我国居民医疗需求影响因素的实证分析--有序probit模型的半参数估计

Determinants of Chinese Residents' Demand for Medical Care——An Application of Semiparametric Estimation of Ordered Probit Model

本文以个人潜在医疗需求为潜变量,以表示患病情况的有序离散变量为被解释变量,建立了有序probit回归模型,目的在于考察年龄、收入、性别、教育等主要个人人口特征和经济社会地位变量对医疗需求的影响。本文实证分析采用的是中国健康与营养调查2004年的微观调查数据,采用半参数方法对模型进行了估计。估计结果显示:年龄、性别、婚姻状态、居住在农村、收入水平、教育水平对个人的医疗需求有不同程度的影响。实证分析结果对于预测我国医疗费用的发展趋势和缓和医疗服务中的不平等问题都有参考价值。

To investigate determinants of individual demand for medical care, we build up an ordered probit model where the extent of morbidity is the explained variable and variablesindicafing individual demographic characteristics and social-economic status are chosen asexplaining variables, with the demand for medical care being the corresponding latent variable. This econometric model is estimated by semiparametric estimation using data from China Health＆ Nutrition Survey in 2004. Regression results show that age, gender, marriage status, living in rural areas, incomelevel and education have effects on demand for medical care to various extents, which is useful when forecasts of national medical expenses in the future are to be made and when measures are to be made to mitigate inequality in medical care consumption.