Beyond 2011: assessing alternatives to a traditional census Local Authority Research and Intelligence Association (LARIA) conference, 22 February 2012 Head of Demography Division National Records Scotland (NRS) NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Overview • Background • Statistical options • Use of admin data sources • Record linkage • How we protect privacy NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Key Drivers • Cost and viability of traditional census • User needs and expectations: quality and efficiency • New opportunities: technology and data • House of Commons Treasury Committee Report ‘Counting the Population’, 2008 NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
McLelland Review ‘Although sensitive to information protection and privacy issues, the citizen would expect that public bodies will share and move information across internal boundaries, particularly where there is an advantage to the citizen or the community when this happens’ NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Purpose and Scope • To explore the potential of improving the quality of population and socio-demographic statistics after the 2011 Census • To report in 2014 on feasibility research (phase 1), followed by a longer term programme of work to pilot, and implement an alternative model(s) Successful Scottish Spending Review bid, Sept 2011 • Assessing alternatives to the census • Wider data linkage framework to support social policy research Office for National Statistics (ONS) Beyond 2011 programme for England & Wales established April 2011 NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Beyond 2011 : Statistical options Traditional Census (long form to everyone) Census Rolling Census (over 5/10 year period) options Short Form (everyone), Long form (Sample) Headcount + Annual Survey (US model) Aggregate analysis Administrative (Intermediate) Sample linkage e.g. 1% of postcodes data options 100% linkage to create ‘statistical population spine’ Survey Address register + Survey option(s) NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Record Linkage • The traditional census involves a number of data collections and uses identified records and record linkage in the production process • Census uses: Scottish Longitudinal Study, Scottish Health and Ethnicity Linkage Study • Organisational and technical infrastructure for the traditional census can be applied • This project: a wider range of sources set up for other, administrative purposes which can be joined together NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Data sources being considered • NHS Central Register • Community Health Index • Higher Education Student Record • Further Education Student Record • School Census • Electoral Register • Department of Work and Pensions (DWP)/Her Majesty’s Revenue and Customs (HMRC) Customer Information System (CIS) • National Insurance No allocations to migrant workers • Surveys • Address registers (Census and One Scotland Gazeteer) • Other local data NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Beyond 2011 – statistical options FRAME SOURCES DATA ESTIMATION OUTPUTS All National to Small Area Population Data Address CENSUS Register Population estimates Adjusting for missing data and error Coverage Assessment Quality Population distribution incl. under & over-coverage measurement provides weighting - by survey and admin data? for attributes Household structure etc Adjusting for non response bias in survey (or sources) Socio demographic Household Attribute Data Attribute estimates Communal Longitudinal Establishments data Maintained national address Interactional gazetteer – provides frame for Analysis population data & surveys NATIONAL RECORDS OF SCOTLAND E.g. TTWA preserving the past; recording the present; informing the future
Beyond 2011 – statistical options FRAME SOURCES DATA ESTIMATION OUTPUTS All National to Small Area Population Data Address Admin Register Source Population Admin estimates Source Admin Source Adjusting for missing data and error ?? Coverage Assessment increasing later? Quality Population distribution incl. under & over-coverage measurement provides weighting Comm - by survey and admin data? for attributes Source Household Commercial structure etc sources? Adjusting for non response bias in survey (or sources) Surveys to fill gaps Socio demographic Household Attribute Data Attribute Socio demographic estimates Survey(s ) Communal Longitudinal Establishments data Maintained national address Interactional gazetteer – provides frame for Analysis population data & surveys NATIONAL RECORDS OF SCOTLAND E.g. TTWA preserving the past; recording the present; informing the future
Challenges in using admin data • Collected for specific purpose, rules and definitions used not usually consistent with the needs of estimating resident population • Coverage: geographic or demographic • Timing and currency: variable • Quality (completeness, accuracy, consistency): often unknown • No cross-systems consistency NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
National Health Service Central Register (NHSCR) population counts compared with NRS Mid-Year population estimates: W e s t L o th ia n W e s t D u n b a rto n s h ire S tirlin g S o u th L a n a rk s h ire S o u th A y rs h ire S h e tla n d Is la n d s S c o ttis h B o rd e rs preserving the past; recording the present; informing the future R e n fre w s h ire 2010, Council areas, % difference P e rth & K in ro s s O rk n e y Is la n d s N o rth L a n a rk s h ire NATIONAL RECORDS OF SCOTLAND N o rth A y rs h ire M o ra y Local Authority M id lo th ia n In v e rc ly d e H ig h la n d G la s g o w C ity F ife F a lk irk E ile a n S ia r E d in b u rg h , C ity o f E a s t R e n fre w s h ire E a s t L o th ia n E a s t D u n b a rto n s h ire E a s t A y rs h ire D u n d e e C ity D u m frie s & G a llo w a y C la c k m a n n a n s h ire A rg y ll & B u te A n g u s A b e rd e e n s h ire A b e rd e e n C ity S C O T L A N D 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% -2.0% e g a t % n u o c r e v O
NHSCR population counts compared with NRS Mid- Year population estimates: 2010, age, % difference 25% Glasgow 20% City 15% Overcount %age Aberdeen 10% City 5% Perth & 0% Kinross All Ages 0 - 4 5 - 9 10 - 14 15 - 19 20 - 24 25 - 29 30 - 34 35 - 39 40 - 44 45 - 49 50 - 54 55 - 59 60 - 64 65 - 69 70 - 74 75 - 79 80 - 84 85 - 89 90+ -5% Shetland -10% Islands -15% -20% Age groups NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
DWP CIS population counts compared with NRS Mid-Year W est Lothian W est D unbartonshire S tirling S outh Lanarkshire S outh A yrshire S hetland S cottish B orders preserving the past; recording the present; informing the future 2010, Council areas, % difference R enfrew shire P erth & K inross O rkney Islands N orth Lanarkshire NATIONAL RECORDS OF SCOTLAND population estimates: N orth A yrshire M oray M idlothian Local authority Inverclyde H ighland G lasgow C ity F ife F alkirk E ilean S iar E dinburgh, C ity E ast R enfrew shire E ast Lothian E ast D unbartonshire E ast A yrshire D undee C ity D um fries & G allow ay C lackm annanshire A rgyll & B ute A ngus A berdeenshire A berdeen C ity S C O T LA N D 10.0% 5.0% 0.0% -5.0% -10.0% -15.0% -20.0% e g a t % n u o rc e v O
Main conclusions • No source uniquely capable of providing a solution NHSCR CIS • Need to use in combination Other Local Data • Good understanding of the Electoral register School relevant quality Census HESA characteristics is a key Resident Population prerequisite NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
The research: process and objectives • Link each source with the 2011 census to help understand coverage and quality • Build a prototype ‘statistical population spine’ (a count by age, sex and small area) by merging two or more sources • Assess against a benchmark. – Two options: prototype is independent of the 2011 census (may not be possible for a while) or prototype uses census information (more likely in the short term) NATIONAL RECORDS OF SCOTLAND preserving the past; recording the present; informing the future
Recommend
More recommend
