Population Trends Autumn 1999 - Office for National Statistics

Linking 1991 population statistics to the 1998 local government geography of 1 Great Britain Tom Wilson and Phil Rees University of Leeds

This article describes how 1991 population statistics and other data based on the 1991 Census areas can be produced for the 1998 local authority geography of Great Britain. Since 1991 a large number of boundary changes have taken place rendering comparisons of population data before and after the changes problematic. Re-basing 1991 statistics on the current set of boundaries allows real demographic change over the period since 1991 to be measured. The article focuses on two sets of population statistics: 1991 Census data and mid-1991 estimates.

I NTRODUCTION Since the last Census was taken in April 1991 a large number of changes have taken place to the boundaries of local authorities in Great Britain. Mostly this stems from the wholesale reorganisation of local government which occurred between 1995 and 1998,2 but it is also the result of a large number of minor boundary alterations made throughout the 1990s. The existence of a new local government geography has created the need to recalculate population statistics for pre-1998 years for the new local government areas so that valid comparisons can be made over time. The Office for National Statistics, the Welsh Office and the General Register Office for Scotland have produced some summary 1991 Census statistics and other demographic data for the new areas in a number of publications. 3,4,5 The emphasis of this article, however, is on how such reworked statistics can be produced. Through the use of look-up tables, it outlines simple methods for producing 1991 Census statistics, and mid1991 population estimates by single years of age, based on the new boundaries. A selection of the results are presented.

S TRUCTURAL AND GEOGRAPHICAL CHANGES At the time of the 1991 Census, much of Great Britain was covered by a two-tier system of local government: the upper and lower tiers were respectively counties and districts in England and Wales, and regions and districts in Scotland; single-tier areas existed in the metropolitan areas of England. In the early 1990s the government decided to initiate a restructuring process which was to replace the two-tier structure in many parts of the country with a single-tier. The account given here will simply summarise the geographical outcome of the reorganisation process; discussions of the process itself, the reasons behind it, and comments on its outcome can be found in a number of other articles. 6,7,8 The switch in Scotland and Wales to single-tier authorities took place

37

Office for National Statistics

Population Trends 97

Table 1

Autumn 1999

The local government structure of the Great Britain, 1991 and 1998

1991

1998

Local authority type

Structure

England

32 London boroughs City of London 36 metropolitan districts 39 counties containing 296 districts Isles of Scilly

single-tier single-tier single-tier upper tier lower tier single-tier

32 London boroughs City of London 36 metropolitan districts 34 counties containing 238 districts 46 unitary authorities Isles of Scilly

Wales

8 counties containing 37 districts

upper tier lower tier

22 unitary authorities

9 regions containing 53 districts 3 island areas

upper tier lower tier single tier

32 council areas

Scotland

Source:

Local authority type

Structure single-tier single-tier single-tier upper tier lower tier single-tier single-tier

single-tier single-tier

ONS. Gazetteer for reorganised statistical regions and local authorities in the United Kingdom.

on 1st April 1996; in England the new unitary authorities and revised counties were introduced in four phases on 1st April in each of the years from 1995 to 1998. Although it was originally thought that much of England would be covered in unitary authorities, in the end this was not the case: 34 out of the 39 shire counties were retained in structure, though not always in geography. Table 1 shows how the local government structure of 1991 compares with that of 1998; Figure 1 contains three maps of the new geography for respectively England, Wales, and Scotland. This change of structure was generally accompanied by boundary changes, although it is important to note that boundary changes and changes of status did not necessarily occur together. There are a number of new authorities, particularly in England, which are geographically equivalent to 1991 county districts. Similarly there are quite a number of areas which did not undergo a change in status between 1991 and 1998 but which nevertheless were subject to geographical alterations. Although these changes were mostly small in area, in demographic terms they can be significant. For example, the London borough of Brent’s mid-1991 population on 1991 boundaries was 248.6 thousand whilst on 1998 boundaries it was 242.3 thousand. The estimated population of the borough in mid-1997, on the new boundaries, was 249.6 thousand.4 The real increase in population between 1991 and 1997 was 3.0 per cent but if the change was incorrectly calculated using the different boundaries of 1991 and 1998 then the increase would work out as 0.4 per cent.

