The Construction and Use of Approximations for Missing Quarterly Observations: A Model-Based Approach
用基于模型的方法为荷兰劳动力市场年度数据推导季度估计值，作为代理变量用于估计结构方程，并分析代理变量估计方法的性质。
Abstract We use a model-based approach to derive quarterly figures on several variables for the aggregate labor market in the Netherlands that are only observed annually. These approximations are conditional expectations derived from univariate and bivariate quarterly time series models for the series under consideration. They are subsequently used as proxies to estimate and analyze the structural labor market equations. Attention is given to the properties of estimation procedures based on proxy variables. KEY WORDS: Proxy variablesInterpolationSmoothingStructural modelKalman filter