Ties with your neighbourhood, before anything happens

It works if the event has a local phase, which is how almost all of them reach a household: blackout, natural disaster, pandemic, economic crisis.

Evidence. High, consistent across independent methods and with hard outcomes. In the Chicago heatwave of July 1995, a case-control study with 339 pairs matched by neighbourhood and age measured the risk of death from heat: not leaving the house daily, odds ratio 6.7; living alone, 2.3. What was protective was having social contacts, group activities or friendships in the area ^[948]Heat-related deaths during the July 1995 heat wave in ChicagoSemenza, Jan C.; Rubin, Carol H.; Falter, Kenneth H. et al. · 1996 · paperView source ↗verified through CrossrefAccessed on 9 September 2026. In 133 municipalities on the Tohoku coast after the 2011 tsunami, where mortality ranged from zero to ten per cent of the population, social capital strongly influenced the outcome ^[48]The physical and social determinants of mortality in the 3.11 tsunamiAldrich, Daniel P.; Sawada, Yasuyuki · 2015 · paperView source ↗verified through CrossrefAccessed on 9 September 2026. The evidence closest to causal comes from an intervention: in Ofunato, those who took part in a community centre created and run by older people from the neighbourhood reported greater recovery than comparable individuals who did not take part, with propensity score matching ^[632]Social capital building interventions and self-reported post-disaster recovery in Ofunato, JapanLee, Juheon; Aldrich, Daniel P.; Kiyota, Emi et al. · 2022 · paperView source ↗Accessed on 9 September 2026.

It does not work if there is no local phase. And the bridge towards gradual disempowermentGradual disempowermentThe idea that humanity could lose control over its own future without any catastrophe, simply because more and more economic, political and cultural decisions pass to AI systems.For exampleLike a town that hands each of its services to an outside company until one day it notices it no longer decides anything of its own. is a reasonable argument, not a measurement: none of these studies is about AI ^[47]Building Resilience: Social Capital in Post-Disaster RecoveryAldrich, Daniel P. · 2012 · bookView source ↗Accessed on 9 September 2026.

Cost. The lowest in money on the whole list and the highest in time. It cannot be bought.

What it does NOT solve. It stops nothing: it changes who recovers. And it cannot be improvised on the day of the event, because what these papers measure is pre-existing social capital.