminution of

the population between 1851 and 1861 were,
happily, not of the twofold character to which
the decrease was to be attributed between
1841 and 1851. In the latter decade it may be
said to have been owing to the great mortality
45

Registrar General Northern Ireland Annual Report 2011

and emigration which originated with the famine

consequent upon the failure of the potato crop
for several years, commencing with 1845.
The diminution between 1851 and 1861 may
be said to have directly due to emigration, no
fatal epidemic, we are happy to state, having
prevailed during that period.6
2.1.60 Indeed, the total number of emigrants
from Irish ports in the decade leading up to the
1861 Census according to the Registrar Generals
returns was 1,208,350, being some 97 per cent
of all persons departing from UK shores in this
period. However, it should be remembered that the
Census was, at that time, still the only source of
mortality data in Ireland, and the Commissioners
acknowledged the estimates of the annual numbers
of deaths reported were likely to have been less
than the actual occurrences.
2.1.61 Despite this loss of population,
particularly among the agricultural classes, what
surprised the Commissioners was that the amount
of arable land had increased significantly from 13.5
million acres in 1841 to 15.5 million in 1861, and
made up some 74 per cent of the total land area
of Ireland at the time of the Census though they
noted that the amount of arable land that was
recorded as being under crops was more or less
unchanged in the decade 1851-1861.

2.1.62 In form and content, 1861 differed

significantly from the previous census only in that,
for the first time, the Act prescribing it allowed a
question on religion to be included. After making
specific provisions prohibiting the inclusion of such
a question in the legislation for the previous four
censuses, Parliament had to be convinced that to
allow it this time was both justified and safe. As
had been shown to be the case in Great Britain
in 1851, it was clear that there was no lack of
interest in the topic (though the reasons for the
enquiry in Ireland were very much different), and
the analysis of the responses to the question,
together with the enquiry into education, required
a whole volume to itself. Moreover, in Ireland the
question was included on a mandatory basis, and
became an integral component of the Irish census
throughout the rest of the century.
2.1.63 Both in the 1861 and 1871 Censuses,
Donnelly and Wilde gave much attention to the
analysis of religious profession and education
in combination. The General Report to the 1871
Census alone devoted some 90 pages to the
subject. In introducing the commentary, they noted
(rather smugly perhaps)

We now come to deal with the two elements

religion and education whose presence serves
more than anything else to distinguish the Irish
Census returns from those of Great Britain, and
to bring them, into harmony with the census
returns of all the world beside. It is no duty of
ours to speculate upon the reasons which have
determined the Legislature to maintain for Great
Britain an isolation, which, in this respect at all
event, entitles the statisticians of the world to
qualify the inhabitants of the sister Island as
Penitus toto divisos orbe Britannos.

There were sufficient reasons we doubt not;

but whatever those reasons may have been, it
is the fact, meanwhile, that every principality
in the world with a claim to civilisation
England and Scotland alone excepted takes
careful note and register of the confessional
differences of its subjects. The self-governing
Colonies, great and small, of our Empire are at

Figure 2 The leaving of Ireland

Emigrants leave Ireland, Henry Doyle

46

one in this respect with all the Governments of

Christendom; and while forbearing to discuss
the policy that excepts Great Britain from
the general rule a policy which we have no
commission to examine we cannot forbear
congratulating ourselves that Ireland has been
exempted from the application of that policy.7
2.1.64 They went on to note the importance of
having accurate counts of the numbers within each
religious denomination on which to reach

safe conclusions based on authentic figures

rather than having to rely, as Great Britain did, on

valueless conjecture.
2.1.65 Such was the considered importance
of obtaining accurate returns for the question on
religion in Ireland, that in both the 1861 and 1871
Censuses, where responses were unclear (for
example, when the simple response Protestant
was given) follow-up enquiries were instigated
in order to find out the particular denomination
in each case. However, this was not always
successful, and the Commissioners had to admit
that

In a very few cases the entries were facetious,

and although we afforded, through our queries,
an opportunity to the persons so returning
themselves, of making a more becoming return,
we did not think it necessary or judicious to
reproduce their silliness here.

But, in reassurance, Donnelly and Wilde noted:

The number of entries characterised by levity

or eccentricity was, exceedingly small, nor
can there be any doubt that, upon the whole,
the returns were made under a proper sense
of responsibility, and are therefore entirely
trustworthy.

2.1.66 The 1871 Irish Census very much

followed the 1861 model in almost all other
aspects also. However, in reporting the results
the Commissioners expanded Larcoms 1841
classification of dwelling type by subdividing the
fourth class and distinguishing those whose walls
were constructed from mud, from those where brick

or stone were used. They also decided to code and

tabulate occupations according to the classification
adopted in England and Wales.
1881-1901: few further developments
2.1.67 In their concluding remarks on the
General Report of the 1871 Census, Donnelly and
Wilde recommended that, in the absence of any
likelihood that the responsibility for the census in
Ireland would be taken up by a statistical office,
as was becoming the case in continental Europe,
the appointment of the Commissioners before the
next census should be made at least eight months
before Census day to allow sufficient time to
enable all the preliminary work to be completed. As
Commissioners, the 1871 Census had been their
third, and they recognised that

The difficulties attendant upon the

execution of so great a work as ours by a
temporary Commission were, no doubt,
materially lessened in our own regard by the
circumstances that the Commissioners brought
a large experience to the performance of their
duties7

2.1.68 The same Commissioners clearly

did not expect to direct a fourth census, and
indeed the new Registrar General, Dr Thomas
Wrigley Grimshaw, appointed as such in 1879 in
succession to William Malachy Burke, took up the
post of Chief Commissioner on 8 December 1880.
While this was not as early as the eight months
that Donnelly had suggested (Census Day would
be 3 April 1881), Grimshaw had in fact discussed
arrangements for the Census with the Treasury
as early as October 1879 even before the Bill for
Census had been presented to Parliament.
2.1.69 The 1881 Census (and the 1891 Census
also conducted by Grimshaw), again followed
the 1861 model. There were no new topics, but
Grimshaw had to report a continued decline in
population to the level of just 5.1 million - barely
above that of the figure estimated from the
incomplete 1813 Census though he noted that
the percentage decrease had declined in each
successive decade since 1841-51, being just 4.4
per cent since 18718.

47

Registrar General Northern Ireland Annual Report 2011

2.1.70 Grimshaw did, however, introduce a

special enquiry into agricultural holdings, including
the value of the land, at the suggestion of the
Statistical and Social Inquiry Society of Ireland as
a means of collecting information on the number of
people on farms of different acreages. He regarded
the benefit of this survey as

providing a standard of comparison for those

who desire to compare the results annually
published by the Registrar General in his report
on the Agricultural Statistics of Ireland, with the
actual condition of the agricultural holdings and
their occupiers.

2.1.71 Occupational coding was again carried

out along the lines of the Census in England
and Wales. However, the classification had been
significantly revised since the previous census,
and the delay in receiving the details of the new
classification from the English Registrar General,
George Graham, (just days before Census itself)
meant that considerable effort was involved in
quickly revising the Irish indexes and instructions
for the clerks in time for the coding to start on 1
June that year. However, Grimshaw experimented
with the introduction of a social classification of
occupations for the city of Dublin that preceded
that developed by T H C Stevenson in England and
Wales by some 30 years.
2.1.72 The publication of all the main tabulations
by the end of the following year was a considerable
achievement for the new Commissioners. But
Grimshaw was even more industrious at the
next Census of which the General Report9 was
submitted to the Lord Lieutenant General, the Earl
of Zetland, as early as 23 June 1892, less than 16
months after Census day. Again, no new questions
were included this time, but the special study into
agricultural holdings was repeated. Indeed, this
enquiry was becoming a distinctive and valuable
element of the Census in Ireland, not comparable
with any other part of the UK.
2.1.73 The 1891 Census was otherwise
uneventful - save for one thing. During the
enumeration there was an outbreak of smallpox
48

