**Background:** Fast-food restaurants near schools are a well-documented contributor to adolescent obesity, with 40% of students eating fast food daily and the majority of US schools having a fast-food restaurant within a 1-mile radius. Zoning restrictions have largely failed, especially in urban, racially diverse communities. This research applies the activity space framework from geography, positing that a nearby location becomes a draw when it is perceived as a social activity space where peers congregate. The authors hypothesize that students who strongly identify with their student community will perceive a nearby fast-food restaurant as such a space, increasing patronage, and that social liability messages (but not health liability messages) can counteract this effect.
**Methods:** Six studies were conducted. Study 1 was an observational analysis of 5,986 eleventh-grade students from the California Healthy Kids Survey, combined with GIS data on fast-food restaurant locations. Studies 2–6 were experimental. Study 2 (N=153) used a behavioral field experiment with actual coupon redemption at near vs. far restaurant locations, manipulating identification via an essay task. Study 3 (N=188) tested mediation through perceived social activity space. Study 4 (N=251) tested a mild social liability message (a social media post). Study 5 (N=178) tested a stronger social liability message (a student boycott poster) among participants all manipulated to feel strong identification. Study 6 (N=379) tested a health liability message emphasizing unhealthy food.
**Key Results:** In Study 1, 94% of students had a fast-food restaurant near school. There was no main effect of nearness on patronage (b=0.10, p=0.40), but identification negatively predicted patronage (b=-0.16, p<0.001), and the interaction was significant (b=0.23, p<0.01). Floodlight analysis showed the nearby location effect was significant at identification levels ≥4.25 (p<0.05). In Study 2, strong identifiers redeemed coupons at higher rates when the restaurant was near vs. far (44% vs. 7%; b=1.27, p=0.001), while weak identifiers showed no difference (28% vs. 19%; p=0.35). Study 3 confirmed mediation: among strong identifiers, the effect of nearness on patronage intent was mediated by perceiving the restaurant as a social activity space (indirect effect=0.18, 95% CI 0.01–0.46). In Study 4, a three-way interaction (F(1,238)=6.89, p=0.009) showed that a social liability message eliminated the nearby location effect among strong identifiers (M=4.87 vs. 5.24, p=0.34), whereas a control message preserved it (M=5.48 vs. 4.10, p=0.009). Study 5 replicated this with a boycott message: strong identifiers showed higher intent for near vs. far under control (M=4.75 vs. 3.32, p<0.001) but not under the social liability message (M=3.44 vs. 2.91, p=0.130). Study 6 found that a health liability message failed to reduce attraction among strong identifiers (M=4.31 vs. 3.26, p<0.001) and actually increased attraction among weak identifiers (M=4.62 vs. 2.96, p<0.001), producing a three-way interaction (F(1,370)=4.97, p=0.03).
**Clinical Implications:** These findings challenge the prevailing reliance on zoning restrictions to address fast-food proximity to schools. Instead, school-based educational and social marketing interventions that reframe fast-food restaurants as social liabilities—through student activism, boycotts, or peer messaging—may be more effective. Health-focused messages alone are not only ineffective but may be counterproductive, particularly among weakly identified students who may experience reactance. The results highlight the importance of addressing social perceptions rather than merely health knowledge in adolescent obesity prevention efforts. Limitations include a focus on patronage rather than actual consumption, use of college students in most studies, and lack of data on what was purchased or with whom students ate.