LOOK- UP TABLES To produce statistics for the 1998 geography from any dataset based on the 1991 Census geography it is necessary to know precisely how the two sets of areas relate to one another. To achieve this a look-up table is required. Geographical look-up tables can take many forms, but the two principal types are old-to-new and new-to-old. The former lists the old areas in some kind of order and then next to each old area the new area to which it belongs, whilst the latter lists the new areas in order and next to each new area all the old areas included within it. Examples of these two types of look-up table are shown in parts a and b of Table 2. Ideally the old areas should be at a spatial resolution fine enough to aggregate perfectly to the new areas, but sometimes this is not possible and population weights must used. A population weight is a proportion of an old (or new) area’s population that falls within the new (or old) area. For example, the 1991 Census ward PBGB, which at the time of the Census was in Ryedale district, North Yorkshire, is now split


38

between York Unitary Authority (UA) and the revised county of North Yorkshire. Using 1991 Census usually resident populations for enumeration districts it is possible to calculate the proportion of the Census population of PBGB in York UA as 0.573. A look-up table of 1991 Census wards to 1998 areas (with ward population weights) could be used to aggregate data only available at ward level (such as the 1991 Census Local Base Statistics). An example is given in part c of Table 2. Two look-up tables were used in the work reported here. The first, an old-to-new look-up table, is a two column file containing in the first column a list of 1991 Census enumeration districts (EDs) and output areas (OAs) for all of Great Britain in the standard Census alphanumeric order, and in the second column the (authors’) code for the 1998 local authority to which each 1991 ED or OA belongs. This look-up table was used to aggregate to the new geography the mid1991 small area estimates produced by the Estimating with Confidence project (see later). The second, a new-to-old look-up table, contains the same information but with the column of 1998 area codes in order. This was used to create the re-worked 1991 Census Small Area Statistics populations (five year age groups), and a version with ward population weights was used to create re-worked Census populations from the Local Base Statistics (single year age groups), which were used in the estimation of the mid-1991 populations for the new geography (see later).

REWORKED 1991 CENSUS SMALL AREA STATISTICS (SAS) DATA 1991 Census data is available to academic users on the MIMAS server at Manchester University9 and can be extracted by programs written in the special extraction software, SASPAC (Small Area Statistics PACkage). The method used by the authors took advantage of the fact that SASPAC allows the creation of new system files for user-defined geographical areas. In addition, it is possible in SASPAC to create these new areas by both adding and subtracting 1991 Census output zones at different levels of geography, rather than by simply adding together the data from lots of small areas. The advantage of this feature is that the confidentiality blurring applied to Census data at the ward and ED level (in England and Wales) and the equivalent postcode sector and OA level (in Scotland) is minimised. For example, if a 1998 local authority area consists of a 1991 district minus three EDs, then Census data for the new area is found by subtracting the data for the three (blurred) EDs from the district data, rather than by adding together hundreds of blurred figures from EDs.


Table 2

Extracts from three look-up tables linking 1991 Census areas and 1998 local government areas

a) 1991 to 1998

b) 1998 to 1991

c) 1991 to 1998 weighted

1991 Census ED code

1998 local authority code


1991 Census ED code

1991 Census ward code

PDFL01 PDFL02 PDFL03 PDFL04 PDFL05 PDFL06 PDFL07 PDFL08 PDFL09 PDFM01 PDFM02 PDFM03 PDFM04 PDFM05 PDFM06 PDFM07 PDFN01 PDFN02 PDFN03 PDFN04 PDFN05 PDFN06 PDFN07 PDFP01 PDFP02 PDFP03 PDFQ01 PDFQ02 PDFQ03 PDFQ05 PDFQ06 PDFR01 PDFR02 PDFR03 PDFR04 PDFS01 PDFS02 PDFS03 PDFS04 PDFS05 PDFS08 PDFT01 PDFT02 PDFT03 PDFT04 PDFT05 PDFT06 PDFT07 PDFU01 PDFU02 PDFU03 PDFU04

