Identifying Distributional Characteristics in Random Coefficients Panel Data Models
研究了固定T时线性个体特定系数面板模型的识别问题，证明了在条件不相关下效应方差的识别性，并讨论了误差服从自回归移动平均过程时个体效应概率分布的识别，最后用GMM估计矩并构建密度非参数估计，应用于母亲孕期吸烟对婴儿出生体重的影响。
We study the identification of panel models with linear individual-specific coefficients when <it>T</it> is fixed. We show identification of the variance of the effects under conditional uncorrelatedness. Identification requires restricted dependence of errors, reflecting a trade-off between heterogeneity and error dynamics. We show identification of the probability distribution of individual effects when errors follow an Autoregressive Moving Average process under conditional independence. We discuss Generalized Method of Moments estimation of moments of effects and errors and construct non-parametric estimators of their densities. As an application, we estimate the effect that a mother smoking during pregnancy has on her child's birth weight.