Register-based Statistics WILEY SERIES IN SURVEY METHODOLOGY Established in Part by Walter A Shewhart and Samuel S Wilks Editors: Mick P Couper, Graham Kalton, Lars Lyberg, J N K Rao, Norbert Schwarz, Christopher Skinner A complete list of the titles in this series appears at the end of this volume Register-based Statistics Statistical Methods for Administrative Data Second Edition Anders Wallgren and Britt Wallgren Formerly of the Department of Research and Development at Statistics Sweden This edition first published 2014 © 2014 John Wiley & Sons, Ltd Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988 All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher Wiley also publishes its books in a variety of electronic formats Some content that appears in print may not be available in electronic books Designations used by companies to distinguish their products are often claimed as trademarks All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners The publisher is not associated with any product or vendor mentioned in this book Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose It is sold on the understanding that the publisher is not engaged in rendering professional services and neither the publisher nor the author shall be liable for damages arising herefrom If professional advice or other expert assistance is required, the services of a competent professional should be sought Library of Congress Cataloging-in-Publication Data Wallgren, Anders, author Register-based statistics : statistical methods for administrative data / Anders Wallgren and Britt Wallgren – Second edition pages cm Includes bibliographical references and index ISBN 978-1-119-94213-9 (cloth) Register-based statistics I Wallgren, Britt, author II Title HA31.23.W35 2014 519.5–dc23 2014003205 A catalogue record for this book is available from the British Library ISBN: 978-1-119-94213-9 Set in Times New Roman 11/12 pt by the authors 2014 Contents Preface Chapter xi Register Surveys – An Introduction 1.1 1.2 1.3 1.4 The purpose of the book The need for a new theory and new methods Four ways of using administrative registers Preconditions for register-based statistics 1.4.1 1.4.2 1.5 Basic concepts and terms 1.5.1 1.5.2 1.5.3 1.5.4 1.5.5 1.6 1.7 Chapter What is a statistical survey? What is a register? What is a register survey? The Income and Taxation Register The Quarterly and Annual Pay Registers Comparing sample surveys and register surveys Conclusions 10 10 11 13 14 16 20 23 The Nature of Administrative Data 25 2.1 2.2 2.3 2.4 2.5 2.6 25 26 27 29 30 32 32 34 36 Different kinds of administrative data How are data recorded? Administrative and statistical information systems Measurement errors in statistical and administrative data Why use administrative data for statistics? Comparing sample survey and administrative data 2.6.1 2.6.2 2.7 Chapter Reliable administrative systems Legal base and public approval A questionnaire to persons compared with register data An enterprise questionnaire compared with register data Conclusions Protection of Privacy and Confidentiality 37 3.1 38 38 39 41 41 43 44 Internal security 3.1.1 3.1.2 3.2 No text in output databases Existence of identity numbers Disclosure risks – tables 3.2.1 3.2.2 3.2.3 Rules for tables with counts, totals and mean values The threshold rule – analyse complete tables Frequency tables are often misunderstood CONTENTS vi 3.2.4 3.3 3.4 Chapter 47 4.1 47 53 53 54 56 57 58 59 60 60 62 63 64 65 70 72 74 4.2 4.3 A register model based on object types and relations 4.4 Standardised variables in the register system Derived variables Variables with different origins Variables with different functions in the system Using the system for micro integration Three kinds of registers with different roles Register systems and register surveys within enterprises Conclusions The Base Registers in the System 5.1 5.2 Characteristics of a base register Requirements for base registers 5.2.1 5.2.2 5.2.3 5.3 5.4 5.5 5.6 5.7 5.8 Defining and deriving statistical units Objects and identities – requirements for a base register Coverage and spanning variables in base registers The Population Register The Business Register The Real Estate Register The Activity Register Everyone should support the base registers Conclusions 77 77 78 78 80 81 83 88 93 94 98 101 How to Create a Register – Matching and Combining Sources 103 6.1 6.2 Preconditions in different countries Matching methods and problems 6.2.1 6.2.2 6.2.3 6.3 6.4 Chapter How to produce consistent register-based statistics Registers and time Populations, variables and time The variables in the system 4.4.1 4.4.2 4.4.3 4.4.4 4.5 4.6 4.7 4.8 The register system and protection of privacy The register system and data warehousing Organising the work with the system The populations in the system 4.3.1 4.3.2 4.3.3 Chapter 45 45 46 The Register System 4.1.1 4.1.2 Chapter Combining tables can cause disclosure Disclosure risks – microdata Conclusions Deterministic record linkage