113 113 113 113 113 113 113 113 113 149 149 149 149 149 149 149 149 149 149 149 149 149 149 113 113 113 149 113 113 149 113 113 113 113 113 149 149 149 149 149 149 113 113 113 113 113 113 113 113 113 113 113

148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 148 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149 149

QAGH06 QAGH07 QAGH08 QAGH09 QAGH10 QAGH11 QAGH12 QAGJ01 QAGJ02 QAGJ03 QAGJ04 QAGJ05 QAGJ06 QAGJ07 QAGJ08 QAGJ09 QAGK01 QAGK02 QAGK03 QAGK04 QAGK05 QAGK06 QAGK07 QAGK08 QAGK09 QAGK10 NZFT01 NZFT04 NZFT05 NZFT06 NZGF01 NZGF02 NZGF03 NZGF04 NZGF05 NZGF06 NZGF07 PBFD01 PBFD02 PBFD03 PBFD04 PBFD05 PBFD06 PBFD07 PBFG01 PBFG02 PBFG03 PBFG04 PBFG05 PBFG06 PBFG07 PBFG08

PDFA PDFA PDFB PDFB PDFC PDFD PDFE PDFF PDFG PDFH PDFJ PDFK PDFL PDFM PDFN PDFP PDFQ PDFQ PDFR PDFS PDFT PDFU PDFW PDFX PDFY PDFZ PDGA PDGB PDGC PDGD PDGE PDGF PDGG PDGH PDGJ PDGK PDGL PDGM PEFA PEFB PEFC PEFD PEFE PEFF PEFG PEFH PEFJ PEFK PEFL PEFM PEFN PEFP

Notes: QA PD PE 113 148 149

= = = = = =

Autumn 1999


149 113 149 113 113 113 149 113 113 113 113 113 113 149 149 113 149 113 113 149 113 113 149 113 113 113 113 113 113 113 113 113 113 113 113 113 149 113 149 149 149 149 149 149 149 149 149 149 149 149 149 149

Population weight

0.254 0.746 0.551 0.449 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.369 0.631 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

1991 Census code for the Wrekin (Shropshire) 1991 Census code for Selby (North Yorkshire) 1991 Census code for York (North Yorkshire) authors’ code for North Yorkshire (1998 boundaries) authors’ code for Telford and Wrekin UA (1998 boundaries) authors’ code for York UA (1998 boundaries).

