I have a new paper on the age-time-cohort identification problem. The substantive issue we address is the age profile of cross-sectional variance in income (a key statistic for heterogeneous agent models). 1/