Register-based Statistics
Administrative Data for Statistical Purposes
Anders and Britt Wallgren
Statistics Sweden, Sweden
Contents
Preface ix
Chapter 1 Register-based Surveys – An Introduction 1
1.1 Do we need a theory on register-based surveys? 1
1.2 What is a statistical survey? 3
1.3 What is a register? 4
1.4 What is a register-based survey? 4
1.5 Administrative and statistical information systems 10
1.6 Why use administrative data for statistics? 12
1.7 An overview of this book 15
Chapter 2 How to Structure a Register System 19
2.1 A register model based on object types and relations 19
2.2 The system of base registers 23
2.3 The register system as a whole 29
2.4 Building and using the system 30
2.5 Standardised variables in the register system 33
2.6 Statistical register systems outside Statistics Sweden 34
Chapter 3 A Terminology for Register-based Surveys 41
3.1 Terminology – different language 41
3.2 Register terms 42
3.3 Terms for different kinds of variables 49
Chapter 4 Sample Surveys and Registers 59
4.1 How can sample surveys benefit by the register system? 59
4.2 Combining register-based surveys and sample surveys 61
4.3 Comparing sample surveys and register-based surveys 63
Chapter 5 How to create a Register – The Population 67
5.1 How should register-based surveys be structured? 67
5.2 Determining the research objectives 70
5.3 Making an inventory of different sources 72
5.4 Defining a register’s object set 72
5.5 Defining and deriving objects 84
5.6 How to produce regional register-based statistics 88
vi CONTENTS
Chapter 6 How to create a Register – The Variables 91
6.1 Deciding the register’s variable content 91
6.2 Forming derived variables using models 93
6.3 Editing and correcting register variables 100
6.4 Creating longitudinal registers 112
Chapter 7 Estimation Methods 115
7.1 Estimation in sample surveys and register-based surveys 116
7.2 Register-based surveys – Fundamental estimation methods 117
7.3 Using weights in register-based surveys 119
7.4 Estimation using weights – calendar year registers 121
7.5 Calibration of weights in register-based surveys 123
Chapter 8 Calibration and Imputation 127
8.1 The nonresponse problem 127
8.2 Estimation methods to correct for overcoverage 138
8.3 Methods to correct for level shifts in time series 140
Chapter 9 Estimation with Combination Objects 147
9.1 Aggregation errors 147
9.2 Estimation methods for multi-valued variables 149
9.3 Linking of time series using combination objects 168
Chapter 10 Quality of Register-based Statistics 173
10.1 Specific quality issues for register-based statistics? 175
10.2 Errors in sample surveys and register-based surveys 176
10.3 The users’ and the producers’ view of quality 181
10.4 Detailed knowledge of a register’s characteristics 182
10.5 Overall appraisal of quality 190
10.6 Main quality issues in different kinds of surveys 192
Chapter 11 Metadata and IT-systems 193
11.1 Primary registers – the need for metadata 193
11.2 Changes over time – the need for metadata 195
11.3 Integrated registers – the need for metadata 196
11.4 Classification and definitions database 197
11.5 The need for metadata for registers 198
11.6 IT systems for register-based statistics 200
Chapter 12 Protection of Privacy and Confidentiality 209
12.1 Internal security 210
12.2 Disclosure risks – tables 212
12.3 Disclosure risks – micro data 216
CONTENTS vii
Chapter 13 Coordination and Coherence 217
13.1 Content-related coordination 217
13.2 Coherence 219
13.3 Consistent and coherent enterprise statistics 220
Chapter 14 Conclusions 227
References 231
Glossary 235
Index 245
Administrative Data for Statistical Purposes
Anders and Britt Wallgren
Statistics Sweden, Sweden
Contents
Preface ix
Chapter 1 Register-based Surveys – An Introduction 1
1.1 Do we need a theory on register-based surveys? 1
1.2 What is a statistical survey? 3
1.3 What is a register? 4
1.4 What is a register-based survey? 4
1.5 Administrative and statistical information systems 10
1.6 Why use administrative data for statistics? 12
1.7 An overview of this book 15
Chapter 2 How to Structure a Register System 19
2.1 A register model based on object types and relations 19
2.2 The system of base registers 23
2.3 The register system as a whole 29
2.4 Building and using the system 30
2.5 Standardised variables in the register system 33
2.6 Statistical register systems outside Statistics Sweden 34
Chapter 3 A Terminology for Register-based Surveys 41
3.1 Terminology – different language 41
3.2 Register terms 42
3.3 Terms for different kinds of variables 49
Chapter 4 Sample Surveys and Registers 59
4.1 How can sample surveys benefit by the register system? 59
4.2 Combining register-based surveys and sample surveys 61
4.3 Comparing sample surveys and register-based surveys 63
Chapter 5 How to create a Register – The Population 67
5.1 How should register-based surveys be structured? 67
5.2 Determining the research objectives 70
5.3 Making an inventory of different sources 72
5.4 Defining a register’s object set 72
5.5 Defining and deriving objects 84
5.6 How to produce regional register-based statistics 88
vi CONTENTS
Chapter 6 How to create a Register – The Variables 91
6.1 Deciding the register’s variable content 91
6.2 Forming derived variables using models 93
6.3 Editing and correcting register variables 100
6.4 Creating longitudinal registers 112
Chapter 7 Estimation Methods 115
7.1 Estimation in sample surveys and register-based surveys 116
7.2 Register-based surveys – Fundamental estimation methods 117
7.3 Using weights in register-based surveys 119
7.4 Estimation using weights – calendar year registers 121
7.5 Calibration of weights in register-based surveys 123
Chapter 8 Calibration and Imputation 127
8.1 The nonresponse problem 127
8.2 Estimation methods to correct for overcoverage 138
8.3 Methods to correct for level shifts in time series 140
Chapter 9 Estimation with Combination Objects 147
9.1 Aggregation errors 147
9.2 Estimation methods for multi-valued variables 149
9.3 Linking of time series using combination objects 168
Chapter 10 Quality of Register-based Statistics 173
10.1 Specific quality issues for register-based statistics? 175
10.2 Errors in sample surveys and register-based surveys 176
10.3 The users’ and the producers’ view of quality 181
10.4 Detailed knowledge of a register’s characteristics 182
10.5 Overall appraisal of quality 190
10.6 Main quality issues in different kinds of surveys 192
Chapter 11 Metadata and IT-systems 193
11.1 Primary registers – the need for metadata 193
11.2 Changes over time – the need for metadata 195
11.3 Integrated registers – the need for metadata 196
11.4 Classification and definitions database 197
11.5 The need for metadata for registers 198
11.6 IT systems for register-based statistics 200
Chapter 12 Protection of Privacy and Confidentiality 209
12.1 Internal security 210
12.2 Disclosure risks – tables 212
12.3 Disclosure risks – micro data 216
CONTENTS vii
Chapter 13 Coordination and Coherence 217
13.1 Content-related coordination 217
13.2 Coherence 219
13.3 Consistent and coherent enterprise statistics 220
Chapter 14 Conclusions 227
References 231
Glossary 235
Index 245