39



Figure 1a

Autumn 1999

The local government geography of England, 1998

North Tyneside

Newcastle upon Tyne Sefton

South Tyneside

Wirral

sley ow Kn er L iv

St Helens

po

Gateshead Sunderland

ol

Rochdale Bolton Wigan

Bury

Northumberland

Tameside

Tyne and Wear

n che

ster

Salford

Sheffield Hartlepool

Durham Cumbria

Bradford

Rotherham

Stockport

Ma

Trafford

D o n c a s t er

B a r n s l ey

Oldham

Redcar and Cleveland Middlesbrough Stockton on Tees Darlington

Leeds N o r t h Yo r k s h i r e

Calderdale Wakefield

York

Kirklees

East Riding of Yorkshire

L a n c a s h i re

Blackpool

Kingston upon Hull

West Yorkshire

Blackburn with Darwen

North Lincolnshire

Greater Manchester

North East Lincolnshire

South Yorkshire

Merseyside Walsall

Wolverhampton

Halton Warrington

D e r by s h i re

in g

ha

m

Sandwell Dudley B

rm

Cheshire

i

Solihull

Coventry

Lincolnshire

Nottinghamshire Stoke-on-Trent Nottingham Derby Staffordshire

Telford and Wrekin

L e i c e s t e r s h i re

18 26

Shropshire

16

24 21 27 31 28

5 33 30 2 6 3 15 10 12 14 1 4 7 11 23 17 13 8 9

Wo rce st

29 32

W

25

M id

la n d s

Warwickshire er

sh

Suffolk

N

Milton Bedfordshire Keynes

re shi ham

O x fo rd s h i re Swindon

City of Bristol North Somerset Bath and North East Somerset

Lu

king

Gloucestershire

South Gloucestershire

Cambridgeshire

a

th or

Herefordshire

19

20

ire

sh

on

t mp

B uc

City of London Camden Hackney Hammersmith and Fulham Haringey Islington Kensington and Chelsea Lambeth Lewisham Newham Southwark Tower Hamlets Wandsworth Westminster Barking and Dagenham Barnet Bexley Brent Bromley Croydon Ealing Enfield Greenwich Harrow Havering Hillingdon Hounslow Kingston Upon Thames Merton Redbridge Richmond upon Thames Sutton Waltham Forest

Peterborough

ire

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

Norfolk

Rutland

Leicester

22

Wi Sl Re West Berkshire Wo Bk

Wiltshire

Hertfordshire

Greater London Thurrock Southend-on-Sea Medway

Surrey

K e n t

Hampshire

Somerset

We s t S u s s e x

East Sussex

So D e v o n

Br

Dorset

Isles of Scilly

Isle of Wight Bournemouth

Cornwall

Torbay

Plymouth


40

Essex

Poole

Portsmouth

So Br Re Wo Wi Bk Sl Lu

Southampton Brighton and Hove Reading Wokingham Windsor and Maidenhead Bracknell Forest Slough Luton


Figure 1b

Autumn 1999

The local government geography of Scotland and Wales, 1998

41

Office f or National Statistics


Table 3

Autumn 1999

Scotland by five year age and sex groups. Details of the work of the EwC project are given in a recent Population Trends article,13 and the MIMAS website pages on the project explain how the data can be extracted from MIMAS (for academic users) or purchased from the EwC project itself.

Welsh speakers by unitary authority

Unitary authority Welsh name Abertawe Blaenau Gwent Bro Morgannwg Caerdydd Caerffili Casnewydd Castell-nedd Port Talbot Ceredigion Conwy Gwynedd Merthyr Tudful Pen-y-bont ar Ogwr Powys Rhondda, Cynon,Taf Sir Benfro Sir Ddinbych Sir Fynwy Sir Gaerfyrddin Sir y Fflint Sir Ynys MÙn Tor-faen Wrecsam Cymru

Percentage aged 3 and above who speak Welsh

English name Swansea Blaenau Gwent The Vale of Glamorgan Cardiff Caerphilly Newport Neath Port Talbot Ceredigion Conwy Gwynedd Merthyr Tydfil Bridgend Powys Rhondda, Cynon, Taff Pembrokeshire Denbighshire Monmouthshire Carmarthenshire Flintshire Isle of Anglesey Torfaen Wrexham Wales

13 2 7 7 6 2 18 59 31 72 7 8 21 9 18 27 2 55 14 62 2 14 19

This method was used to create SASPAC programs which produced the new geography SAS system files for all the 1998 local authority areas of Great Britain. These new system files are now available to academic users on MIMAS. Alternatively, and for non-academic users, the original SASPAC programs used to create the new system files can be downloaded from the first named author’s web page.10 The new system files allow the user to extract not just population counts, but any of the SAS tables or data cells. An example of this is shown in Table 3, which reveals the proportion of the population aged 3 and above who could speak Welsh in the new unitary authorities of Wales at the time of the 1991 Census.

REWORKED MID -1991 ESTIMATES While many users of demographic statistics will find geographicallyrebased 1991 Census populations useful, others require mid-year estimates. Population estimates for 1991 differ from the Census figures in three ways: • • •

they are for mid-year rather than Census night; they count students as resident at their term-time addresses rather than at their parental addresses; and they allow for deficiencies in the Census such as underenumeration, and are thus a better indicator of the population of an area.

