Title page
Contents
Abstract 2
1. Introduction 4
2. Clustering methodologie 7
2.1. The model 7
2.2. Static clustering membership 8
2.3. Cluster switching 9
2.4. Identification and the number of clusters 11
3. The data 11
4. Results 16
4.1. Linear model with two-way fixed effects 16
4.2. Clusters and heterogeneity 17
4.3. Allowing for cluster membership switches 23
4.4. Robustness 28
5. Conclusions 38
References 43
Appendix A. Additional results 46
Table 1. Sample selection and observation numbers 14
Table 2. Country transitions for K = 2 clusters and H = 2 time partitions 30
Figure 1. CPIA and Financial Development (FD) descriptives 15
Figure 2. Linear panel-data regression results 17
Figure 3. Cluster validation indices for (CPIA, FD) 19
Figure 4. Group membership in the bivariate model without regressors and without cluster switching 20
Figure 5. Population-weighted time-varying cluster mean FD 21
Figure 6. Panel-data regression results with cluster-specific coefficients 22
Figure 7. Cluster switches 24
Figure 8. Panel-data regression coefficients with cluster membership switches 26
Figure 9. Panel-data group-specific time-effects with cluster membership switches 27
Figure 10. Cluster membership results under switching 29
Figure 11. Cluster validation indices for 7-variables 31
Figure 12. Group membership for CPIA plus seven FD constituents 32
Figure 13. Cluster means over time in the 7-variate model 33
Figure 14. Cluster validation indices for 4-variables 34
Figure 15. Group membership for CPIA plus three FI constituents 35
Figure 16. Population-weighted time-varying cluster means the three FI constituents 36
Figure 17. Linear panel-data regression results with the three FI components 37
Figure 18. Panel-data regression results with cluster-specific coefficients for CPIA plus three FI constituents 38
Figure 19. Linear panel-data regression results with FSI 39
Figure 20. Linear panel-data regression results with FSI regressed on the components of FI 39
Figure 21. Group membership in the bivariate model without regressors and without cluster switching 40
Figure 22. Group membership in the bivariate model without regressors and without cluster switching 40
Figure 23. Panel-data regression results with cluster-specific coefficients using FSI as the dependent variable 41
Figure 24. Panel-data regression results with cluster-specific coefficients using FSI as the dependent variable and the components of FI as covariates 42
Table A.1. Transitions from partition 1 (rows) to 2 (columns), 2 clusters, 3 partition total, 4 controls 51
Table A.2. Transitions from partition 2 (rows) to 3 (columns), 2 clusters, 3 partition total, 4 controls49 51
Table A.3. Transitions from partition 1 (rows) to 2 (columns), 3 clusters, 2 partition total, 4 controls 52
Table A.4. Transitions from partition 1 (rows) to 2 (columns), 3 clusters, 3 partition total, 4 controls 53
Table A.5. Transitions from partition 2 (rows) to 3 (columns), 3 clusters, 3 partition total, 4 controls 54
Figure A.1. Time-varying cluster assignments, 3 clusters and 2 partitions 46
Figure A.2. Time-varying cluster assignments, 2 clusters and 3 partitions 47
Figure A.3. Time-varying cluster assignments, 3 clusters and 3 partitions 48
Figure A.4. Cluster averages, 2 clusters and 2 partitions 49
Figure A.5. Cluster averages, 2 clusters and 3 partitions 50