Probabilistic record linkage Four causes of matching errors Matching sources with different object types Conclusions 103 105 105 106 112 114 120 How to Create a Register – The Population 121 7.1 7.2 121 125 125 How should register surveys be structured? Register survey design 7.2.1 Determining the research objectives CONTENTS 7.2.2 7.2.3 7.3 Defining a register’s object set 7.3.1 7.3.2 7.3.3 7.3.4 7.3.5 7.3.6 7.3.7 7.4 Chapter Units and identities when creating primary registers Using administrative objects instead of statistical units Creating longitudinal registers – the population Conclusions 128 128 131 131 134 135 136 137 138 141 142 143 144 145 146 How to Create a Register – The Variables 147 8.1 147 148 149 150 151 151 152 153 154 157 158 159 160 161 161 165 169 The variables in the register 8.1.1 8.1.2 8.1.3 8.1.4 8.2 8.3 Exact calculation of values using a rule Estimating values with a rule Estimating values with a causal model Derived variables and imputed variable values Creating variables by coding Activity data 8.3.1 8.3.2 8.3.3 8.4 8.5 Variable definitions Variables in statistical science Variables in informatics Creating register variables – checklist Forming derived variables using models 8.2.1 8.2.2 8.2.3 8.2.4 8.2.5 Chapter Defining a population Can you alter data from the National Tax Agency? Defining a population – primary registers Defining a population – integrated registers Defining a calendar year population Defining a population – frame or register population? Base registers should be used when defining populations Defining the statistical units 7.4.1 7.4.2 7.5 7.6 Making an inventory of different sources Analysing the usability of administrative sources vii Activity statistics Activity data aggregated for enterprises and organisations Activity data aggregated for persons: multi-valued variables Creating longitudinal registers – the variables Conclusions How to Create a Register – Editing 171 9.1 171 173 175 178 180 181 181 183 184 185 185 186 191 192 Editing register data 9.1.1 9.1.2 9.1.3 9.1.4 9.2 Case studies – editing register data 9.2.1 9.2.2 9.2.3 9.3 Editing work within the Income and Taxation Register Editing work with the Income Statement Register What more can be learned from these examples? Editing, quality assurance and survey design 9.3.1 9.3.2 9.3.3 9.4 Editing one administrative register Consistency editing – is the population correct? Consistency editing – are the units correct? Consistency editing – are the variables correct? Survey design in a register-based production system Quality assessment in a register-based production system Total survey error in a register-based production system Conclusions CONTENTS viii Chapter 10 Metadata 10.1 10.1.1 10.1.2 10.1.3 10.2 10.3 10.4 10.5 10.6 193 Primary registers – the need for metadata Documentation of administrative sources Documentation of sources within the system Documentation of a new register Changes over time – the need for metadata Integrated registers – the need for metadata Classification and definitions database The need for metadata for registers Conclusions Chapter 11 Estimation Methods – Introduction 11.1 11.2 11.3 11.4 11.5 Estimation in sample surveys and register surveys Estimation methods for register surveys that use weights Calibration of weights in register surveys Using weights for estimation Conclusions Chapter 12 Estimation Methods – Missing Values 12.1 12.2 12.3 12.4 12.5 Make no adjustments, publish ‘value unknown’ Adjustment for missing values using weights Adjustment for missing values by imputation Missing values in a system of registers Conclusions Chapter 13 Estimation Methods – Coverage Problems 13.1 Reducing overcoverage and undercoverage 13.1.1 13.1.2 13.2 13.3 13.4 Coverage problems in the Population Register Coverage problems in the Business Register Estimation methods to correct for overcoverage Undercoverage in the administrative system Conclusions Chapter 14 Estimation Methods – Multi-valued Variables 14.1 14.2 Multi-valued variables Estimation methods 14.2.1 14.2.2 14.2.3 14.2.4 14.2.5 14.2.6 14.3 14.4 Application of the method Linking of time series using combination objects 14.4.1 14.4.2 14.5 Occupation in the Activity and Occupation Registers Industrial classification in the Business Register Importing many multi-valued variables Consistency between estimates from different registers Multi-valued variables – what is done in practice? Additional estimation methods Linking time series Changed industrial classification in the Business Register Conclusions 193 194 194 195 195 196 197 198 200 201 202 203 204 207 208 209 210 214 215 218 220 221 221 221 222 224 226 228 229 229 232 232 236 238 242 245 247 251 254 254 256 258 CHAPTER 16 Conclusions The previous chapters in this book contain many proposals for change New terms and new methods have been presented with the aim that register systems and register-based statistics can be developed and function in a better way than they today Dillman (1996) is rather pessimistic