Mid-1991 estimates are particularly important as they act as reference points for both earlier and later population estimates. Later estimates are often produced using the cohort component method, and earlier estimates can be adjusted slightly to allow for migration errors that have been introduced since the last Census.11,12 The mid-year figure is also used as an approximate population at risk (the denominator) for calculating calendar year demographic rates, for example fertility rates and mortality rates. The method used by the authors for producing the mid-1991 estimates made use of the Estimating with Confidence (EwC) project small area estimates and data from the Census Local Base Statistics (LBS). The EwC project involved the creation of the best possible mid-1991 estimates for wards and EDs in England and Wales, and postcode sectors and OAs in


42

The first stage of the method involved the aggregation of the small area estimates to the 1998 geography through the use of a computer program and an old-to-new look-up table. This gave estimates for the new areas by five year age group. The second stage, disaggregation to single years of age, was achieved by calculating conditional probabilities of being in a single year age group given membership of the relevant five-year age group (e.g. the probability of being aged 12 given membership of age group 10-14). These probabilities were estimated from the available single year of age detail from the Census. Counts of residents in households (Table L38) were used once adjustments had been made so that students were placed at their term-time addresses rather than their parental addresses (Tables L10 and L37). To calculate these LBS counts for the 1998 geography new SASPAC LBS system files were prepared in a similar manner to the reworked SAS. The only difference was that the LBS are only available down to the ward level so it was necessary to use a look-up table of wards to new areas (with wards weighted where they were divided between new areas). The method described above does of course involve some approximations. First, we assume that the EwC figures are correct. Second, there is a small timing difference between the Census information and the EwC estimates (21st April 1991 and 30th June 1991) which we ignore. In addition, the only Census counts by single years of age for all ages are in table L38 – counts of residents in households – which we have used as a proxy for all residents whether in households or in communal establishments. Furthermore, the student adjustment is not perfect because there is no information in the Census on boarding school pupils under 16. Despite these approximations, the mid-1991 estimates for the new areas presented in this paper are likely to be good estimates of the single year of age population figures for the vast majority of areas. Now that the methodology has been explained, some of the results of our estimation procedure are described. The age and sex characteristics of populations are usefully illustrated by means of population pyramids. These reveal features of an area’s population that would remain hidden in tables of numbers. Pyramids for Aberdeenshire, Conwy, the East Riding of Yorkshire, and Milton Keynes are shown in Figure 2. Aberdeenshire’s age-sex profile shows the relatively large numbers in the middle working years reflecting the importance of services for the off-shore oil industry in that region. In Conwy the population pyramid shows the impact of retirement migration in the large elderly age groups, and in addition one can see the differential mortality of men and women in these age groups. In the East Riding of Yorkshire unitary authority – the East Riding minus its large urban centre, Hull – the pyramid shows a fairly old population and a particularly marked post World War II baby boom. In contrast Milton Keynes has the most youthful population of all England’s new unitary authorities, a result of the significant in-migration of those in the working ages to the new town. The single year of age estimates used to create these pyramids, along with figures for all the other authorities in Great Britain (as listed in Table 4), are now available on the MIMAS and the first named author’s webpage. 9,10 Total mid-1991 populations for the new set of local authorities are listed in Table 4. Where local authorities existing in 1998 were also in existence in 1991, populations for the 1991 boundaries are given as well; where boundary changes have occurred to authorities that were in existence in 1991 these have been marked with an asterisk. As the table shows, the largest of the new English unitary authorities is the City of Bristol with a mid-1991 population of 379 thousand; Rutland is the smallest with just 33 thousand. In Wales the largest and smallest unitary authorities are


Figure 2

Population pyramids for each ethnic group, Great Britain, 1995-1997

(a) Aberdeenshire

(b) Conwy Females

Males 90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 -0.8 -0.6 -0.4 -0.2

0

Percentage of total population

0 Birth cohort

0.2

0.4

0.6

0.8

Age in completed years


Males 90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 -1

1931 1941 1951 1961 1971 1981 1991

Females

-0.6

-0.4

-0.2

0

0 Birth cohort

0.2

0.4

0.6

0.8

1


Males

90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 0

0 Birth cohort

0.2

0.4

0.6

0.8

1

●

●

90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 -0.8 -0.6 -0.4 -0.2

0


It is important to note that these are the estimates produced by the authors and are not the official figures produced by the national statistical offices. Minor discrepancies with figures published by ONS and GRO(S) exist for a small number of areas. Three factors can explain these differences. ONS used the EwC ward level population estimates. Where wards were split between 1998 local authorities Census data for parishes was used to give population weights.14 We used the EwC mid-1991 estimates at the level of EDs and OAs instead.