in the northern counties which necessitated

special disinfection measures to be taken for any
persons who were expected to handle the census
forms. Grimshaw held emergency discussions with
Medical Commissioner of the Local Government
Board, the Secretary of the General Post Office and
officials at the Stationery Office and Department
of Public Works in order to install disinfecting
equipment at the Census Office and to prepare
impervious covers in which to wrap each census
form in order to prevent the spread of the infection.
These were, happily, successful, and Grimshaw
was able to report that no case of smallpox
occurred among his staff during the course of the
Census.
2.1.74 As was the case in Great Britain, the
first census of the new century in Ireland in 1901
offered few changes, but in reporting on it in
June 1902, the Commissioners, who were now
Robert E Matheson, the new Registrar General,
his deputy Robert J Brew, and T J Bellingham
Brady, commented on the difficulty in planning
the enumeration resulting from the compete rearrangement of the territorial divisions following
the Local Government (Ireland) Act in 1898. This
required the necessary preparatory work to start
significantly earlier than usual, and, indeed, a
complete plan for the census was submitted to
the Government, together with an estimate of
costs, as early as 3 November 1898, some five
months before the Census Office was officially reestablished in April the following year.
2.1.75 But little else seemed to have troubled
the new Commissioners. Though they had to report
a continued decline in Irelands population from
4,704,750 to 4,458,775 in the ten years since
the 1891 Census, this percentage reduction at 5.2
per cent was only a little more than half the decline
over the previous decade10. The rate of population
decrease was slowing and would continue to do
so thereafter.
1911: the last all-Ireland Census
2.1.76 The last census (so it turned out) to
be carried out in what was an unpartitioned

Ireland, was, much like the 1901 Census had

been, little different in conduct and content to
its predecessors, other than the addition of the
same fertility enquiry that was being carried out
for the first time in Great Britain. The 50 pages of
the Commissioners commentary in the General
Report11, published in February 1913, was,
consequently, rather perfunctory; even the new
fertility data did not demand much attention and,
as in Scotland, there was no separate Report.
2.1.77 The number of forms and schedules
to be completed either by an Enumerator or the
householder had, however, now grown to almost
unmanageable proportions. There were some
27 different types of document that required
completion for each of the 208 enumeration
districts:
Form A. Family return
Form B1. House and building return
Form B2. Return of out-offices and farm-steadings
Form B3. Shipping return
Form C. Return of the sick
Form D. Return of lunatics and idiots, not in
institutions
Form E. Return of paupers in workhouses
Form F. Hospital return
Form G. College and boarding school return
Form H. Return of military, R. I. constabulary, or
metropolitan police, in barracks
Form I. Return of lunatics and idiots in public
institutions and private lunatic asylums
Form K. Prison, Bridewell, and police station return
Form L1. College and boarding school return
Form L2. Return of scholars attending schools
Form L3. Return of children on school roll
Form L4. University and college return
Form M1. Return of agricultural holdings
Form M2. Enumerators abstract of returns relating
to agricultural holdings
Form M3. Amalgamation of holdings in occupation
of same person

Form N. Enumerators abstract for a townland or

street
Form O1. Enumerators summary
Form O2. Enumerators summary of houses, by
classes
Form P1. Names of Enumerators
Form P2. Townlands, etc., in enumeration district
Form P3. Enumerators return showing material
discrepancies in the houses, and the population of
the townlands, etc
Form Q. Schedule of forms
Form R. Statutory declaration return
with copious notes and instructions accompanying
them. The processing of all the data contained
therein remained essentially a clerical operation.
Mechanical data processing would not be
introduced in any part of Ireland until the 1920s.
2.1.78 The Parliamentary debate caused little
fuss, certainly in comparison with the hours of
argument that were simultaneously ensuing over
the legislation for the census in Great Britain. Most
of the few points that were raised were resolved
amicably and with relatively little discussion. John
Boland, Irish Parliamentary Party MP for South
Kerry, for example, had, at the Bills second
reading, questioned the collection of information
on literacy relating to children as young as five;
he felt that such returns would swell the figures
to an extent that would mask the true level of
illiteracy in Ireland. He observed that in America
the minimum age for measuring illiteracy was ten.
At first, Augustine Birrel, the Chief Secretary for
Ireland, accepted the point and agreed to consider
the matter, but in Committee, two days later, he
explained that the tabular results would be able to
show levels of literacy by any age, thereby making
ready comparison with other countries possible.
2.1.79 Thomas Scanlan, Irish Parliamentary
Party MP for North Sligo, was concerned about the
continued use of members of the Irish constabulary
as Enumerators, citing instances in England when
the enquiries of the Distress Committee had largely
broken down because the investigations had been
49

Registrar General Northern Ireland Annual Report 2011

entrusted to the police, and noting that in Ireland

the constabulary was far more unpopular since it
represented, what he considered to be, a foreign
garrison. He tabled an amendment to have the Bill
require that the

officials to be appointed by the county

councils for the several districts under their
respective jurisdictions, and by the local
authority of each borough and urban district
which is without the jurisdiction of a county
council

should act as the Enumerators instead. Birrel

responded by reminding the House that the Royal
Irish Constabulary and the Dublin Metropolitan
Police had carried out the enumerations in Ireland
since 1841 without any particular obstruction
from the public, and that a Treasury Committee,
appointed in 1890 to consider the way in which the
Census could be improved, while hearing testimony
of some inefficiency in the manner in which the
enumeration was carried out in England, had,
thanks to Grimshaws evidence, singled out Ireland
for special commendation, and had reported that it
had been uncommonly well done.
2.1.80 The Registrar General had traditionally
commented on marriage and fertility rates in
the Irish Census reports (using his own vital
registration data), so the inclusion of the new
fertility enquiry was generally welcomed by
Members, particularly by Birrel himself, who saw
the benefit of collecting consistent information
across the UK. However, in reporting the reintroduction of the question on duration of
marriage, that had been asked in earlier Irish
censuses but had since been abandoned, he
sardonically observed

50

it was found, as anybody could have

foreseen, that it is rather difficult to rely upon
the value of statistics which you get from a
somewhat hazy memory. I could not myself tell
anybody how long I have been married without
considerable research, though my memory is
still strong enough to say with some confidence
the age and the number of my children.

2.2 The pre-war Northern

Ireland censuses
The delayed census in 1926
2.2.1
At the time when the preparations for the
Great Britain 1921 Census started towards the
end of 1919 particular emphasis was laid on the
need to harmonise the census questions across all
parts of the United Kingdom. To this end a Census
Joint Committee was entrusted with the task of
achieving the maximum degree of comparability
in census statistics in each constituent country.
Their efforts to do so were however balked by
the political conditions in Ireland that rendered it
impracticable to proceed with the census there as
planned.
2.2.2
Even at the time when the Census Bill
for Ireland was introduced in the Commons, there
were serious concerns about its viability. Joseph
Montague Kenworthy, Liberal MP for Central Hull
believed that most people in the south and west
of Ireland would certainly refuse to take part. He
asked the Attorney General for Ireland to tell the
House

how many tanks he will require for the

collection of the census forms,

2.2.3
Captain William Wedgewood Benn, Liberal
MP for Leith outside Edinburgh agreed that it would
be totally impracticable to do so. Nevertheless, the
Bill proceeded along its due course. In Committee
there was very little discussion on the content
of the census, but more on the risk of expending
government funds on such a risky venture. The
Rt Hon Sir Donald Charles Hugh Maclean, then
leader of the Liberal Opposition, in particular, did
not believe that any sane person would think for a
moment that there was any chance whatsoever of a
census taking place in Ireland. He said

If conditions arise in Ireland under which a

census can be taken, then there will be an
Irish authority for taking it, and not this House.
Ireland is now in a state of chaos, if not of de
facto civil war, and the Government are asking
this House to support them by giving authority
for the expenditure of 94,000.