regarding innovation and change in government survey organisations, especially when dealing with nonsampling issues One reason is the gap between operations and research cultures We agree with Dillman that it is difficult to bring about change To change a system of surveys is even more difficult, as many managers responsible for different surveys must decide together what changes should be made The implementation of new methods must be supported not only by managers, but also by a dialogue between the researchers developing new methods and those working with the surveys that should be improved The methods we propose in this book have been developed while discussing register issues with those operating register products We have spent many hours in seminars and study groups to promote new ideas and methods This implementation work will be necessary in all statistical offices where register-based statistics will be developed A new approach is necessary A new approach towards administrative data is necessary: – There should be no prejudice that administrative data are of bad quality If we compare the quality of the huge amounts of administrative data that tax authorities collect via tax forms from individuals and enterprises with the quality of the same data collected by the statistical office, we must admit that the tax authorities collect the better data Scheuren and Petska (1993) are of the opinion that ‘the detailed income and expenditure data on tax returns are generally regarded as more reliable than similar survey data’ – On the other hand, administrative data should not be used as they are They should be processed so that they can be used for statistical purposes The most important part of this processing is the integration of many sources A new approach towards registers and statistical science is necessary Some statisticians say there are no special methodological issues related to register statistics, and that there is no difference compared with ordinary censuses This is a misunderstanding caused by their lack of awareness of the methodological issues which are unique to register surveys The integration phase in register surveys and the Register-based Statistics: Statistical Methods for Administrative Data, Second Edition Anders Wallgren and Britt Wallgren © 2014 John Wiley & Sons, Ltd Published 2014 by John Wiley & Sons, Ltd 298 CONCLUSIONS methods used here largely determine the quality, and they have no similarity with the methods used for censuses Therefore, the development of register-based statistics should be recognised as an important field for statistical science The register system Administrative data from many sources are used to create a system of coordinated statistical registers This register system can be used to produce register-based statistics, create new registers, and create frames for sample surveys or censuses The register system should also be used for quality assessment of both sample surveys and register surveys If the register system has been created in the right way, it will be an important factor promoting consistency and coherence between all surveys conducted at a statistical office Even countries which are new as producers of register-based statistics will benefit from coordinating their registers into one system It should be noted that we propose that one system is created for all surveys If different subsystems are created (e.g one system for social statistics and one system for economic statistics), then it will be difficult to combine data from these two systems Productivity by industry can be estimated by dividing: – value added by industry, from a register survey based on the Business Register, with – hours worked by industry, from the Labour Force Survey, which is a sample survey based on the Population Register To achieve good quality here, the register system must be one coordinated system, and the estimation methods used must take into consideration that industry is a multi-valued variable Integration errors, in this case aggregation errors, can give rise to inconsistencies between industry in the Business Register and industry in the Labour Force Survey Good coverage and consistency are the important advantages of register-based statistics, if the register system has been created according to the principles in Chapters 4–9 All surveys based on the register system can benefit from this, and the National Accounts will gain more consistent data without undercoverage Frame errors Twelve of the 13 chapters in Cochran (1963) are devoted to sampling errors In the last chapter, Cochran mentions measurement errors and nonresponse During the last few decades, much effort has been devoted to nonsampling errors; both measurement and nonresponse issues are today regarded as central issues in survey methodology In the book by Särndal and Lundström (2005), 13 out of 14 chapters are devoted to nonresponse issues; frame errors are mentioned in the last chapter Today, there are no established methods for handling