Abstract
One of the most significant demographic changes predicted for this
mid-century is in the age distribution of the U.S. population. Few aggregate
empirical studies in the alcohol demand literature account for the entire age
distribution of a jurisdiction in the statistical analyses. Herein, we propose
the use of the attraction CODA empirical framework to model age cohort impacts
to state-level beer, wine and spirits consumption shares in the U.S. for the
years 2008 - 2020. This compositional construct, based in simplicial geometry,
allows researchers to keep intact the entirety of states' age distributions
without designed transformation. Age related results show that declines in
states' beer shares are attributed to, mainly, proportional increases of states'
35-54 year-old cohorts. States with increasing older to elder population
proportions experienced increases in wine shares and states' 35-54 year-old
cohorts again, not younger populations, are key in driving the increased
states' spirits consumption shares.
JEL classification numbers: C10, C19, L60.
Keywords: Age cohorts,
Alcohol consumption, Compositional regressions.