2.2.4
And though the Bill was approved,
Macleans foresight proved astute. Following
threats from Dil irean - the parliament set
up in Southern Ireland in 1919 - to boycott the
census, the British government decided, on 4
April 1921 that it would take the advice of the
Census Commissioners in Ireland to postpone the
census there indefinitely. Just a month later the
Government of Ireland Act 1920 came into force
and separate jurisdictions were established in the
north and the south of the island. It was not until
10 November 1925 that separate legislation for
the census in Northern Ireland took effect.
2.2.5
In the course of the debate on the 1925
Bill for the first census in Northern Ireland, the
Minister for Finance, Hugh MacDowell Pollock,
noted that it had been nearly 15 years since the
last census in Ireland, and that intervening events
such as the Great War and the 1918-19 influenza
epidemic had affected the demographic picture
of the country, and thus required a census to be
taken. It was eventually held on 18 April 1926 on
the same day as the similarly delayed census in
the Republic of Ireland.
2.2.6
Northern Ireland continued to have
specific legislation for each new Census until the
1969 Act provided a legal framework similar to the
permanent 1920 Census Act in Great Britain.
2.2.7
The schedule for the 1926 Census in
Northern Ireland, conducted by its first Registrar
General, L A Bullwinkle, had basically the same
content as the 1911 Census except that the
fertility enquiry was dropped (as it had been in the
1921 Census throughout Great Britain). Moreover,
neither the question on Irish language nor the longstanding enquiry into agricultural holdings, so long
a feature of Irish censuses and which had been
the focus of a special study in 1911, were included
north of the border. Instead the Northern Ireland
form contained an additional question relating to
health insurance and an enquiry on dependency
in childhood and orphanhood, arising from the
particular concerns resulting from the loss of
fathers in the Great War.
2.2.8
Following the previous practice, the
services of the local police force, the now Royal
Ulster Constabulary, were again utilised in the
delivery and collection of the Census returns.

2.2.9
Data processing was, for the first time
in Ireland, carried out along the lines adopted in
Great Britain and utilised the punched card and
mechanical sorting machines first used in the
1911 Census there.
2.2.10 The Registrar General used the General
Report to explain the need for, and the intricacies
of, the Hollerith tabulating machinery that it had
been necessary to introduce in order to remove the
limitations (and reduce the costs) of the manual
ticking method of coding and tabulating the
information from the census returns.
2.2.11 Bullwinkle accepted that the only
satisfactory way of compiling the bulk of the
Northern Ireland census statistics was by machine
tabulation. Even though the amount of processing
that was required was only a fraction of that
in England and Wales or Scotland, Bullwinkle
acknowledged in his Report that, from the
experience of the 1921 Census in Great Britain, it
was

abundantly clear that all the tabulation,

excepting that of the housing and
accommodation particulars and the populations
of small areas, could be carried out more
quickly, accurately and cheaply by using modern
machine methods.12

2.2.12 Coding the responses into a series of

numeric codes was done in groups of topics, with,
for example, usual residence, birthplace and
nationality coded together, but where the more
complex topics of occupation and industry
were coded separately by different coders. After
this process the returns were passed to the card
punchers who transferred the coded responses
on to a unique card for each person, where the
position of the each punched hole represented a
specific question and code (Figure 3).
2.2.13 Two types of punching machines
were used; the gang punch was able to punch
mechanically batches of cards where there was
a common code for each data item (such as area
codes for the Enumeration Districts and place of
usual residence; and hand operated key punches
were used for the remaining data items. It was
through such holes that electrical contacts were
made when the cards went subsequently through
the counting machines, which identified and
recorded each item response at a rate of 24,000
cards per hour.
51

Registrar General Northern Ireland Annual Report 2011

Figure 3 The 1926 Census punched card

2.2.14 Bullwinkle recorded that all the card

punching was carried out by just seven young
women and a supervisor in the seven months
between October 1926 and April 1927. The
relatively small amount of data processing which
was not carried in this way was done by the more
traditional and laborious clerical ticking process.
1937
2.2.15 The next census in Northern Ireland
was again out of phase with the rest of the
United Kingdom. In an attempt to bridge the gap
between the 1926 Census and the UK-wide census
scheduled for 1941, W A Carson the second
Registrar General for Northern Ireland at first
planned to carry out a census in 1936 to coincide
with the then proposed mid-term enumeration in
Great Britain. But when the plans for the latter
were abandoned because of the general economic
crisis, Carson postponed the census in Northern
Ireland until 28 February 1937, when he carried
out a more limited enumeration by omitting not only
the questions on key topics such as occupation
and industry, but the long-running question on
infirmities.
2.2.16 No detailed national report was
published; the outbreak of the war in 1939 and
the resulting paper shortages meant that, after the
County Reports had been produced, the remaining

52

publication programme was put on hold. Carson

reported, in a much shortened Summary Report

After the close of hostilities the question of

issuing a final volume unless a new census is
then in preparation will be considered.13

But neither a final report nor any such new census

ever came to fruition.
The 1939 National Registration and identity cards
2.2.17 On 3 September 1939 World War II
broke out, and less than a month later the civilian
population was enumerated under the direction
of the three UK Registrars General, but under
separate legislation (the National Registration
Act 1939), so that a National Register might be
complied as a war-time general security measure.
2.2.18 Preparations for such a National Register
had already begun at the end of 1938 and were
virtually complete by April 1939, and had the
country still been at peace these preparations
would have been used for a 1941 Census.
2.2.19 Originally the Registration Regulations
prescribed that all persons present within the UK
and the Isle of Man on National Registration Day,
29 September 1939, were to be enumerated,
with the exception of those persons serving in the
armed forces or who were registered merchant
seaman, for whom a separate register was to be

kept. However, at the time of the Registration, no

requirement for such a mercantile marine register
had been identified, and merchant seamen were,
in the event, included in the National Register.
The information collected on each individual was
kept to the bare minimum (as the exercise was
not intended to serve any statistical function)
and referred only to sex, age, marital status and
occupation, together with some additional details,
where appropriate, regarding membership of
Naval, Military or Air Force Reserves, Auxiliary
Forces, or Civil Defence Services.
2.2.20 Some of the statistics for Northern
Ireland from the 1939 enumeration were published
in 1944 in a volume covering the UK and the
Isle of Man, but they were not comparable with
previous censuses, since the enumeration itself
was primarily an administrative exercise used for
issuing National Identity Cards, and as the basis
for a variety of other war-time measures such
as rationing and the deployment of labour in the
essential industries and services.

2.3 The post-war censuses:

1951-1971
1951
2.3.1
The first post-war census - the first in
which the census in Northern Ireland took place at
the same time as the rest of the United Kingdom
- was no exception to the general rule that more
suggestions for questions to be included are
received by the Census Office than can possibly
be contained on the questionnaire. Nevertheless,
the new Registrar General for Northern Ireland, L C
Mulligan, tried to accommodate as many demands
for new information as possible, and consequently
the topic content for the 1951 Census was far
wider in scope than any of its predecessors. The
fact that there had been an interval of 14 years
since the previous Northern Ireland census (nomid-term census was even contemplated for
1946), during which time there had been significant
legislative and social changes, no doubt justified a
larger enquiry.
2.3.2
Mulligan agreed with his counterparts
in England and Wales and Scotland (George
North and Edmund Hogan respectively) that the
content of the census in Northern Ireland should
be broadly the same as in Great Britain. Thus,
questions on age, marital status, relation to
head of household, usual residence, birthplace,
nationality, number of rooms, occupation and
industry were similar throughout the UK. The main
omissions in Northern Ireland were the fertilityrelated questions on date of marriage and number
of children born in marriage for women aged under
50, and the questions on age at which full-time
education ceased and employers address, and
the enquiry into the exclusive or shared use
household amenities such as piped water. On
the other hand the Northern Ireland census form
still kept the traditional voluntary question about
religion and included also a new question on length
of residence for persons born outside Northern
Ireland.

