본문 바로가기 주메뉴 바로가기
국회도서관 홈으로 정보검색 소장정보 검색

목차보기

Title page 1

Contents 1

Abstract 3

1. Introduction 4

2. Background and Data 8

2.1. Background 8

2.2. Main Data Sources 10

2.3. Descriptive Statistics 13

3. Research Design and Main Results 13

3.1. Estimating Equation 13

3.2. Instrumental Variable Strategy 14

3.3. Protest size and hate crimes 18

3.4. Robustness Checks 19

4. Mechanisms 23

4.1. Dynamic Effects 23

4.2. Spatial Diffusion 26

4.3. Media coverage 29

4.4. Counter-Mobilization 33

4.5. Coordination 33

5. Conclusion 36

References 38

Appendix A: Additional Results 42

Appendix B: Robustness Checks 50

Appendix C: Data Appendix 61

Tables 20

Table 1. Protest Participation increases probability of hate crimes 20

Table 2. Diffusion: geographic and (right-wing) social media proximity to other large protest locations 30

Table 3. Newspaper and Social Media Coverage pro- & anti-PEGIDA 32

Table 4. Large Protests and Counter-Mobilization Offline and Online 32

Table 5. Effect of Larger Protests on Hate-Crime Characteristics 35

Figures 12

Figure 1. Hate crimes and PEGIDA protest across municipalities 12

Figure 2. Hate crimes and PEGIDA protest over time 12

Figure 3. Residualized protest participation and weather conditions 16

Figure 4. Time variation of PEGIDA protests on pleasant days 17

Figure 5. Municipality characteristics and protest on pleasant days 18

Figure 6. Event Study: Hate Crimes Following Protests on Pleasant Days 25

Appendix Tables 44

Table A.1. Summary Statistics for Ever-Treated and Never-Treated Samples 44

Table A.2. First Stage: pleasant weather and protest participation 45

Table A.3. Evidence against reporting bias 46

Table A.4. Empirical results consistent with signaling, agitation, and coordination 47

Table A.5. Football match types and hate crimes 48

Table A.6. Football matches and hate crimes 49

Table B.1. Robustness: Protest participation and hate crime - spatial correlation 53

Table B.2. Protest Participation increases probability of hate crimes - more controls 54

Table B.3. Robustness: varying the fixed effects structure 55

Table B.4. Chaisemartin and D'Haultfoeuille (2020) TWFE negative weights 56

Table B.5. Protest Participation increases probability of hate crimes (without Monday) 57

Table B.6. Robustness: treatment definition and non-linear estimation 58

Table C.1. Examples of Hate Crimes 71

Table C.2. Description of Data Sources 72

Table C.3. Football derbies with data sources 73

Appendix Figures 42

Figure A.1. PEGIDA and other right-wing protests over weekdays (2015-2020) 42

Figure A.2. Protest participation and weather cut-offs 42

Figure A.3. Baseline result with alternative weather instruments 43

Figure A.4. Baseline result with LASSO-selected instrumental variable 43

Figure B.1. Robustness to excluding potential outliers 59

Figure B.2. Event study: first PEGIDA protest and hate crimes at the day-level (left panel) and week-level (right panel) 60

Figure B.3. Event study: PEGIDA protest and hate crimes at the day-level controlling for weather on each day (left panel) and lagged cumulative protest (right panel) 60