Abstract
'Demographic transition theory' assumes that fertility decline is irreversible. This commonly held assumption is based on observations of recent and historical reductions in fertility that accompany modernization and declining mortality. The irreversibility assumption, however, is highly suspect from an evolutionary point of view, because demographic traits are at least partially influenced by genetics and are responsive to social and ecological conditions. Nonetheless, an inevitable shift from high mortality and fertility to low mortality and fertility is used as a guiding framework for projecting human population sizes into the future. This paper reviews some theoretical and empirical evidence suggesting that the assumption of irreversibility is ill-founded, at least without considerable development in theory that incorporates evolutionary and ecological processes. We offer general propositions for how fertility could increase in the future, including natural selection on high fertility variants, the difficulty of maintaining universal norms and preferences in a large, diverse and economically differentiated population, and the escalating resource demands of modernization.