Females

90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 -1


respectively Cardiff and Merthyr Tydfil with populations of 300 thousand and 60 thousand. In Scotland the equivalent council areas are Glasgow City (631 thousand) and the Orkney Islands (19 thousand).

●

1921




Males


1911

-0.8


90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

1901

(d) Milton Keynes

90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 -0.8 -0.6 -0.4 -0.2

Females

90+ 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

1

(c) East Riding of Yorkshire

-1

Autumn 1999

0 Birth cohort

0.2

0.4

0.6

0.8

1


ONS used the EwC estimates to produce 1998 geography mid1991 estimates for England, but not for Wales. At the time of preparing the Wales figures the EwC data was only available for England.14 Instead, Census populations for small areas were aggregated to the new unitary authorities and then the student, timing and underenumeration adjustments were made to obtain the mid-year estimates. GRO(S) used the same method for Scotland.15 While ONS have published mid-1991 populations for the new local authorities, they do not make retrospective revisions to the population figures of local authorities which undergo boundary changes (e.g. many of the London boroughs).

43



Table 4

Autumn 1999

Mid-1991 local authority populations based on two different geographies

Local authority

1991 geography

1998 geography thousands

ENGLAND London London boroughs City of London * Barking and Dagenham * Barnet * Bexley Brent * Bromley * Camden Croydon * Ealing * Enfield Greenwich * Hackney * Hammersmith and Fulham * Haringey * Harrow * Havering Hillingdon Hounslow * Islington * Kensington and Chelsea * Kingston upon Thames * Lambeth * Lewisham * Merton Newham * Redbridge * Richmond upon Thames * Southwark * Sutton Tower Hamlets * Waltham Forest * Wandsworth * City of Westminster * Metropolitan districts Bolton Bury * Manchester * Oldham Rochdale Salford * Stockport * Tameside Trafford * Wigan * Knowsley * Liverpool St Helens * Sefton Wirral Barnsley * Doncaster Rotherham * Sheffield * Gateshead Newcastle upon Tyne North Tyneside South Tyneside Sunderland * Birmingham * Coventry * Dudley * Sandwell Solihull * Walsall Wolverhampton * Bradford * Calderdale Kirklees * Leeds * Wakefield *

48,208.1 6,889.9

48,208.1 6,890.4

4.1 146.2 299.9 219.4 248.6 294.7 181.7 319.2 281.8 263.2 213.6 187.9 156.2 211.8 203.8 232.5 236.8 209.1 173.5 145.3 137.5 256.6 240.8 171.8 221.3 231.2 164.3 227.2 171.4 168.1 217.7 265.3 187.7

5.6 155.9 300.3 219.4 242.3 295.1 181.7 319.4 285.9 263.2 212.4 188.3 156.0 211.4 204.7 232.5 236.8 206.0 172.5 145.6 138.1 255.0 241.9 171.8 220.4 223.6 166.2 228.3 171.4 167.6 217.3 265.7 188.1

262.9 179.1 438.5 219.6 204.8 230.9 288.3 219.8 215.8 310.9 156.9 480.7 180.9 295.2 336.0 224.4 293.3 254.9 529.3 203.1 278.2 195.5 157.2 296.4 1,006.5 305.6 309.4 294.8 201.2 263.4 248.5 475.4 194.0 381.5 717.4 316.2

262.9 179.7 434.1 219.6 204.8 233.0 290.5 219.8 217.5 308.3 156.7 480.7 181.3 295.2 336.0 225.1 293.3 255.5 528.7 203.1 278.2 195.5 157.2 297.5 1,016.2 305.1 309.9 294.8 201.7 263.4 249.9 475.0 194.0 381.1 718.0 316.0