53

Registrar General Northern Ireland Annual Report 2011

2.3.3
Though representations were made to
retain the 1926 question on infirmities, Mulligan
decided to accept the view widely held by research
bodies that it was no longer possible to use
the census to collect accurate and consistent
information about physical and mental conditions.
The same conclusion had been reached in
England and Wales some thirty years earlier,
when the question was dropped after the 1911
Census. Indeed as early as the 1881 Census
doubts were being expressed by William Ogle,
the Superintendent of Statistics at the General
Register Office in London about the reliability of the
information recorded in the returns.
2.3.4
Today, however, there is an increasing
demand for the sort of information that only
questions on disability can provide, and the
problem of dementia for example is certainly an
issue of particular current national interest. But
it remains as challenging as it was in Ogles day
to include, on a self-completion form, questions
on disability which yield relevant and accurate
statistics.
2.3.5
In his General Report, Mulligan recorded
no significant change in the field operation from
the approach adopted in the previous census,
and members of the Royal Ulster Constabulary
acted once again as Enumerators. However, the
publicity campaign for raising public awareness
of the census was more extensive than any
yet. It had, of course, been long realised that a
successful enumeration is entirely dependent on
the ready cooperation of the public. As Mulligan
acknowledged

The standard attained in the tabulations

cannot rise above the standard set out in the
returns, and this in turn depends on the extent
to which people have been informed about the
benefits accruing from a census.14

2.3.6
Consequently, not only was the census
given wide publicity on the national and local
press, but there was also a flow of articles in
weekly and other periodicals. The BBC also made
a substantial contribution in ensuring that the
54

public was informed and receptive, and as many

as twenty radio features, totalling over three and
a half hours of air time, were broadcast covering
all three stations - the Home Service, and the
Light and Third Programmes - including the popular
Womans Hour. A four and a half minute slot on
Radio Newsreel was broadcast on Census night.
2.3.7
Neither were cinema audiences
neglected. When not watching the latest
blockbuster Abbott and Costello meet the
Invisible Man they were shown special newsreel
features interspaced with a cartoon produced by
the Central Office for Information.
1961
2.3.8
For the 1961 Census there was much
closer collaboration between the Northern Ireland,
England and Wales, and Scotland Census Offices.
Mulligan decided that, subject to the modifications
necessary to meet local requirements and
the needs of the Northern Ireland government
departments and other users, the topic content
should be harmonised with Great Britain so as to
ensure a high degree of comparability across the
UK as far as possible.
2.3.9
Consequently the list of questions
included in the 1961 Census was considerably
wider in scope than any of its predecessors. All
those asked in 1951 were repeated, although
in some cases there were slight changes in the
wording and application. Thus the questions on
sex, age, marital status, relationship to the head
of the household, usual address, birthplace,
nationality, occupation, industry and religion were
similar to those in 1951, except that the age
qualification for the marital status and occupation
and industry enquiries reverted to 16 or over
to be consistent with Great Britain and with the
censuses in Northern Ireland prior to 1951 (when
the questions were asked of those aged 14 or over
and 15 or over respectively).
2.3.10 There were also new questions, in line
with those being included in Great Britain, on
scientific and technological qualifications,

included at the request of the Ministry for

Science on the advice of the Advisory Council
on Scientific Policy, in order to establish,
amidst concerns about Britains brain drain, the
extent and location of the countrys scientific
manpower, and for whom special analyses were
produced;
length of stay at a persons usual address
if they had lived there more than one year,
or previous address if less than one year,
responses to which, when combined with other
data, provided information about the extent of
internal migration and the characteristics of
such migrants;
the fertility of married women, included at
the recommendation of the Conference of
European Statisticians and at the request of the
Department of Health and Local Government as
a key input into the study of genetics;
age at which full-time education ceased in
order to provide a general measure of level of
educational attainment; and
tenure of accommodation and the extent to
which piped water, cooking and toilet facilities
were available, in order to investigate the extent
and characteristics of private renting and council
housing for the purposes of informing housing
and town planning policy particularly in Belfast.
2.3.11 Additional information, recorded on the
form by the Enumerator, indicated the number of
rooms occupied by the household and the type of
accommodation, the total number of households
occupying the premises, and whether the building
was wholly or only partially residential.
2.3.12 In fact, the only information collected
in England and Wales which was not asked in
Northern Ireland was the persons workplace
address and a selection of key questions which
were asked in respect of those persons usually
resident at the address but who were temporarily
absent on Census night.
2.3.13 The field operation was carried out, as
was now customary, by members of the RUC,

despite the concerns of a number of Northern

Ireland MPs - expressed during the debates on the
Census Bill that it was no longer appropriate or
practicable for them to do so. The Northern Ireland
Minister of Finance, Terence ONeill, emphasised
the experience that the RUC had acquired in
performing the Enumerator duties and the benefits
that this would have to the overall quality of the
census. However, in his General Report of the
census in 1964, the new Registrar General Dr A
T Park noted that insufficient numbers of the RUC
could be released at the time of the Census in
the border areas (in April) and that enumeration
duties had to be undertaken by other appointed
persons15.
2.3.14 Nevertheless, with a common date
fixed throughout the UK, the census in Northern
Ireland benefited from the wider more general
publicity in the UK national press and other media
outlets. The importance of engaging the public
and encouraging them to participate in the census
was increasingly being recognised, and a variety
of media channels (the daily press, journals and
magazines, broadcasting, booklets and posters)
was used to demonstrate the national importance
of the event. Advantage was taken, in particular,
of the increased access to television by referring
to the census in both news items and in popular
programmes on both the BBC and commercial TV
channels, and by featuring short light-hearted fillers
between programmes during the run up to the
Census.
2.3.15 The local activities in Northern Ireland
included the issue of some 4,000 census posters
for display at schools, libraries, post offices,
police stations, transport depots and other
public buildings as well as at large industrial and
commercial business establishments. Almost
8,000 copies of an explanatory leaflet, The Census
of 1961, were distributed to a wide audience
including clergymen and such organisations as
chambers of commerce and young farmers clubs.
2.3.16 Data processing, in the main, developed
the punched card technology used in 1951 but
with a much larger 80-column card. But for the
55

Registrar General Northern Ireland Annual Report 2011

preparation of the final reports, the Registrar

General was able to use the services of an ICT
1500 electronic computer, operated under secure
conditions, by International Computers and
Tabulators Ltd at its London-based offices.
2.3.17 The use of a computer made it
practicable for the first time to have machinereadable output consisting of the precise figures
required for publication, tabulated in the way
that was required by users. For example, the
computer was programmed to derive automatically
social class and socio-economic group (SEG), on
the basis of a standard cross-classification of
occupation and employment status; and for the
purposes of the Fertility Report, where a husband
and wife were enumerated on the same return, the
computer automatically associated their records
and recorded the husbands SEG code on the
wifes record. In addition to this, the total volume
of results could be both increased and reproduced
for different output geographies, particularly for
small areas if required.
2.3.18 Equally important was the fact that the
quality of the results could now be better assured.
The census input records inevitably contained
errors originating both from respondents and
as the result of the coding and keying of data.
Previously, of course, some respondent errors
could be removed by clerical processing, but often
the tables produced by the punched card tabulation
system contained inconsistencies which could
not be detected by inspection alone. In 1951 and
earlier censuses, the subsequent tabulations
had then to be adjusted, in a somewhat timeconsuming process, so that the marginal totals
were consistent throughout.
Figure 4 The ICT 1500 computer

56

2.3.19 But despite this technology the speed

of producing the published reports was not
greatly improved; the final county report (for
Tyrone) appeared in July 1964, compared with
the final county report from the 1951 Census (for
Fermanagh) in September 1954.
Cost of the census
2.3.20 At a time, now, when the ever-increasing
cost of the census is a matter of serious public and
political concern, it is perhaps interesting to note
that such increases seem almost always to have
been the case. The total cost that Park estimated
for the 1961 Census was 105,732, being 1s.6d
per head of population, and therefore equivalent, in
new money, to about 3p per head per year over the
entire period of the census operation. But this was
more than double the total cost for 1951 Census.
2.3.21 The longer than usual gap of 14 years
since the previous census and the much reduced
size of the whole operation in 1937 means that
costs then (amounting to just 9,953) are not
comparable. A more realistic comparison should,
perhaps made with the cost of the 1926 Census,
at 18,126.
2.3.22 Generally though, the costs of the field
operation in these early Northern Ireland censuses
were artificially low compared with the situation
in Great Britain since, other than providing for
travelling expenses for the Enumerators, the
Registrar General was able to acquire the use of
the RUC as part of their constabulary duties.
1966: the only mid-term census
2.3.23 The legislation providing for the census
in Great Britain had, since its introduction in 1920,
always allowed for a mid-term census, should the
need arise. However, despite repeated calls from
the Royal Statistical Society and others to do so,
until 1966 no Registrar General had chosen to
exercise the option to conduct a mid-term census
relying instead on the decennial results to meet the
needs of users for population data throughout each
decade. But a rapidly changing socio-demographic
profile in the 1960s - with increasing birth and