frame errors We believe that this kind of nonsampling error has been overlooked and that the errors can be substantial Development in this field is necessary, and these errors can only be reduced by register-statistical methods If we learn how to create registers with good coverage, all surveys using these registers will benefit from the good coverage CONCLUSIONS 299 The first step is to become aware of the frame errors At a statistical office where sample survey theory is the predominant paradigm, registers are used to produce frames and thereafter data are collected As a rule, the quality of the frame population will never be known Instead, new frames will be created followed by a new round of data collection A statistical office, where those responsible for, say, a business survey want to use administrative data, may follow the same procedure – except that instead of sending questionnaires to the sampled enterprises, they use administrative data Response burden and costs will decrease, but the frame errors will be the same If administrative data are used in this restricted way, the most important quality of administrative sources has not been used – the capacity for good coverage If registers are used to create both frames and calendar year populations, then it will be possible to become aware of the frame errors The preliminary estimates for the sample surveys based on frames can also be revised with information from the calendar year register, and the methods used to create frames can also be improved so that frame errors become smaller What more is needed? Apart from developing the existing registers, staff at a statistical office should constantly be discovering new administrative sources that can be used to create new statistical registers and products New types of registers and databases that are created outside the public sector may also be relevant sources in the future Individuals and enterprises leave numerous electronic tracks that are stored in databases by private enterprises These new kinds of data are sometimes called ‘Big Data’ When statistical agencies want to use these sources, there will be an introductory process that will resemble the process that took place when administrative data from the public sector started to be used for statistics production Can we use this kind of data? Can we gain access to such new data? How we protect privacy? Methodology work, negotiations and legislation will be required again These types of Big Data sources not originate from relevantly defined populations; the definition of an enterprise’s database is determined by the enterprise’s contacts with their customers, suppliers, etc National statistical offices with developed register systems could possibly create the relevant populations that will be required when these new sources are used for statistics production Big Data issues are discussed by Ploug (2013) and Elias (2013) We suggest that the potential of administrative data from the public sector should be utilised first, before resources are spent on new kinds of Big Data The ability to structure databases for statistical purposes and to analyse the data taken from administrative systems in a statistically meaningful way will be skills that are required in many new fields Register-statistical skills are therefore also required outside government survey organisations Universities and higher education must pursue research and provide teaching on register surveys This teaching and research should relate to society as well as enterprise register-based statistics References Argüeso, A and Vega, J (2013) A population census based on registers and a ‘10% survey’ Paper presented to the 59th ISI World Statistics Congress, Hong Kong, Session STS063 Berka, C., Humer, S., Moser, M., Lenk, M., Rechta, H and Schwerer, E (2012) Combination of evidence from multiple administrative data sources: quality assessment of the Austrian register-based Census 2011 Statistica Neerlandica, 66(1), 18–33 Biemer, P (2010) Total survey error – design, implementation and evaluation Public Opinion Quarterly, 74(5), 817–848 Biemer, P and Lyberg, L (2003) Introduction to Survey Quality Hoboken, NJ: John Wiley & Sons, Inc Carfagna, E and Carfagna, A (2010) Alternative sampling frames and administrative data What is the best data source for agricultural statistics? In R Benedetti, M Bee, G Espa and F Piersimoni (eds), Agricultural Survey Methods, pp.45– 61 Chichester: John Wiley & Sons, Ltd Cochran, W G (1963) Sampling Techniques, 2nd edn New York: John Wiley & Sons, Inc Commission of the European Communities (2009) Communication from the Commission to the European Parliament and the Council on the production method of EU statistics: a vision for the next decade Brussels 10.8.2009 COM(2009) 404 final Daas, P and Arends-Tóth, J (2012) Secondary Data Collection, Statistics Methods (201206) The Hague: Statistics Netherlands Daas, P., Ossen, S and Tennekes, M (2010) The determination of administrative data quality: recent results and new developments Paper presented to the European Conference on Quality in Official Statistics 2010, Helsinki Daas, P., Ossen, S., Tennekes, M., Zhang, L.