Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper shows that most methods of incorporating demographic variation into separable models can be represented in a form that is identical to Barten equivalence scales, except that the scales themselves depend on the exact mix of goods that comprise each group, as well as on demographic variables. This generalization of Barten scales is shown to be more plausible than ordinary scales, can be used to increase the efficiency of demand system estimation, and can overcome Muellbauer's under-identification result for cross-sectional estimation of equivalence scales.