The Econometrics of Complex Survey Data: Theory and Applications

COVER
THE ECONOMETRICS OF COMPLEX SURVEY DATA: THEORY AND APPLICATIONS
COPYRIGHT PAGE
CONTENTS
LIST OF CONTRIBUTORS
INTRODUCTION
SAMPLING DESIGN
VARIANCE ESTIMATION
ESTIMATION AND INFERENCE
BUSINESS, HOUSEHOLD AND CRIME SURVEYS
PART I SAMPLING DESIGN
CAN INTERNET MATCH HIGH-QUALITY TRADITIONAL SURVEYS? COMPARING THE HEALTH AND RETIREMENT STUDY AND ITS ONLINE VERSION
ABSTRACT
1. INTRODUCTION
2. METHODS AND OUTLINE
3. HRS AND UAS DESCRIPTIONS
3.1. The Health and Retirement Study
3.2. The Understanding America Study
3.3. UAS Sampling and Weighting Procedures
4. COMPARING SOCIOECONOMIC VARIABLES IN THE HRS, UAS AND CPS
4.1. Representativeness of the HRS and UAS Samples
5. COMPARING SURVEY OUTCOMES IN THE HRS AND UAS
5.1. Mode Effects
6. CONCLUSIONS
NOTES
REFERENCES
APPENDIX
EFFECTIVENESS OF STRATIFIED RANDOM SAMPLING FOR PAYMENT CARD ACCEPTANCE AND USAGE
ABSTRACT
1. INTRODUCTION
2. DATA DESCRIPTION
2.1. Hungarian Payments Landscape
2.2. Data Description
2.3. Key Variables
3. METHODOLOGY
3.1. Estimating Card Acceptance
3.2. Models of Card Acceptance and Usage
4. RESULTS
4.1. Estimates of Card Acceptance
4.2. Regression Models of Card Payments
4.2.1. Acceptance
4.2.2. Usage
5. CONCLUSIONS
ACKNOWLEDGEMENTS
REFERENCES
APPENDIX
A.1. Review of Literature
A.2. Logistic Regression Models of Card Acceptance and Usage
A.2.1. Card Acceptance: Variables
A.2.2. Card Acceptance: Results
A.2.3. Card Usage: Variables
A.2.4. Card Usage: Results
PART II VARIANCE ESTIMATION
WILD BOOTSTRAP RANDOMIZATION INFERENCE FOR FEW TREATED CLUSTERS
ABSTRACT
1. INTRODUCTION
2. CLUSTER-ROBUST INFERENCE
2.1. The Wild Cluster Bootstrap
3. RANDOMIZATION INFERENCE
3.1. The Problem of Interval p Values
4. WILD BOOTSTRAP RANDOMIZATION INFERENCE
5. ALTERNATIVE PROCEDURES
6. SIMULATION EXPERIMENTS
7. EMPIRICAL EXAMPLE
8. CONCLUSION
ACKNOWLEDGEMENTS
NOTES
REFERENCES
VARIANCE ESTIMATION FOR SURVEY-WEIGHTED DATA USING BOOTSTRAP RESAMPLING METHODS: 2013 METHODS-OF-PAYMENT SURVEY QUESTIONNAIRE
ABSTRACT
1. INTRODUCTION
2. RAKING RATIO ESTIMATOR
2.1. Classical Raking Estimator
2.2. Maximum Likelihood Raking Estimator
2.3. Generalized Regression Estimator
3. VARIANCE ESTIMATION
3.1. Variance Estimation via Linearization
3.2. Variance Estimation via Bootstrap
4. EMPIRICAL APPLICATION OF THE 2013 MOP
4.1. Implementation Under Stata
5. SUMMARY
ACKNOWLEDGMENTS
NOTES
REFERENCES
PART III ESTIMATION AND INFERENCE
MODEL-SELECTION TESTS FOR COMPLEX SURVEY SAMPLES
ABSTRACT
1. INTRODUCTION
2. NONNESTED COMPETING MODELS AND THE NULL HYPOTHESIS
3. TESTING UNDER STRATIFIED SAMPLING
3.1. Standard Stratified Sampling
3.2. Variable Probability Sampling