-C., Hendriks, C., Foldal Haugen, K., Cerroni, F., Di Bella, G., Laitila, T., Wallgren, A and Wallgren B (2011) Report on methods preferred for the quality indicators of administrative data sources Second deliverable of work package of the BLUE Enterprise and Trade Statistics project Register-based Statistics: Statistical Methods for Administrative Data, Second Edition Anders Wallgren and Britt Wallgren © 2014 John Wiley & Sons, Ltd Published 2014 by John Wiley & Sons, Ltd 302 REFERENCES Daas, P., Ossen, S and Tennekes, M (2012) Quality report card for administrative data sources including guidelines and prototype of an automated version Third deliverable of work package of the BLUE Enterprise and Trade Statistics project De Waal, T (2009) Statistical data editing In D Pfefferman and C.R Rao (eds), Sample Surveys: Design, Methods and Applications, Handbook of Statistics Volume 29A, pp 187–214 Amsterdam: Elsevier Deville, J and Särndal, C-E (1992) Calibration estimators in survey sampling Journal of the American Statistical Association, 87, 376–382 Dillman, D (1996) Why innovation is difficult in government surveys Journal of Official Statistics, 12(2), 113–124 Elias, P (2013) Global integration of new forms of data: problems and possibilities Paper presented to the 59th ISI World Statistics Congress, Hong Kong, Session STS027 Eltinge, J., Kozlow, R and Luery, D (2003) Imputation in three federal statistical agencies Paper prepared for the Federal Economic Statistics Advisory Committee, October Eltinge, J., Biemer, P and Holmberg, A (2013) A potential framework for integration of architecture and methodology to improve statistical production systems Journal of Official Statistics, 29(1), 125–145 Eurostat (1997) Proceedings of the Seminar on the Use of Administrative Sources for Statistical Purposes, 15–16 January 1997 Luxembourg: Office for Official Publications of the European Communities Färnström, A (2013) The policeman and the statistician Paper presented to the 59th ISI World Statistics Congress, Hong Kong, Session IPS071 Federal Committee on Statistical Methodology (2005) Report on Statistical Disclosure Limitation Methodology Statistical Policy Working Paper 22 Washington, DC: Statistical Policy Office, Office of Information and Regulatory Affairs, Office of Management and Budget Fellegi, I.P and Sunter, A B (1969) A theory for record linkage Journal of the American Statistical Association, 64, 1183–1210 Goossens, H (2013) Building the statistical warehouse to improve statistics Paper presented to the 59th ISI World Statistics Congress, Hong Kong, Session IPS023 Granquist, L and Kovar, J (1997) Editing of survey data: how much is enough? In L.E Lyberg, P Biemer, M Collins, E.D de Leeuw, C Dippo, N Schwarz and D Trewin (eds), Survey Measurement and Process Quality New York: John Wiley & Sons, Inc Greijer, Å (1995) Overcoverage of foreign born in the Population Register estimated with the Labour Force Survey (in Swedish), Metodrapport från BoRavdelningen 1995:3, Statistics Sweden Greijer, Å (1996) Overcoverage in the Population Register – an analysis of returned mail (in Swedish) Metodrapport från BoR-avdelningen 1996:7, Statistics Sweden REFERENCES 303 Greijer, Å (1997a) Overcoverage in the Population Register estimated with the Income Register (in Swedish) Metodrapport från BoR-avdelningen 1997:11, Statistics Sweden Greijer, Å (1997b) Overcoverage in the Population Register estimated with the Labour Force Survey (in Swedish) Metodrapport från BoR-avdelningen 1997:12, Statistics Sweden Groves, R and Lyberg, L (2010) Total survey error – past, present and future Public Opinion Quarterly, 74(5), 849–879 Groves, R., Fowler, F., Couper, M., Lepkowski, J., Singer, E and Tourangeau, R (2004) Survey Methodology Hoboken, NJ: John Wiley & Sons, Inc Herzog, T., Scheuren, F and Winkler, W (2007) Data Quality and Record Linkage Techniques New York: Springer Holt, D (2001) Comment on Platek and Särndal Journal of Official Statistics, 17(1), 55–61 Hoogland, J., van der Loo, M., Pannekoek, J and Scholtus, S (2011) Data editing – Detection and correction of errors, Statistical Methods (201110) The Hague: Statistics Netherlands Johansson, D (1997) The Number and the Size Distribution of Firms in Sweden and Other European Countries IUI Working Paper no 483, Research Institute of Industrial Economics, Stockholm Johansson, D (2001) The Dynamics of Firm and Industry Growth – The Swedish Computing and Communications Industry TRITA-IEO R 2001:05, Royal Institute of Technology, Stockholm Kardaun, J.W.P.F and Loeve, J.A (2005) Longitudinal analysis in statistical offices Statistics Netherlands Discussion Paper 05010 Kott, P and Vogel, F (1995) Multiple-frame business surveys In B Cox, D Binder, N Chinnappa, A Christianson, M Colledge and P Kott (eds), Business Survey Methods, pp.185–201 New