We study the effects of the size of older cohorts on labor force participation (LFP) and wages of older workers. In the standard relative supply framework usually applied to relative cohort size, we would expect larger older cohorts to experience lower wages and hence lower employment or LFP. However, there are two reasons that we might find a positive effect. First, we might expect the age structure of the population to affect the composition of consumption and hence labor demand; it is possible that the age structure of employment is such that relative labor demand for an age cohort increases when the relative size of that cohort increases. Second, a large older cohort implies that the old cohort is large relative to at least some other narrowly-defined age cohorts. If two age cohorts are substitutable, then a decline in the relative size of one of them can imply an increase in the relative demand for the other.
We use panel data on states, treating the age structure of the population as endogenous, owing to migration. We find that when older cohorts are large relative to a young cohort, the evidence fits the relative supply hypothesis. But when older cohorts are large relative to 25-49 year-olds, the evidence points to a relative demand shift. Thus, we need a more nuanced view than simply whether the older cohort is large relative to the population; the cohort they are large relative to matters.