4. TESTS STATISTICS UNDER MULTISTAGE SAMPLING
5. MODEL-SELECTION TESTS WITH PANEL DATA
6. EXOGENOUS STRATIFICATION
7. EXAMPLES
8. A SMALL SIMULATION STUDY
9. CONCLUSION
REFERENCES
INFERENCE IN CONDITIONAL MOMENT RESTRICTION MODELS WHEN THERE IS SELECTION DUE TO STRATIFICATION
ABSTRACT
1. INTRODUCTION
2. THE MODEL
2.1. Conditional Moment Equalities
2.2. Variable Probability Sampling
2.3. Identification
2.4. Endogenous and Exogenous Stratification
3. INFERENCE
3.1. Related Literature and Our Contribution
3.2. Efficiency Bounds
3.3. Efficient Estimation
3.4. Testing
4. SIMULATION STUDY
4.1. Design
4.2. Discussion
5. CONCLUSION
ACKNOWLEDGMENTS
NOTES
REFERENCES
APPENDIX A: COMPARING THE ASYMPTOTIC VARIANCE OF LS AND GMM ESTIMATORS UNDER EXOGENOUS STRATIFICATION
APPENDIX B: COMPUTATION
NONPARAMETRIC KERNEL REGRESSION USING COMPLEX SURVEY DATA
ABSTRACT
1. INTRODUCTION
2. MODEL-ASSISTED NONPARAMETRIC REGRESSION ESTIMATOR
2.1. Local Constant Estimator
2.2. Model-assisted Local Constant Estimator
2.3. Asymptotic Properties
3. BANDWIDTH SELECTION
4. MONTE CARLO SIMULATIONS
4.1. Sample Mean Squared Error
4.2. Bandwidths
5. APPLICATION
6. CONCLUSION
ACKNOWLEDGMENTS
REFERENCES
APPENDIX A. PROOFS
A.1. Proof of Theorem 2.1
A.2. Proof of Theorem 2.2
A.3. Least Squares Cross-Validation
NEAREST NEIGHBOR IMPUTATION FOR GENERAL PARAMETER ESTIMATION IN SURVEY SAMPLING
ABSTRACT
1. INTRODUCTION
2. BASIC SETUP
3. MAIN RESULTS
4. REPLICATION VARIANCE ESTIMATION
5. SIMULATION STUDY
6. CONCLUDING REMARKS
ACKNOWLEDGMENTS
REFERENCES
APPENDICES
APPENDIX A: PROOF FOR THEOREM 1
APPENDIX B: PROOF FOR THEOREM 2
APPENDIX C: ASSUMPTIONS FOR KERNEL FUNCTIONS
APPENDIX D: SIEVES ESTIMATION
APPENDIX E: PROOF FOR THEOREM 3
PART IV
APPLICATIONS IN BUSINESS,HOUSEHOLD, AND CRIME SURVEYS
IMPROVING RESPONSE QUALITY WITH PLANNED MISSING DATA: AN APPLICATION TO A SURVEY OF BANKS
ABSTRACT
1. INTRODUCTION AND LITERATURE
2. THE STANDARD APPROACH
3. THE CHALLENGE
4. PLANNED MISSING DATA DESIGN
5. RESULTS
6. CONCLUSION AND NEXT STEPS
ACKNOWLEDGMENT
NOTES
REFERENCES
DOES SELECTIVE CRIME REPORTING INFLUENCE OUR ABILITY TO DETECT RACIAL DISCRIMINATION IN THE NYPD’S STOP-AND-FRISK PROGRAM?
ABSTRACT
1. INTRODUCTION
2. STOP-AND-FRISK: DATA AND INSTITUTIONAL KNOWLEDGE
3. THEORY
3.1. Pedestrians
3.2. Police Officers
3.3. Existence of an Equilibrium
3.4. Characterization of the Equilibrium
4. Empirical Strategy
5. Results
6. CONCLUSION
ACKNOWLEDGMENT
NOTES
REFERENCES
APPENDIX
SURVEY EVIDENCE ON BLACK MARKET LIQUOR IN COLOMBIA
ABSTRACT
1. INTRODUCTION
2. DATA
3. RESULTS
4. MULTIPLE IMPUTATION
5. DISCUSSION
6. CONCLUSION
ACKNOWLEDGMENTS
REFERENCES
APPENDIX A: TABLES
APPENDIX B: MULTIPLE IMPUTATION BY CHAINED EQUATIONS
APPENDIX C: RESIDUAL ANALYSIS
INDEX