York: John Wiley & Sons, Inc Laitila, T., Wallgren, A and Wallgren B (2012) Quality Assessment of Administrative Data – Data Source Quality Part two of third deliverable of work package of the BLUE Enterprise and Trade Statistics project Lenk, M (2009) Methods of Register-based Census in Austria Paper presented at the Seminar on Innovations in Official Statistics, United Nations, New York, February http://unstats.un.org/unsd/statcom/statcom_09/seminars/innovation /innovations_seminar.htm Lohr, S (2009) Introduction to Part 1: Sampling and survey design In D Pfefferman and C.R Rao (eds), Sample Surveys: Design, Methods and Applications, Handbook of Statistics Volume 29A, pp 3–8 Amsterdam: Elsevier Nanopoulos, P (2001) Comment on Platek and Särndal Journal of Official Statistics, 17(1), 77–86 Pannekoek, J and de Waal, T (2005) Automatic edit and imputation for business surveys: The Dutch contribution to the EUREDIT project Journal of Official Statistics, 21(2), 257–286 Platek, R and Särndal, C-E (2001) Can a statistician deliver? Journal of Official Statistics, 17(1), 1–20 304 REFERENCES Ploug, N (2013) New forms of data for official statistics Paper presented to the 59th ISI World Statistics Congress, Hong Kong, Session STS027 Särndal, C.-E and Lundström, S (2005) Estimation in Surveys with Nonresponse Chichester: John Wiley & Sons, Ltd Scheuren, F and Petska, T (1993) Turning administrative systems into information systems Journal of Official Statistics, 9(1), 109–119 Selander, R., Svensson, J., Wallgren, A and Wallgren, B (1998) Administrative Registers in an Efficient Statistical System – New Possibilities for Agricultural Statistics? How Should We Use IACS Data? Statistics Sweden, Eurostat Statistics Canada (2009) Statistics Canada Quality Guidelines, 5th edn Statistics Denmark (1995) Statistics on Persons in Denmark – A register-based statistical system Eurostat Statistics Finland (2004) Use of Registers and Administrative Data Sources for Statistical Purposes – Best practices of Statistics Finland Statistics Finland (2012) Finland’s Public Sector Data Registers – a national asset http://www.vm.fi/vm/en/04_publications_and_documents/01_publications/ 08_other_publications/20110705Finlan/name.jsp Statistics Netherlands (2004) The Dutch Virtual Census of 2001 – Analysis and Methodology Statistics Sweden (2007a) Register-based Activity Statistics (in Swedish) Background Facts – Labour and Education Statistics 2007:2 Statistics Sweden (2007b) Register-based economic statistics base on a standardised register population – A calendar year version of the business register with consistent microdata designed for the yearly National Accounts (in Swedish) Background Facts – Economic Statistics 2007:6 Statistics Sweden (2013) Revising Surveys – Linking Old and New Data Background facts on Labour and Education Statistics 2013:2 http://www.scb.se/ statistik/_publikationer/AM0401_2013A01_BR_AM76BR1302.pdf Steel, D and McLaren, C (2009) Design and analysis of surveys repeated over time In D Pfefferman and C.R Rao (eds), Sample Surveys: Inference and Analysis, Handbook of Statistics Volume 29B, pp 289–313 Amsterdam: Elsevier Struijs, P., Camstra, A., Renssen, R and Braaksma, B (2013) Redesign of statistics production within an architectural framework: the Dutch experience Journal of Official Statistics, 29(1), 49–71 UN/ECE (2007) Register-based statistics in the Nordic countries – Review of best practices with focus on population and social statistics United Nations Wallgren, A and Wallgren, B (1999) Administrative Registers in an Efficient Statistical System – How Can we Use Multiple Administrative Sources? Statistics Sweden and Eurostat Wallgren, A and Wallgren, B (2010) Using administrative registers for agricultural statistics In R Benedetti, M Bee, G Espa and F Piersimoni (eds), Agricultural Survey Methods, pp.27–44 Chichester: John Wiley & Sons, Ltd Wallgren, A and Wallgren, B (2011) To understand the possibilities of administrative data you must change your statistical paradigm! Paper at JSM 2011 REFERENCES 305 Wallgren, A and Wallgren, B (2012) Linking a system of time series – Recalculation of Labour Force Surveys 1987–1992 Research and Development – Methodology reports from Statistics Sweden2012:2 http://www.scb.se/statistik /_publikationer/ AM0401_2012A01_BR_X103BR1202.pdf Willenborg, L and de Waal, T (2001) Elements of Statistical Disclosure Control New York: Springer Winkler, W (1995) Matching and record linkage In B Cox, D Binder, N Chinnappa, A Christianson, M Colledge and P Kott (eds), Business Survey Methods New York: John Wiley & Sons, Inc Winkler, W (2006) Overview of Record Linkage and Current Research Directions Research Report Series 2006/2, US Bureau of the Census Winkler, W (2008) Record linkage In D Pfefferman and C.R Rao (eds), Sample Surveys: Design, Methods and Applications, Handbook of Statistics Volume 29A, pp 351–380 Amsterdam: Elsevier Zhang, L.