–

164.8

Unitary authorities Bath and North East Somerset


44

Blackburn and Darwen Blackpool Bournemouth Bracknell Forest Brighton and Hove City of Bristol Darlington Derby East Riding of Yorkshire Halton Hartlepool Herefordshire City of Kingston upon Hull Leicester Luton Medway Middlesborough Milton Keynes North East Lincolnshire North Lincolnshire North Somerset Nottingham Peterborough Plymouth Poole Portsmouth Reading Redcar and Cleveland Rutland Slough South Gloucestershire Southampton Southend-on-Sea Stockton-on-Tees Stoke-on-Trent Swindon Telford and Wrekin Thurrock Torbay Warrington West Berkshire Isle of Wight Windsor and Maidenhead Wokingham York Two-tier counties Bedfordshire * Buckinghamshire * Cambridgeshire * Cheshire * Cornwall Cumbria Derbyshire * Devon * Dorset * Durham * East Sussex * Essex * Gloucestershire Hampshire * Hertfordshire * Kent * Lancashire * Leicestershire * Lincolnshire Norfolk Northamptonshire Northumberland North Yorkshire * Nottinghamshire * Oxfordshire Shropshire * Somerset Staffordshire * Suffolk Surrey * Warwickshire West Sussex * Wiltshire * Worcestershire

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

137.9 149.8 158.8 98.8 243.9 397.0 99.9 225.4 295.5 124.9 91.5 161.4 266.5 284.7 174.6 243.3 146.4 179.2 161.7 153.6 179.8 280.9 154.6 254.4 135.1 188.8 136.2 146.4 33.2 105.6 223.2 207.3 162.5 175.4 253.1 172.9 141.9 129.6 122.6 185.2 139.3 126.3 134.2 142.0 173.6

532.4 639.1 668.7 966.1 474.1 489.2 943.1 1,038.7 660.5 605.8 715.6 1,546.9 539.4 1,581.9 988.7 1,536.1 1,409.9 894.4 591.0 759.4 586.6 306.7 719.1 1,020.2 580.9 411.6 468.4 1,049.8 653.8 1,033.6 489.2 712.3 571.8 –

357.9 458.4 514.2 656.0 474.1 489.2 717.7 661.7 367.0 504.8 471.7 1,253.4 539.4 1,184.2 990.2 1,292.8 1,122.3 576.4 591.0 759.4 586.6 306.7 545.2 739.3 580.9 269.7 468.4 794.8 653.8 1,029.9 489.2 713.6 400.2 514.4


WALES Unitary authorities Abertawe (Swansea) Blaenau Gwent Bro Moorgannwg (The Vale of Glamorgan) Caerdydd (Cardiff) Caerffili (Caerphilly) Casnewydd (Newport) Castell-nedd Port Talbot (Neath Port Talbot) Ceredigion Conwy Gwynedd Merthyr Tudful (Merthyr Tydfil) Pen-y-bont ar Ogwr (Bridgend) Powys Rhondda, Cynon, Taf (Rhondda, Cynon, Taff) Sir Benfro (Pembrokeshire) Sir Ddinbych (Denbighshire) Sir Fynwy (Monmouthshire) Sir Gaerfyrddin (Carmarthenshire) Sir y Fflint (Flintshire) Sir Ynys MÙn (Isle of Anglesey) Tor-faen (Torfaen) Wrecsam (Wrexham)

2,891.5

2,891.5

– – – – – – – – – – – – – – – – – – – – – –

231.5 73.0 119.4 299.8 171.5 136.9 139.5 66.6 108.5 116.0 59.9 129.3 120.2 237.4 113.0 91.7 80.4 170.5 142.7 69.4 91.4 122.9

SCOTLAND Council areas Aberdeen City Aberdeenshire Angus Argyll and Bute Clackmannanshire Dumfries and Galloway Dundee City East Ayrshire

