Hate in Context
Week 9: Reducing Hate Crime I — Contact
Fickle Prosocial Violence Response model
SSM support increases post-Pulse
Pulse: June 2016 massacre at Pulse, an LGBTQ+ nightclub in Orlando
Post-Pulse respondents more supportive of same-sex marriage
Pulse effect attenuates over time
The prosocial shift is short-lived, consistent with the “fickle” FPVR prediction
Anti-gay attitudes after Pulse
Top: post-Pulse cuts implicit and explicit anti-gay bias (D-score, straight bias, heterocentrism) across bandwidths.
Bottom: that reduction holds for ~50 days, then fades. Same “fickle” pattern as SSM support
Meaningful contact + physical touch
HBS experiments: pairs who shake hands cooperate more and reach better joint deals than pairs who do not.
Schroeder et al. (2019)
Experimental design
- Setting: Mohamed Salah joins Liverpool FC (June 2017), a highly visible, practicing Muslim star
- Design: synthetic control for Merseyside hate crimes, Twitter among LFC fans vs other clubs, and a survey experiment
- Treatment: parasocial exposure to a successful, openly Muslim celebrity athlete
- Outcomes: hate crimes, anti-Muslim tweets, and survey attitudes toward Islam/Muslims
Does Salah fulfill the criteria for successful parasocial contact?
✓ Repeated exposure
✓ Positive/likable portrayal
✓ Salient out-group identity
Figure 1 · Attitudes toward Islam
Context: share saying Islam clashes with British values rises through 2015–17, then eases slightly.
Figure 2a · Hate crimes: observed vs imputed
After Salah joins, Merseyside hate crimes fall ~16% below the synthetic (imputed) counterfactual.
Figure 2b · Hate crimes: treatment vs placebos
Merseyside’s post-treatment drop stands out against placebo units (other police stations), not a generic UK trend.
Figure 3 · All crime types in Merseyside
Decline is specific to hate crimes. Other Merseyside crime types do not show the same pattern.
Figure 4a · Twitter: observed vs imputed
Liverpool fans’ anti-Muslim tweets fall ~50% below the imputed path after Salah arrives (3.8% vs 7.3%).
Figure 4b · Twitter: ATT vs other clubs
Liverpool’s ATT is more negative than placebo clubs. The effect is concentrated among LFC fans.
Figure 5 · Survey evidence
Religiosity cue raises pro-Muslim/“Islam compatible” attitudes. Weaker or null on commonality and immigrants.
Conditional on performance?
“I am a German when we win and an immigrant when we lose.”
“When things were going well, … they were calling me Romelu Lukaku, the Belgian striker. When things weren’t going well, they were calling me Romelu Lukaku, the Belgian striker of Congolese descent.”
“You get a bit older and you realise that if you’re playing badly that’s when media starts saying ‘Nigerian-born’… when you do well you’re British; when you’re not doing so well you’re the Nigerian.”
Conditional on positive coverage?
Conditional on not taking political stances?
Colin Kaepernick: kneeling during the U.S. national anthem (from 2016) against racial injustice and police violence
Experimental design
- Setting: amateur soccer leagues for displaced men in post-ISIS Northern Iraq
- Sample: Christian players randomized within league teams
- Treatment: ~3 Muslim teammates on an otherwise Christian squad (~25% Muslim)
- Control: all-Christian teams in the same leagues
- Outcomes: on-field vs off-field behaviors + attitudinal indices (weeks to months later)
Does this experiment fulfill the criteria for successful intergroup contact?
✓ Equal status
✓ Common goals
✓ Cooperative environment
✓ Institutional support
Figure 1 · Behavioral results
On-field: more likely to train with Muslims, vote a Muslim for sportsmanship, and sign up for a mixed team next season. Off-field: little change for a Mosul restaurant visit or a mixed social event.
Did competition create hostility?
Yellow/red cards as a proxy for on-pitch aggression:
- Card rates do not differ by match type (mixed vs all-Christian)
- No extra hostility when mixed teams play all-Christian teams
- Referee’s religion also unrelated to carding
→ Competitive contact here does not look like backlash/escalation on the pitch.
Figure 1 · Experimental treatment arms
Table 2 · Negative attributes
No meaningful prejudice reduction from UYVT, mixed classes, or mixed partners. UYVT shows small increase in negative attributes among Christians.
Table 3 · Positive attributes
No contact effects on endorsement of positive out-group attributes. UYVT shows small negative effects among Christians.
Table 4 · Out-group evaluation
Bipolar “thermometer” ratings (lazy–hardworking, etc.), not stereotype checklists. Same story: no contact gains; tiny negative UYVT effect among Christians.