-C (2011) A unit-error theory for register-based household statistics Journal of Official Statistics, 27(3), 415–432 Index Adjoined variable, 63, 68, 71, 151, 161 Aggregated variable, 63, 68, 71, 161f Aggregation error, 230f, 236, 241f, 247f Anonymised, 46, 127 Auxiliary variable, 52, 186, 202 Calendar year register, 58–59, 86, 137, 139, 230, 281, 293 Calibration, 202, 210, 225 Census, 10f, 22, 265 Classification, 5, 61, 78 Classification database, 197–198 Classification error, 82, 157, 285 Coding, 158, 189, 271 Coherence, 3, 74 Cohort, 166–167 Combination object, 145, 233 Communication variable, 64, 71, 84 Consistency, 3, 56 Consistency editing, 172–181 Coverage error, 19, 66f, 133, 280, 295 Cross-sectional quality, 269 Current stock register, 58, 86, 91, 139 Data warehouse, 53, 291–294 Definitions database, 197–198 Demographic event, 58–59, 64, 77, 84–86, 145 Derived object, see derived unit Derived unit, 15, 80, 142–145, 178 Derived variable, 62–64, 169 Deterministic matching or record linkage, 61, 105 Disclosure risk, 22, 46 Events calendar, 195–196 Events register, 59 Flow variable, 59, 86, 91, 137–138 Foreign key, 64–67 Frame population, 58, 86, 132 Fundamental estimation methods, 202–203 Historical register, 59 Identifying variable, 64 Imported variable, 63–64, 71, 195, 210 Imputation, 56, 215 Imputation error, 284, 286, 289 Integrated data collection, 79, 128 Integrated register, 51, 57, 70–71 Integration error, 230, 279, 295, 298 Link, 48, 61 Linking of time series, 254 Local primary variable, 63–64, 71 Local variable, 63 Locally derived variable, 63–64, 71 Longitudinal quality, 165, 269 Longitudinal register, 59, 71, 145, 165 Matching error, 105, 112, 176, 203, 279 Matching key, 105, 109–112 Measurement error, 29, 282, 288 Metadata, 13–14, 35, 124, 128, 193f Model error, 29, 147, 153, 284, 289 Multi-valued variable, 145, 150, 161–163, 229f Register-based Statistics: Statistical Methods for Administrative Data, Second Edition Anders Wallgren and Britt Wallgren © 2014 John Wiley & Sons, Ltd Published 2014 by John Wiley & Sons, Ltd 308 Natural random variation, 203, 289 Nonresponse adjustment, 136, 209 Object type, 13, 47 Overcoverage, 6, 16, 66, 100, 133, 139–140, 221–226 Primary key, 64 Primary variable, 63 Primary statistical register, 70–71, 125 Record linkage, 65, 103f Reference variable, 64–65, 84 Register Calendar year register, 58–59, 86, 137, 139, 230, 281, 293 Current stock register, 58, 86, 91, 139 Events register, 59 Historical register, 59 Integrated register, 51, 57, 70–71 Longitudinal register, 59, 71, 145, 165 Primary statistical register, 70–71, 125 Register at a specific point in time, 58 INDEX Survey design, 122–125, 130, 185–186, 264–266 Survey system design, 121, 265, 277 Technical variable, 64, 84 Time reference, 13, 51, 64, 82–83 Time series quality, 269 Total survey error, 191, 261 Undercoverage, 6, 16, 67, 133, 139–141, 226f Value set, 197 Variable Adjoined variable, 63, 68, 71, 151, 161 Aggregated variable, 63, 68, 71, 161f Auxiliary variable, 52, 186, 202 Classification, 5, 61, 78 Communication variable, 64, 71, 84 Derived variable, 62–64, 169 Flow variable, 59, 86, 91, 137–138 Identifying variable, 64 Imported variable, 63–64, 71, 195, 210 Local primary variable, 63–64, 71 Local variable, 63 Locally derived variable, 63–64, 71 Matching key, 105, 109–112 Register maintenance survey, 89, 99, Multi-valued variable, 145, 150, 161– 100, 275, 280 163, 229f Register population, 132, 146 Primary key, 64 Relational object, 49, 94, 159 Primary variable, 63 Relevance error, 22, 133, 284, 295 Reference variable, 64–65, 84 Response variable, 65 Response variable, 65 Single-valued variable, 150, 162 Single-valued variable, 150, 162 Spanning variable, 65, 81–82 Spanning variable, 65, 81–82 Standardised variable, 60–61 Standard, see classification Stock variable, 59–60, 137–138 Standardised population, 56–57, 77, Technical variable, 64, 84 104, 141 Time reference, 13, 51, 64, 82–83 Standardised variable, 60–61 Stock variable, 59–60, 137–138 Weight-generating variable, 137–138, Supplementary estimation methods, 202, 235 209, 221 WILEY SERIES IN SURVEY METHODOLOGY Established in Part by WALTER A SHEWHART AND SAMUEL S WILKS Editors: Mick P Couper, Graham Kalton, J N K Rao, Norbert Schwarz, Christopher Skinner Editor Emeritus: Robert M Groves The Wiley Series in Survey Methodology covers topics of current research and practical interests in survey methodology and sampling While the emphasis is on application, theoretical discussion is encouraged when it supports a broader understanding of the subject matter The authors are leading academics and researchers in survey methodology and sampling The readership includes professionals in, and students of, the fields of applied statistics, biostatistics, public policy, and government and corporate enterprises ALWIN · Margins of Error: A Study of Reliability in Survey Measurement BETHLEHEM · Applied Survey Methods: A Statistical Perspective BETHLEHEM, COBBEN, and SCHOUTEN · Handbook of Nonresponse in Household Surveys BIEMER · Latent Class Analysis of Survey Error *BIEMER, GROVES, LYBERG, MATHIOWETZ, and SUDMAN · Measurement Errors in Surveys BIEMER and LYBERG · Introduction to Survey Quality BIEMER · Latent Class Analysis of Survey Error BRADBURN, SUDMAN, and WANSINK ·Asking Questions: The Definitive Guide to Questionnaire Design—For Market Research, Political Polls, and Social Health Questionnaires, Revised Edition BRAVERMAN and SLATER · Advances in Survey Research: New Directions for Evaluation, No 70 CALLEGARO, BAKER, BETHLEHEM, GÖRITZ, KROSNICK, and LAVRAKAS (editors) · Online Panel Research: A Data Quality Perspective CHAMBERS and SKINNER (editors) · Analysis of Survey Data COCHRAN · Sampling Techniques, Third Edition CONRAD and SCHOBER · Envisioning the Survey Interview of the Future COUPER, BAKER, BETHLEHEM, CLARK, MARTIN, NICHOLLS, and O’REILLY (editors) · Computer Assisted Survey Information Collection COX, BINDER, CHINNAPPA, CHRISTIANSON, COLLEDGE, and KOTT (editors) · Business Survey Methods *DEMING · Sample Design in Business Research DILLMAN · Mail and Internet Surveys: The Tailored Design Method FULLER · Sampling Statistics GROVES and COUPER · Nonresponse in Household Interview Surveys GROVES · Survey Errors and Survey Costs GROVES, DILLMAN, ELTINGE, and LITTLE · Survey Nonresponse GROVES, BIEMER, LYBERG, MASSEY, NICHOLLS, and WAKSBERG · Telephone Survey Methodology GROVES, FOWLER, COUPER, LEPKOWSKI, SINGER, and TOURANGEAU · Survey Methodology, Second Edition *HANSEN, HURWITZ, and MADOW · Sample Survey Methods and Theory, Volume 1: Methods and Applications *HANSEN, HURWITZ, and MADOW · Sample Survey Methods and Theory, *Now available in a lower priced paperback edition in the Wiley Classics Library Volume II: Theory HARKNESS, BRAUN, EDWARDS, JOHNSON, LYBERG, MOHLER, PENNELL, and SMITH (editors) · Survey Methods in Multinational, Multiregional, and Multicultural Contexts HARKNESS, VAN DE VIJVER, and MOHLER (editors) · Cross-Cultural Survey Methods HUNDEPOOL, DOMINGO-FERRER, FRANCONI, GIESSING, NORDHOLT, SPICER, and DE WOLF · Statistical Disclosure Control KALTON and HEERINGA · Leslie Kish Selected Papers KISH · Statistical Design for Research *KISH · Survey Sampling KORN and GRAUBARD · Analysis of Health Surveys KREUTER (editor) · Improving Surveys with Paradata: Analytic Uses of Process Information LEPKOWSKI, TUCKER, BRICK, DE LEEUW, JAPEC, LAVRAKAS, LINK, and SANGSTER (editors) · Advances in Telephone Survey Methodology LESSLER and KALSBEEK · Nonsampling Error in Surveys LEVY and LEMESHOW · Sampling of Populations: Methods and Applications, Fourth Edition LUMLEY · Complex Surveys: A Guide to Analysis Using R LYBERG, BIEMER, COLLINS, de LEEUW, DIPPO, SCHWARZ, TREWIN (editors) · Survey Measurement and Process Quality LYNN · Methodology of Longitudinal Surveys MADANS, MILLER, and MAITLAND (editors) · Question Evaluation Methods: Contributing to the Science of Data Quality MAYNARD, HOUTKOOP-STEENSTRA, SCHAEFFER, and VAN DER ZOUWEN · Standardization and Tacit Knowledge: Interaction and Practice in the Survey Interview PORTER (editor) · Overcoming Survey Research Problems: New Directions for Institutional Research, No 121 PRESSER, ROTHGEB, COUPER, LESSLER, MARTIN, MARTIN, and SINGER (editors) · Methods for Testing and Evaluating Survey Questionnaires RAO · Small Area Estimation REA and PARKER · Designing and Conducting Survey Research: A Comprehensive Guide, Third Edition SARIS and GALLHOFER · Design, Evaluation, and Analysis of Questionnaires for Survey Research, Second Edition SÄRNDAL and LUNDSTRÖM · Estimation in Surveys with Nonresponse SCHWARZ and SUDMAN (editors) · Answering Questions: Methodology for Determining Cognitive and Communicative Processes in Survey Research SIRKEN, HERRMANN, SCHECHTER, SCHWARZ, TANUR, and TOURANGEAU (editors) · Cognition and Survey Research SNIJKERS, HARALDSEN, JONES, and WILLIMACK · Designing and Conducting Business Surveys STOOP, BILLIET, KOCH and FITZGERALD · Improving Survey Response: Lessons Learned from the European Social Survey SUDMAN, BRADBURN, and SCHWARZ · Thinking about Answers: The Application of Cognitive Processes to Survey Methodology UMBACH (editor) · Survey Research Emerging Issues: New Directions for Institutional Research No 127 VALLIANT, DORFMAN, and ROYALL · Finite Population Sampling and Inference: A Prediction Approach WALLGREN and WALLGREN · Register-based Statistics: Administrative Data for Statistical Purposes, Second Edition ... of administrative data How are data recorded? Administrative and statistical information systems Measurement errors in statistical and administrative data Why use administrative data for statistics? ... theory We formulate four basic principles for using administrative registers (Chart 1.1) Chart 1.1 Four principles for using administrative registers for statistics Transformation principle Administrative. .. during the 1960s when paper -based administrative registers were transformed into computer -based flat files The preconditions for using administrative registers for statistical purposes were good