5,107.0

5,107.0

– – – – – – – –

215.0 216.5 108.3 93.5 48.4 147.7 156.6 124.3

Key points ●

●

●

●

●

●

The change in local government geography between 1991 and 1998 has presented the need to geographically rebase 1991 population statistics to the new boundaries. The 1991 and 1998 geographies can be linked by the use of look-up tables, and these can be one of two main types: old-to-new or new-to-old. The look-up tables used in this paper can be used to rework to the 1998 areas any type of data that is based on the 1991 Census geography. 1991 Census data can be reworked by the use of new-to-old look-up tables embedded in SASPAC programs. Mid-1991 population estimates by single years of age can be calculated for the new geography by aggregating the EwC estimates for small areas using an old-to-new look-up table , and then disaggregating to single year age groups using Census data. Further information on the methodology and results is available on the web sites listed in the notes below.

East Dunbartonshire East Lothian East Renfrewshire City of Edinburgh Eilean Siar Falkirk Fife Glasgow City Highland Inverclyde Midlothian Moray North Ayrshire North Lanarkshire Orkney Islands Perth and Kinross Renfrewshire Scottish Borders Shetland Islands South Ayrshire South Lanarkshire Stirling West Dunbartonshire West Lothian

1

A more detailed account of the work reported in this article can be found in Wilson T and Rees P. Look-up tables to link 1991 population statistics to the 1998 local government areas. Working Paper 98/5, School of Geography, University of Leeds (Leeds: 1998). Also included is a description of the iterative proportional

– – – – – – – – – – – – – – – – – – – – – – – –

110.5 84.9 86.3 439.5 29.4 143.1 349.4 631.2 204.1 91.6 80.2 84.2 139.0 329.0 19.6 127.6 176.6 104.1 22.5 113.5 304.8 81.5 97.8 146.3

Sources: 1991 geography figures: ONS. Mid-1997 population estimates for England and Wales. Monitor PP1 98/1. Registrar General for Scotland. Mid-1995 Population Estimates Scotland. 1998 geography figures: Authors’ computations. Notes The 1998 geography mid-1991 estimates are those produced by the authors, and not by the national statistical offices. – not applicable because authority not in existence * change of boundaries between 1991-98 (only applicable to authorities in existence in 1991)

2

3

4

5

6

7

8

9 10 11

12

NOTES

Autumn 1999

13

14

15

fitting method used to produce the Northern Ireland district council area populations. This paper can be viewed at http:// www.geog.leeds.ac.uk/wpapers/98-5.pdf. Some details of the reorganisation were given in Jackson G and Lewis C. Local government reorganisation in Scotland and Wales. Population Trends 83 HMSO (1996), 43-51. Office for National Statistics and the Welsh Office. Key Population Statistics 1991-1994: Welsh Unitary Authorities, HMSO (1996). Mid-1991 population totals for English local authority areas which underwent reorganisation in 1995-1997 are included in Office for National Statistics. Mid-1997 population estimates for England and Wales, ONS Monitor PP1 98/1 (1998). Mid-1991 population totals for Scottish council areas are given in Registrar General for Scotland. Mid-1995 Population Estimates Scotland. HMSO (1996). Johnston R and Pattie C. Local government in local governance: the 1994-95 restructuring of local government in England. International Journal of Urban and Regional Research 20 (1996), 671-696. Leach R. Restructuring local government. Local Government Studies (1994), 345-360. Chisholm, M. Some lessons from the review of local government in England. Regional Studies 29 (1995), 563-580. See: http://www.mimas.ac.uk/census/. See http://www.geog.leeds.ac.uk/pgrads/t.wilson/. Population Statistics, OPCS. Rebasing the annual population estimates. Population Trends 73 (1993), 27-31. Armitage B and Bowman J. Accuracy of rolled-forward population estimates in England and Wales, 1981-91. Occasional Paper 44, HMSO (London: 1995). Simpson S, Cossey R and Diamond I. 1991 population estimates for areas smaller than districts. Population Trends 90 (1997), 31-39. Personal communication from ONS Population Estimates Unit to the authors, 13th January 1999. Personal communication from GRO(S) to the authors, 8th September 1998.

45