migration rates, greater mobility and changing

employment patterns - together with the need for
information that would be more up-to-date than
that being provided by the delayed results from
the 1961 Census in Great Britain (the General
Report for example, was not published until 1968),
prompted the decision around the end of 1963, to
undertake, for the first time, a mid-decade census.
2.3.24 Thus censuses were taken throughout
Great Britain on 17 April 1966. To reduce costs,
however, it was proposed that, for the first time,
the census should be carried out on a 10 per cent
sample basis.
2.3.25 The case for a census in Northern Ireland
was equally as strong. However, the Northern
Ireland Census was to be a full enumeration.
The Bill for the Census was debated in Stormont
and concerns were expressed that a 10 per cent
sample would seriously limit the use of data
since such samples, with their inherent errors,
would prevent figures for areas of less than
50,000 population being published. The other big
difference in Northern Ireland was that the date of
the Census was deferred until 9 October.
2.3.26 The proximity of the 1966 General
Election (called by Harold Wilson on the 31
March) was a matter of some concern to all three
Registrars General. But because the additional cost
of postponement in Great Britain (estimated to be
as much as 350,000) was significant compared
with the total estimated cost of the Census itself
(2 million), the decision was made to stick with
the original date. In Northern Ireland, however,
there were also concerns to avoid the risk of
any disruption due to the parades and activities
to commemorate the fiftieth anniversary of the
Easter Rising on 24April1916. Both these factors
together led to the postponement of the Census in
Northern Ireland from 17 April to 9 October.
2.3.27 The 1966 Census questionnaire was
much shorter than the 1961 form and omitted
the individual person questions on birthplace,
religion, fertility, education and occupation (though
the question on industry was retained to reflect

the changing pattern of employment in a number

of new industries being established in Northern
Ireland). Nor were the questions on housing tenure
and amenities included as it was argued that these
would have changed very little in the five years
since the previous census. However, new questions
on address one year and five years before the
census and on workplace and transport to work
were introduced to gain a better understanding of
migration and travel to work flows.
2.3.28 The proposals were generally welcomed
by the Northern Ireland parliament when the
Census Bill was being debated, but concerns
were expressed, particularly by David Bleakey,
the Northern Ireland Labour Party MP for Belfast
Victoria, that the Census would not address the
issue of emigration, particularly of people qualified
in the teaching and medical professions and in
other skilled trades. In response, the Minister of
Finance, Ivan Neill, noted regretfully that, though
the government was looking for ways and means of
getting a fuller and more up-to-date picture of the
trend in migration:

the airport authorities will not cooperate

with us in the matter of taking the census of
persons leaving the country to remain overseas,
and until we find some way of doing the job with
the voluntary co-operation of those who are
leaving the picture will remain incomplete

2.3.29 In most other aspects the 1966 Census

operation followed very much the pattern of its
predecessors, and the new Registrar General, W G
Nichol noted this in his General Report.
1971: new legislation more questions
2.3.30 And nor did the 1971 Census throw
up anything new or controversial for Northern
Irelands first female Registrar General Miss
S D J Henderson to report in depth. In fact for
the first time since 1937 no general report was
produced. And this was perhaps surprising, since
a significant change in the legislative framework
for the census had taken place in 1969 with the
introduction of a permanent statue (the Census
Act (Northern Ireland) 1969), along broadly similar
57

Registrar General Northern Ireland Annual Report 2011

lines to the 1920 Census Act of Great Britain. This

allowed, particularly in an age of rapid change and
development, for a census to be carried out by the
Registrar General every five years. Now all that
would be required for a particular census to take
place in Northern Ireland, as elsewhere in the UK,
would be secondary legislation in the form of an
Order and Regulations, prescribing the information
to be collected and the arrangements for carrying
out the enumeration respectively.
2.3.31 In taking the second reading of the
Northern Ireland Census Bill on 14 May 1969, the
Minister of Finance, Herbert Kirk, simply moved
the proposal for the provisions of the legislation in
much the same way that the Rt Hon Dr Christopher
Addison, the Minister for Health had done at
Westminster in 1920, seemingly wishing to play
down the significant changes that the Bill was
introducing. Only F V Simpson, Northern Ireland
Labour Party MP for Belfast Oldpark, voiced a
concern over the time it took for the reports to be
published, and asked

In this age of computerisation would it not be

possible to produce a census report in a very
short time so that it would not be out of date by
the time it is made available to the public, which
has been the case in the past? With a new
Registrar General we expect great things.

2.3.32 However, the complexity of the task was

to mean that the Summary Tables for the 1971
Census were only published in 1975.
2.3.33 When the subsequent Northern Ireland
Census Order was approved, a number of new
questions found their way on the census form.
With the aim of maintaining as much harmony as
possible with rest of the UK, the new Registrar
General, J Y Malley, included, as part of the
fertility survey, a question on the dates of birth of
all children born alive to women aged under 60 a
move aimed at providing data to examine trends in
family size and spacing.
2.3.34 A new question was also included on
occupation a year before the census, to provide
data on occupational mobility. But no question was
58

included on part-time or full-time work; instead a

question on hours worked was asked of all persons
in employment.
2.3.35 To extend the scope of the information
collected on educational qualifications in order to
inform planning in the field of higher education and
the deployment of highly qualified manpower, a new
question was asked on the possession of certain
specified qualifications, broadly equivalent to A
level or Higher grade of the Scottish Certificate of
Education. But the 1961 question on age at which
full-time education ceased was dropped.
2.3.36 Following difficulties experienced in
1966 in enumerating households in converted/
shared accommodation, an attempt was made to
identify structurally separate dwellings by asking
if the household shared any room, hall, passage
or staircase with another. And questions on car
availability, garaging and means of transport to
work (for the largest part, by distance, of the
normal journey to the place of primary employment)
were included because of the concerns about
the need to monitor and control traffic in towns
following the Buchanan Report - a major report on
UK Transport Policy published in 196316.
2.3.37 A further change brought about by the
regulations emanating from the 1969 Census
Act was the role of the Census Enumerator.
The tradition of utilising members of the local
constabulary forces, stretching back to 1841,
was broken, and a new temporary census field
management structure was created, with Census
Supervisors appointed and trained to recruit
Census Officers and to take control of the field
operations locally and to recruit, in turn, their own
Enumerators. As a consequence the direct field
costs of the census in Northern Ireland increased
considerably.

2.4 The modern census:

1981-2001
2.4.1
As we progress to more recent censuses
their historical interest naturally diminishes but not
so the significance of the innovations they present.
With the authority and basic format of the census
now well-established, under the provisions of the
1969 Act, key developments in census taking
in Northern Ireland focused on improving and
assuring the coverage and quality of the statistical
results. There was also a need, despite the growth
in users requirements for information, to keep the
demands made upon the public within bounds by
making the form as easy as possible to complete.
These increased demands, and the increasingly
rapid changes in the social and technical
environments meant that a number of new features
have been introduced.
1981
2.4.2
After the 1979 General Election the UK
Government under Margaret Thatcher sought to
reduce the content of the 1981 Census by omitting
the proposed questions on ethnic group, school
qualifications, availability of car, and hours worked.
In contrast the Registrar General for Northern
Ireland E Boston made proposals for the Northern
Ireland census which escaped relatively unscathed.
However, the content of Northern Irelands second
census conducted under the 1969 Act differed
from the 1971 Census.
2.4.3
The question on tenure of
accommodation was expanded to include response
categories to distinguish between renting from
a public authority and a housing association
or charitable trust; the enquiry into household
amenities now covered domestic sewage disposal,
central heating, type of fuel used for heating, and
type of heating insulation (none of which were
included elsewhere in the UK) at the expense of the
previous questions on kitchen cookers and sinks,
and the supply of hot water; the 1971 question on
the garaging of cars was also dropped.

