Adoption Curves and Social Interactions

This paper considers the observational implications of social influences on adoption decisions for an environment of perfect foresight adopters. We argue that social influences can produce two observable effects: 1) discontinuities in unconditional adoption curves and 2) pattern reversals in conditional adoption curves, in which earlier adoption is found for one group of actors versus another when "fundamentals" suggest the reverse ordering should occur; in turn the presence of either of these features can, under weak assumptions, be interpreted as evidence of social influences. As such, these properties are robust implications of social effects.