본문 바로가기 주메뉴 바로가기
국회도서관 홈으로 정보검색 소장정보 검색

목차보기

Title page

Contents

Abstract 3

1. Introduction 4

2. Context 9

3. The "Proud to Belong" training program 10

3.1. Design 10

3.2. Randomization and implementation 12

4. Data and empirical strategy 15

4.1. Survey data and incentivized cheating games 15

4.2. Estimation strategy 19

5. Results: Survey and Incentivized Cheating Game 22

5.1. Balance tests 22

5.2. Treatment effects 24

5.3. Secondary analysis: Mechanisms 29

6. Results: Administrative Data 31

7. Conclusions 35

References 37

APPENDIX A: FIGURES AND TABLES 51

APPENDIX B: THE TRAINING PROGRAM 74

Table 1. Descriptive Statistics 47

Table 2. Treatment effects on survey-generated outcomes 48

Table 3. Treatment effects on game-generated outcomes 49

Table 4. Comparing Baseline Officer Characteristics in Districts with and without Clean Administrative Data 50

Figure 1. Survey-based indexes 44

Figure 2. Officers' behavior in the incentivized cheating game 45

Figure 3. Event Study Estimates of Treatment Effects on Field Behavior 46

Table A1. Aggregate survey-based indexes 54

Table A2. Intent-to-Treat effects on Survey-generated outcomes 55

Table A3. Intent-to-Treat effects on game-generated outcomes 56

Table A4. LATE (2SLS) on survey-generated outcomes 57

Table A5. LATE (2SLS) on game-generated outcomes 58

Table A6. Robustness: Treatment effects on survey-generated outcomes, excluding trained officers in C 59

Table A7. Robustness: Treatment effects on game-generated outcomes, excluding trained officers in C 60

Table A8. Balance Tests - Control vs Central district officers 61

Table A9. Robustness - Treatment effects on survey-generated outcomes, including Central district officers 62

Table A10. Robustness - Treatment effects on game-generated outcomes, including Central district officers 63

Table A11. Treatment effects on survey-generated outcomes, excluding the top 10% improvement in survey scores 64

Table A12. Treatment effects on game-generated outcomes, excluding the top 10% improvement in survey scores 65

Table A13. Heterogeneous effects by initial intrinsic motivations: Survey-generated outcomes 66

Table A14. Heterogeneous effects by initial intrinsic motivations: Game-generated outcomes 67

Table A15. Heterogeneous effects by officer rank: Survey-generated outcomes 68

Table A16. Heterogeneous effects by officer rank: Game-generated outcomes 69

Table A17. Heterogeneous effects by number of trained officers: Survey-generated outcomes (treatment districts only) 70

Table A18. Heterogeneous effects by number of trained officers: Game-generated outcomes (treatment districts only) 71

Table A19. Comparing Baseline Officer Characteristics in C and T Admin Data Districts 72

Table A20. Treatment effects on District-Level Administrative Data (per officer per month) 73

Figure A1. Agent of Change pin delivered to trained officers at the award ceremony 51

Figure A2. Behavior in the baseline Mind Game (Oct 2018) 52

Figure A3. District-level Administrative Data on Field Behavior 53