Recent studies have shown that individuals of species that live in groups tend to have high annual survival, but this link has lacked a theoretical explanation. We evaluate two hypotheses that explain how longevity could have led to the evolution of group living. The first is the territory inheritance hypothesis, and it proposes that longevity increases the probability of nonbreeding subordinates surviving long enough to have the opportunity of inheriting their natal territory. Second, we propose a novel hypothesis, the reciprocal altruism hypothesis, which is that longevity increases local dominance by favoring nonaggression pacts among neighboring residents because longevity increases the likelihood of reciprocal altruism. Birds thus accept subordinate residency because the exclusion of nonlocal birds will mean that, if they survive long enough, they will be likely to actually achieve territory inheritance. The reciprocal altruism hypothesis is supported by a wider array of evidence; becomes progressively more powerful as longevity increases, thus producing a positive feedback; explains the evolution of local dominance (whereas the territory inheritance hypothesis assumes its existence); and provides an explanation for why cooperative breeding should be found more often in aseasonal environments.