2.4.4
As throughout the rest of the UK, a
question on whereabouts on Census night was
now included in Northern Ireland (in order to
collect information on absent usual residents),
and the question on journey to work again adopted
the mode of transport categories asked in Great
Britain, but added a space for the respondent
to record the start time of the journey. But
the enquiries into 5-year migration, school
qualifications, apprenticeships, occupation one
year before the census, and hours worked were
omitted.
2.4.5
As in 1971, a large number of parttime field staff were recruited to undertake the
enumeration activities which included the delivery
and the collection of forms to households. At that
time the census was conducted in particularly
difficult circumstances amidst acute political
tension during the second Maze Prison hunger
strike. A campaign of non-cooperation, which
included the public burning of census forms was
conducted in some areas. Tragically, a female
Enumerator, Mrs Joanne Mathers, was shot
dead on the doorstep of a house while she was
collecting returns. Arising from this, special
arrangements were put in place, most notably the
public were asked to co-operate by sending in their
completed forms by post. The response from the
public was excellent but the enumerated count still
understated the actual population.
2.4.6 The Summary Report, published in
1983 under the pen of the new Registrar General
R McMurray, reported an enumerated population
of 1.49 million and an adjusted Census day
population of 1.51 million, implying that 1.3 per
cent of the population were not among those
enumerated17. However, a number of academics
and official statisticians then examined the extent
of estimated non-response, and compared the
Census returns with other sources such as child
benefit data, the Census of Employment, and
derived variables such as implied fertility rates.
This resulted in a range of estimates of the extent
of the under-enumeration. The accepted mid-year
estimate for 1981 (as currently adopted by NISRA)
59

Registrar General Northern Ireland Annual Report 2011

is 1.54 million suggesting a level of non-response

in the Census in Northern Ireland of the order of
3.6 per cent.
1991
2.4.7
The list of questions asked in the 1991
census was no shorter. Indeed all the topics
included on the 1981 Census form were repeated and some more.
2.4.8
A question on limiting long-term illness
was included at the request of the Department
of Health and local authorities to provide both
a means of helping to measure, generally, the
need for health and personal social services,
and, more specifically, to help improve the
formulae for allocating central government funds
to the health service. After a gap of 80 years
since the last health-related question had been
included in the 1911 Census, it was important
that, notwithstanding the strong demand for
information, a viable and acceptable question
should be devised. Consequently, thorough tests
of two questions - one on long-term illness and
the other on short-term illness were carried out by
the Registrar General for England and Wales. The
results from the long-term question correlated well
with the use of health services, and data users
were confident that the results could be put to
practical use.
2.4.9 The term-time address of students was
collected so that students could be allocated to
area of usual residence as defined in the annual
mid-year population estimates. The question on
hours worked, dropped from the census in 1981,
was reinstated. And for the first time, an attempt
was made to collect information from wholly
absent households, albeit on a voluntary basis, by
asking householders who were away from home on
Census night to complete a form on their return.
2.4.10 A question on Irish language was reintroduced for the first time since partition, so
that similar questions on Celtic languages were
now included in Wales, Scotland and Northern
Ireland. Furthermore, the format of three of the
GB Census questions were modified in Northern
60

Ireland in order to meet specific user requirements

for local information: an extra category of unpaid
work in a family business including a shop or farm
was added to the economic activity question;
and the education question ascertained levels of
educational achievement in specific fields of study.
Samples of anonymised records and the need for
confidentiality
2.4.11 A further innovation in the 1991 Census
was the production of microdata that was made
available to users in the form of two Samples of
Anonymised Records (SARs) covering the whole of
the UK. These were commissioned by the Economic
and Social Research Council.
2.4.12 The decision to issue these samples was
only taken (a) after the three Registrars General
had been assured that the measures taken to
protect the data and to prevent identification of
individual persons and households were sufficient,
and (b) after legal advice had been given about
the authority of the Registrars General to publish
such samples as abstracts under the provisions
of the Census Acts. The advent of computerised
processing and tabulation production had
demanded a whole new interpretation of the
wording of the 1920 and 1969 Acts, and it was
important to have this legally defined.
2.4.13 The SARs thus represented a major
milestone in census processing on a par with
the advances in outputs made possible by the
introduction of Hollerith technology in 1926 and the
computer in 1961.
2.4.14 Maintaining the confidentiality aspects of
the SARs was always a key factor in their design
and dissemination. Indeed, all three UK Census
Offices have always regarded the confidentiality and
security of the Census to be a matter of the gravest
importance, and went to even greater lengths
this time to assure Parliament and the public
alike about the confidentiality of the information
recorded and the security of the data processing
systems. This assurance included, as was now
the usual practice, commissioning an independent
review of the computing arrangements. The Data

Protection Registrar had expressed support for

such a review, and following a competitive tender,
the British Computer Society (BCS) was awarded
the contract to review the security arrangements to
protect against unauthorised access to personal
data for all activities from the receipt of the forms
in the processing sites, through all the processing
stages, to the subsequent removal and transfer of
the forms to permanent storage.
2.4.15 In its main findings the BCS reported that
the Census Offices arrangements were

fully in keeping with the very high standards

of confidentiality required, and that the plans
and intentions current at the time were
consistent with that standard.18

2.4.16 But even before the 1991 Census was

taken, the issue of confidentiality was brought to
a head requiring new legislation to be hurriedly
drafted. Two years before, parts of the 1911
Official Secrets Act were repealed, including, in
particular, section 2 on which the Census Offices
had relied until then to safeguard personal
information after the census had been taken.
2.4.17 A new Act was therefore required to
strengthen the confidentiality provision of the
Northern Ireland and GB Census Acts. There was,
fortunately, unanimous support for the measure
and the resulting Census (Confidentiality) Act 1991,
together with an equivalent Order in Northern
Ireland, was passed only days before the Census.
It provided for an offence of unlawful disclosure
of personal census information, and extended
the previous confidentiality provisions to include
post-enumeration census-related surveys and to
encompass any person employed for the purpose
of taking the census.

2001 Census Logo

2001: new challenges, new solutions
2.4.18 Under the supervision of the new and
current Registrar General, Dr T N Caven, the broad
strategy for the first census of the new century
was to base it on those successful methods
used for the 1991 Census, but to strengthen the
operation by the introduction of some innovatory
measures aimed at meeting the ever changing
needs of users, the societal environment and the
opportunities of utilising the rapidly developing
technologies offered by outside service provides.
2.4.19 A question on relationship to each
member of the household (rather than just to
the head of household) was included to provide
more detailed information on household structure
to assist in planning housing needs for multiple
families within households.
2.4.20 Recognising the increasing amount of
informal (and often unrecognised) help given to
people with ill health, the Census also included a
new question on the provision of unpaid personal
care aimed at improving the understanding of local
variations in the need for care and the pressure on
social services in an attempt to target resources
more effectively. This was linked to another new
question on general health, which, in addition
to the question on long-term illness, aimed to
allow people to assess their own health over the
preceding 12 months as either Good, Fairly
good, or Not good.
2.4.21 Questions on the lowest floor level of
accommodation and on the number of floor levels
in the accommodation were included to provide a
measure of the numbers of people and households
living in potentially unsuitable accommodation,
such as households with young children, or elderly

61

Registrar General Northern Ireland Annual Report 2011

residents, or people with long-term illness living

several floors above the ground.
2.4.22 While a specific question on religion was
being included throughout the rest of the UK for
the first time (following a necessary change to the
1920 Census Act), Northern Irelands long-standing
enquiry was expanded to include a question on
religion in which the person was brought up for
those who indicated in the main question that they
had no religion; this information would assist in the
monitoring of policies on equality issues.
Figure 5 The 2001 Census question on ethnic
group in Northern Ireland

2.4.23 But the most significant change to the

Northern Ireland census form was the inclusion of
a question on ethnicity. Such a question had, after
many years of research and testing been included
in the census in Great Britain for the first time in
1991. The GB question provided baseline figures
against which the Government was able to monitor
possible racial disadvantage and deprivation within
minority groups.
62

2.4.24 The question was not at that time

included in the census in Northern Ireland, where
user consultation had not indicated a sufficiently
strong enough case for such information. But its
undoubted success meant that its extension to
Northern Ireland in 2001 was broadly welcomed.
The classification of ethnic groups used in the
Census is now widely regarded as a standard for
inter-censal surveys and ethnic monitoring.
2.4.25 A key development in the field operation
was the use of postal services for the first time
as the primary means of collecting the completed
forms from households. This was a significant
change in data collection methodology, and its
implementation allowed a significant reduction
to be made in the field force necessary to carry
out the 2001 Census. The aim was to free up
resources in order to focus on those areas known
to be difficult to enumerate, with the aim of
minimising the differential undercount experienced
across the country ten years earlier. But it was also
to play a key role in enabling the enumeration to
be completed despite the catastrophic outbreak
of foot and mouth disease affecting farms and the
rural community not only in Northern Ireland but
across the whole of the UK generally, and which
threatened to disrupt the entire census operation.
2.4.26 The foot and mouth outbreak was
sufficiently serious for the government to warn
against travelling in rural areas, and for an early
General Election, originally intended for 3 May (just
five days after the scheduled date for the Census),
to be postponed. Nevertheless, the Census went
ahead as planned, though modified enumeration
procedures had to be quickly developed to conduct
the field work during the emergency. Close liaison
was maintained with the Department for Agriculture
and Rural Development and the several farming
agencies to ensure that the census field work
would not compromise the work being done to
contain the outbreak.

