Household projections at the national and sub-national levels by statistical offices and market analysis agencies still mostly employ the classic headship-rate method. However, the headship-rate method suffers several serious shortcomings and has been criticized widely by demographers for about three decades (Bell & Cooper 1990; Murphy 1991; Mason and Racelis 1992; Spicer et al. 1992; Budlender, 2003; Zeng, Land et al., 2006; Christiansen and Keilman, 2013; Wilson, 2013; Berard-Chagnon, 2015). The criticisms may be summarized in four points. First, the designation of a household head is a vague, ill-defined, and arbitrary choice that is not easy to model, making projections difficult. Second, given the nature of cross-sectional extrapolations of the headship-rates method, it cannot be linked to demographic rates; this makes it impossible to incorporate projected or assumed changes in the propensity/timing of demographic processes. Third, the information on households produced by headship-rate projections is very limited and inadequate for detailed planning and analysis. Fourth, the headship-rate combines all “non-head” household members into one category with no projected information. This makes it impossible to study the household status and living arrangements of the elderly, children and non-head adults who are also significantly relevant to business/academic/policy analysis and planning.