The Estimation of the Prais-Houthakker Model of Equivalence Scales
推导了Prais-Houthakker等价尺度模型的最大似然估计方法，适用于截面和面板数据，并基于英国家庭支出调查数据验证了模型不可识别的问题，加入先验信息虽可识别但模型因零替代假设而不合理。
Given a preference basis for the model, a household's general equivalence scale is seen to be a cost of living index relative to the reference household type defined at constant prices. A maximum likelihood estimation procedure which can be applied both to cross-section data and pooled time-series/cross-sections is derived. The lack of identification of the model is established theoretically and checked empirically on British family expenditure survey data. With prior information, e.g. a nutrition based food scale, identification can be reached but ultimately the model is rendered implausible because of its zero substitution implication and the empirical results bear this out.