Figure 6 Foot and Mouth 2001

2.4.27 Special methods of delivery and

collection were quickly arranged to ensure as
complete a coverage as possible in the areas most
affected. Additional instructions were issued to
field staff on the revised arrangements for delivery
and collection of census forms where access to
premises was not possible or advisable.
2.4.28 These special arrangements worked
well, and there was no evidence that response
was seriously affected by the outbreak. Very few
incidents of alleged breaches of the restrictions
were reported, and even these turned out to be
largely unfounded. Furthermore, since the Royal
Mail had already established procedures for
handling mail in such emergencies, the decision
to adopt a post back methodology can be seen, in
hindsight, to be serendipitous.
The one number census
2.4.29 Every effort is made to ensure everyone
is counted in a Census. However, no Census is
perfect and inevitably some people are missed. As
in previous censuses, the numbers from the 1991
Census were not adjusted for under-enumeration,
but the mid-year population estimates series were.
Differences in the published population statistics
for 1991, some 1.578 million enumerated in the
Census and 1.607 million in the mid-year estimate,
were mainly due to the final estimate of the census
under-enumeration.
2.4.30 But during the extensive consultation
carried out prior to the 2001 Census, users of
census data were adamant that they wanted

one definitive set of estimates from the Census

which were adjusted for the estimated underenumeration, such that Census statistics were
consistent with the mid-year estimate (MYE) series.
The so called One Number Census (ONC) was the
innovative response to these problems, though
in many ways it was a natural progression from
the advances made in previous censuses: the
introduction of automatic editing and imputation of
missing variables in 1981; then the imputation of
absent households in the 1991 Census
2.4.31 The key to the ONC was comparing the
Census with a second large independent survey,
the Census Coverage Survey (CCS), that was
carried out three and half weeks after Census Day,
and then matching the records with those from the
Census itself using a combination of automated
and clerical processes to identify people and
households missed by the Census but found by the
CCS, and vice versa.
2.4.32 A new and ground-breaking methodology
was successfully implemented and for the first
time all the census results were successfully
adjusted for measured under-enumeration. It was
estimated that 95 per cent of the population were
covered in the Northern Ireland 2001 completed
census returns. A crude adjustment to the figures
based on the information, collected separately by
Enumerators, on occupied households that did
not make a return might have increased coverage
to around 98 per cent. It is clear that the ONC
enabled a more accurate population count giving a
more representative picture of the characteristics
of the population.
Electronic outputs
2.4.33 The 2001 Census was also the first
which based the planning of outputs on a strategy
that was agreed at the outset and set out as
government policy. It was based on the simple
premise that

the investment in the Census was only justified

if the results are made available speedily and in
a clear and useable form.19
63

Registrar General Northern Ireland Annual Report 2011

2.4.34 The key UK criterion to meet user needs,

was that outputs should be comprehensive and
comparable between areas. At the same time
NISRAs own key criteria were that they should: be
produced speedily and efficiently, be user-friendly,
accessible and intuitive, and should minimise the
risk of disclosing information about identifiable
individuals.
2.4.35 As a result, and in accordance with
governments policy on e-government, the bulk
of the outputs were disseminated by NISRA
through on-line data systems such as the NISRA
website and the Northern Ireland Neighbourhood
Information Service or NINIS [Link].
uk/ninis. Hardcopy reports were still produced for
laying before the Assembly and ease of reference,
however the dissemination of output at lower
geographical levels was exclusively carried out
through electronic media (CD products and via the
Internet).
2.4.36 By 2001, and after almost two hundred
years of census taking, the population of Northern
Ireland had increased from 1.380 million to a
record level of 1.685 million, though not before
falling to a low of 1.236 million in 1891. This
profile of change is in stark contrast with the
pattern of continuous, though not always steady,
growth in Great Britain, where population totals
increased over 5 times in the same period
(Figure7).
Figure 7 Populations of Northern Ireland (18212001) and Great Britain (1801-2001)
1,800

Potato Famine &

Mass Emigration

Population of Northern Ireland (thousands)

1,600

50

1,400
1,200

40

1,000
30
800
600

20
Population Northern Ireland (thousands)

400

1800

1825

1850

1875

1900
Census Year

64

10

Population Great Britain (millions)

200

1925

1950

1975

2000

Population of Great Britain (millions)

60

2.5 The 2011 Census

2.5.1
At the time of the
publication of this article, the
2011 Census was little more
than a year and half ago and the
bulk of the outputs are yet to
come, and so this account as
a history - effectively ends here. But perhaps we
may conclude by commenting on how the lessons
learned by the Census Offices from the several
reviews and evaluations of the 2001 Census,
together with changing social and legislative
environments, helped drive the innovations and
developments that were introduced into the 2011
Census carried out on 27 March last year.
2.5.2
The key issues raised in the reviews and
evaluations of the 2001 Census covered the need
to:
develop a high quality and up-to-date address
list to increase the efficiency of the delivery of
census forms;
develop field management and form-tracking
systems to enable better central control of field
processes and activities;
undertake more detailed and earlier
engagement with stakeholders to ensure,
for example, that the needs of the disabled
community are taken into account;
review whether or not the coverage survey
design is sufficient to identify under-enumeration
in the hardest-to-count areas;
review the cost-benefit trade-offs in aiming to
produce more timely outputs that are consistent
and harmonised across the UK; and
review the mechanisms to protect statistical
confidentiality without eroding the utility of the
data.
2.5.3
The subsequent design of Northern
Irelands 2011 Census introduced measures,
methodologies and processes to respond to each
of these as well as the need to reflect an ever more
ageing, migrant and mobile population, and an
increasingly less compliant society.

More questions
2.5.4
Not surprisingly, the 2011 Census
introduced yet more new questions to reflect
not only societal change generally and shifts in
Government policy, but also the requirement to
comply with new European legislation to provide the
European Union with a set of census outputs that
are harmonised not only within the UK but across
all 27 Member States. But such questions were
included only after extensive consultation, testing,
assessment and a full justification for national
purposes.
2.5.5
The previous long-standing census
question on whether or not households have
exclusive use of either a bath/shower and/or toilet
amenities was no longer regarded as providing
a sufficiently discriminative indicator in the
allocation of resources for housing development
and regeneration, and was now replaced with a
question on the type of central heating available
to provide a more useful indicator of basic housing
standards. The information will be used to facilitate
future research on fuel poverty and deprivation,
and on renewable energy and energy efficiency.
2.5.6
In a similar manner, a question on
adaptations to accommodation for health
conditions was also asked to evaluate the issue of
accessibility for many people, for example people
with disabilities.
2.5.7
In addition to the enquiry into the
knowledge of the Irish language, a similar question
on Ulster Scots was introduced. And questions on
main language and knowledge/use of English were
included in order: to provide an indication of areas
and communities where foreign language service
provision is necessary; to better understand the
diversity of the population; and, in particular, to
assess the impact of English language ability on
employment and other social inclusion indicators.
The ability to include British/Irish Sign Language
as a main language was introduced to respond to
the particular needs of the deaf.
2.5.8
In total, four specific individual questions
on health and care were included in the 2011

