Contents

Preface xi

I. The Theory and Methodology of Discourse Networks

1. Introduction 3

2. Actor-centered approaches to discourse 7

2.1. The Advocacy Coalition Framework 9

2.2. Veto player analysis 16

2.3. Punctuated Equilibrium Theory 17

2.4. Policy paradigms and social learning 21

2.5. Collective symbolic coping 23

2.6. Civic arenas 25

2.7. Multiple streams 26

2.8. Argumentative discourse analysis and discourse coalitions . 28

2.9. Epistemic communities 30

2.10. Comparison of actor-centered approaches 32

3. Content-oriented approaches to discourse 37

3.1. Critical discourse analysis 38

3.2. Category-based content analysis 40

3.3. Frame mapping and the co-occurrence approach 41

3.4. Clause- or grammar-based methods and knowledge graphs . 43

3.5. Semantic networks 45

3.6. Comparison of content-oriented approaches 48

4. Hybrid approaches: linking actors and contents 53

4.1. Decomposition analysis 53

4.2. Political claims analysis 54

4.3. Conclusions for a new discourse methodology 57

4.4. Policy network analysis 60

5. The methodology of discourse network analysis 61

5.1. Coding procedure 61

5.2. Affiliation networks 62

5.3. Actor congruence networks 64

5.4. Concept congruence networks 69

5.5. Conflict networks 71

5.6. Time window networks 72

5.7. Attenuation networks 78

5.8. Software implementation 84

5.8.1. Encoding statements 84

5.8.2. Network export facilities 86

II. A Showcase: German Pension Politics, 1993-2001

6. German pension politics in the 1990s and the 2001 Riester reform 91

6.1. Dimensions of pension systems and the status quo in the 1990s 93

6.1.1. Pay-as-you-go versus capital cover systems 93

6.1.2. Public versus private pension systems 97

6.1.3. Voluntary versus mandatory contributions 98

6.1.4. Intra-generative redistribution versus equivalence . 98

6.1.5. Risk balance 101

6.1.6. The pension formula 102

6.2. Demographic change and the pension gap 104

6.2.1. Demographic Transition as a complex long-term risk 105

6.2.2. Mortality 106

6.2.3. Fertility 107

6.2.4. Inconsistent solution concepts in the demographic debate 110

6.2.5. Migration 111

6.3. The 2001 Riester reform 112

6.3.1. The 1999 Pension Reform Act 112

6.3.2. Changes in the 2001 reform 113

6.4. Positive analyses of the policy outcome 117

7. Description of the dataset and univariate analysis . . . . . 125

7.1. Coding procedure . . . . . . . . . . . . . . . . . . . . . . 125

7.2. Media bias and validity . . . . . . . . . . . . . . . . . . . . 128

7.2.1. Record coding . . . . . . . . . . . . . . . . . . . . 129

7.2.2. Data coding . . . . . . . . . . . . . . . . . . . . . 133

7.2.3. Indexing . . . . . . . . . . . . . . . . . . . . . . . 134

7.3. Actors in the pension discourse . . . . . . . . . . . . . . . 134

7.3.1. Government . . . . . . . . . . . . . . . . . . . . . 135

7.3.2. Social actors . . . . . . . . . . . . . . . . . . . . . 136

7.3.3. Liberal actors . . . . . . . . . . . . . . . . . . . . 136

7.3.4. Financial sector . . . . . . . . . . . . . . . . . . . 137

7.3.5. Young actors . . . . . . . . . . . . . . . . . . . . . 137

7.3.6. Scientists . . . . . . . . . . . . . . . . . . . . . . . 137

7.3.7. Other actors . . . . . . . . . . . . . . . . . . . . . 137

7.4. Univariate analysis: concepts in the pension discourse . . 138

7.4.1. System change . . . . . . . . . . . . . . . . . . . . 140

7.4.2. Retrenchment within the PAYG paradigm . . . . . 151

7.4.3. Strengthening the insurance principle . . . . . . . . 160

7.4.4. Increasing the number of contributors . . . . . . . 167

8. Empirical analysis of the German pension discourse in the 1990s 183

8.1. Discourse activity over time 184

8.2. Cross-sectional actor congruence, 1993-2001 188

8.3. Longitudinal change 199

8.4. Analysis of political parties 209

8.5. Identification of ideologies 214

8.6. Discussion and conclusion 218

III. Models of Discursive Behavior

9. The contagious dimension of political discourse 225

9.1. Attention and interaction in media discourse 225

9.1.1. Discursive contagion in existing theories 226

9.1.2. Two types of contagion 227

9.1.3. Hypotheses 228

9.2. Methods and data 230

9.2.1. Operationalization 231

9.2.2. Support versus resistance 231

VIII POLICY DEBATES AS DYNAMIC NETWORKS

9.2.3. Exponential random graph models with dyadic de- pendence 232

9.2.4. Estimation 233

9.2.5. Dichotomization of edge weights 234

9.2.6. Control variables 234

9.2.7. Data 236

9.2.8. Potential caveats 237

9.3. Results and goodness of fit 239

9.4. Discussion 244

10. An agent-based model of political discourse 249

10.1. A formal model of political discourse 251

10.1.1. Defi and basic setup 251

10.1.2. Exogenous ideology 252

10.1.3. Endogenous ideology 252

10.1.4. Concept popularity (bandwagoning) 252

10.1.5. Actor similarity (coalition formation) 253

10.1.6. Concept similarity 253

10.1.7. Actor's history (self-consistency) 254

10.1.8. Rare concepts (agenda-setting) 255

10.1.9. Government coherence 255

10.1.10. Normalization 256

10.1.11. Utility functions 257

10.2. Analysis 258

10.2.1. Measurement 258

10.2.2. Betweenness centralization 258

10.2.3. Ideological polarization 259

10.2.4. Number of components 261

10.2.5. Proportion of concepts still alive 261

10.2.6. Number of recent concept changes 262

10.3. Results 262

10.4. Discussion 272

11. Conclusion 275

11.1. Main achievements 275

11.2. Potential weaknesses 276

11.3. Outlook 278

IV. Appendix

Laws and legislative decrees . . . . . . . . . . . . . . . . . . . . . . . 283

List of actors in the dataset . . . . . . . . . . . . . . . . . . . . . . . 285

Software manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

1. Installation . . . . . . . . . . . . . . . . . . . . . . . . . . 291

2. Tutorial for beginners . . . . . . . . . . . . . . . . . . . . 292

3. In-depth description . . . . . . . . . . . . . . . . . . . . . 299

3.1. File format . . . . . . . . . . . . . . . . . . . . . . 299

3.2. Dealing with articles . . . . . . . . . . . . . . . . . 300

3.3. Recoding statements . . . . . . . . . . . . . . . . . 302

3.4. Regular expressions highlighter . . . . . . . . . . . 302

3.5. Within-actor contradictions . . . . . . . . . . . . . 303

3.6. The bottom bar . . . . . . . . . . . . . . . . . . . 303

3.7. Exporting time series statistics . . . . . . . . . . . 305

3.8. Exporting network data . . . . . . . . . . . . . .

4. rDNA. A Package to Control Discourse Network Analyzer from R 307

4.1. Motivation . . . . . . . . . . . . . . . . . . . . . . 315

4.2. Functionality . . . . . . . . . . . . . . . . . . . . . 315

4.3. Examples . . . . . . . . . . . . . . . . . . . . . . 316

5. FAQ: Frequently Asked Questions . . . . . . . . . . . . . . 317

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349