Census; most notably a new individual question on

specific long-term health conditions (e.g. deafness,
blindness etc.) was included.
2.5.9
Following the Civil Partnership Act 2004,
the traditional question on marital status was
expanded to include response categories for those
with a civil partnership status or former status.
2.5.10 A number of new questions were
introduced to help measure and analyse the
pattern of international migration flows a subject
that, with European Union expansion in the
intervening decade since the 2001 Census, has
aroused considerable interest. These included:
country of previous usual residence and month
and year of most recent entry into Northern
Ireland for those people who had lived outside
Northern Ireland for a period of one year or
more;
intended length of stay for those who had
arrived in the UK within a year of the census;
and
citizenship (as determined by asking a question
on the country of passport held)
2.5.11 But in order to reduce the burden on the
public, people who were visiting at an address on
Census night (including overseas visitors) were
required only to provide information on the basic
demographic questions; in previous censuses,
and particularly those before 1991 that were
carried out on a person present (de facto) basis,
such visitors had been required to complete a
full questionnaire at the address where they were
present. Full information on such visitors was,
however, still collected at their place of usual
residence on their return.
Post-out and the Census Address Register
2.5.12 A logical extension of the post-back
methodology adopted to collect the completed
census returns in 2001 was the employment
of a postal service provider for the delivery of
questionnaires. And so, following the successful
trialling of post-out as a delivery mechanism for
census forms in the 2007 Test, a decision was
65

Registrar General Northern Ireland Annual Report 2011

made by Dr T N Caven, the Registrar General for

Northern Ireland, to use the postal service as the
prime means of form delivery.
2.5.13 The rationale for this strategy was that
even with hand delivery in the 2001 Census,
Enumerators failed to make doorstep contact with
households at more than a third of addresses
and resorted to delivering the form through the
letter box. The use of the Royal Mail to perform
this activity was seen as an opportunity to direct
further field resources to those areas where a
more focused approach to follow-up activities was
necessary in order to improve response rates.
2.5.14 Clearly the prerequisite for this was
the availability of a comprehensive, high quality
address register. In the absence of a single
authoritative domestic address register, NISRA
worked with the Land and Property Services
Agency to develop an address register that met
Census requirements. Drawn from the Land
and Property Services Agency POINTER address
database that had been developed for Northern
Ireland jointly with the Royal Mail and local district
councils, the resulting register contained a list of
addresses linked to higher level geographic areas
for management, reporting and output purposes.
This associated every address with an accurate
positional reference, which was used during
the planning of field operations, the printing of
questionnaires and the production of outputs.
More outsourcing more confidentiality
2.5.15 The use of the Royal Mail, now for the
delivery as well as the collection of the census
forms, was not the only example of the benefits
gained by using specialist agencies in the Census.
NISRA, working with the Office for National
Statistics, started a major data processing
procurement in September 2005, and, following an
extensive evaluation a single supplier - Lockheed
Martin UK - was selected in August 2008, in time
for them to work on the 2009 Rehearsal.
2.5.16 The confidentiality and security
arrangements in place to manage not only the
outsourced contracts, but the traditional field
66

enumeration and all in-house activities at each site,

were subject, as was now usual, to an Independent
Information Assurance Review, which was reported
to Parliament in February 2011 and which
concluded that

. the public can be reassured that the

information they provide to the 2011 Census
will be well protected and securely managed.20

The online census

2.5.17 But not all forms were posted back.
One final, but very significant development in
2011 was the faculty for people to complete
their census form online. It had been estimated
from international experience that up to 25 per
cent of households might be prepared to provide
their census information online. Accordingly, the
Northern Ireland Census Office offered the option
of making a census return via the internet for the
first time.
2.5.18 But in planning for this, the actual take
up of this channel was uncertain. Such a facility
afforded the possibility of both encouraging
response among particular sub-groups of the
population who may otherwise be hard to reach,
and of achieving some data processing cost
savings. Furthermore, such an option provided
an opportunity to assess the efficacy of collecting
information in such a way for any future censuses
and surveys.
2.5.19 The online questionnaire had the
same questions as the paper form. A unique
reference number was placed on each paper
questionnaire which formed the basic security code
for accessing and completing an online return.
The address was also checked at the start of the
completion process to confirm identification. A
little disappointingly, perhaps, less than one in five
returns were submitted in this way, but the reaction
from those respondents who did so was generally
very favourable.
2.5.20 If only this technology had been available
to the Reverend James Whitelaw!

References
1

Census of Ireland 1841. Report of the

Commissioners appointed to take the Census
of Ireland for the Year 1841. HMSO, Dublin.

William Shaw Mason (1823). Abstract of the

Answers and Returns Made pursuant to an
Act of the united Kingdom, passed in the 55th
Year of the reign of George III, intituled An
Act for the taking an Account of the Population
of Ireland, and of the Increase or Diminution
thereof. London

James Whitelaw (1805). An Essay on the

Population of Dublin, being the Result of an a
Actual Enumeration taken in 1798 with Great
Care and Precision and Arranged in a Manner
Entirely New. Graisberry and Campbell,
Dublin.
Census of Ireland 1831 (1833). Abstract of
Answers and Returns under the Population
Acts 55 Geo. III chap 120, 3 Geo. IV chap
5, 2 Geo. IV chap 30, 1 Will. IV chap 19.
Parliamentary Papers, Session 1833,Vol.
XXXIX, London.

Census of Ireland 1851 (1856). Part VI,

General Report. HMSO, Dublin.

Census of Ireland 1861 (1864). Part V,

General Report. HMSO, Dublin.

Census of Ireland 1871 (1876). Part III,

General Report with Illustrative Maps and
Diagrams, Summary Tables and Appendix.
HMSO, Dublin.

Census of Ireland 1881 (1826). Part II,

General Report with Illustrative Maps and
Diagrams, Summary Tables and Appendix.
HMSO, Dublin.

Census of Ireland 1891 (1892). Part II,

General Report with Illustrative Maps and
Diagrams, Summary Tables and Appendix.
HMSO, Dublin.

10 Census of Ireland 1901 (1902). Part II,

General Report with Illustrative Maps and
Diagrams, Tables and Appendix. HMSO,
Dublin.
11 Census of Ireland 1911 (1913). Part II,
General Report with Tables and Appendix.
HMSO, Dublin.
12 Census of Population of Northern Ireland
1926 (1929). General Report. HMSO, Belfast.
13 Census of Population of Northern Ireland
1937 (1940). General Summary. HMSO,
Belfast.
14 Census of Population of Northern Ireland
1951 (1955). General Report. HMSO, Belfast.
15 Census of Population of Northern Ireland
1961 (1964). General Report. HMSO, Belfast.
16 Buchanan, Colin (1963). Traffic in Towns.
HMSO/Penguin, London.
17 Census of Northern Ireland 1981 (1983).
Summary Report. HMSO, Belfast.
18 British Computer Society (1991). 1991
Census of Population: Confidentiality and
Computing. Cm 1447. HMSO, London.
19 Economic Secretary to the Treasury, Secretary
of State for Scotland, and the Secretary of
State for Northern Ireland (1999). The 2001
Census of Population. White Paper, Cm 4253.
The Stationery Office, London.
20 Office for National Statistics, General Register
Office for Scotland, Northern Ireland Statistics
and Research Agency (2011). 2011 Census
Security: Report of the Independent Review
Team. ONS/GROS/NISRA.

67

Registrar General Northern Ireland Annual Report 2011

Annex A - Dates of Census

Annex B Registrar Generals

B1 - Registrar Generals of Ireland

1813
1821
1831
1841, June 6-7
1851, March 30-31
1961, April 7-8

W Donnelly

1844-1876

WM Burke

1876-1879

TW Grimshaw

1879-1900

Sir R.E. Matheson

1900-1909

Sir W. Thompson

1909-1926

1871, April 2-3

B2 - Registrar Generals of Northern Ireland
1881, April 3-4
1891, April 5-6
1901, March 31-April 1
1911, April 2-3
1926, April 18-19
1937, February 28 March 1

LA Bullwinkle

1922-1938

WA Carson

1938-1952

LC Mulligan

1952-1962

CF Darling

1962-1963

Dr AT Park

1963-1965

WG Nicholl

1965-1969

JY Malley

1969-1971

J Burns

1971-1974

Miss SDJ Henderson

1974-1976

JY Malley

1976-1978

EJA Boston

1978-1981

R McMurray

1981-1994

Dr TN Caven

1994-Date

1939, September 29
1951, April 8-9
1966, October 9-10
1971, April 25-26
1981, April 5-6
1991, April 21-22
2001, April 29
2011, March 27
NOTES
1

68

War-Time Population Enumeration and not full

Census