无论对个人还是社会来说，基础教育都至关重要，从宏观经济政策的角度出发，公共教育支出的增加能否同向增加个体获得的教育是十分值得关注的问题。借鉴文献经验，本文采取理论分析与实证分析相结合的分析方式。理论部分以公共物品“挤出效应”的理论为基础，以效用最大化的方式推导出公共教育支出变化对个人教育支出的影响。对于效用函数为柯布-道格拉斯形式的个体，增加公共教育经费用于提高学校教育质量会使其用于强制性教育的支出不变，而用于自主教育的支出和教育的总支出都会减少。实证部分使用2016-2018年县级生均教育经费总支出数据和CHIP2018的微观数据对基础模型和滞后模型进行OLS回归和Heckman两阶回归分析，结果显示，公共教育支出对家庭教育支出存在显著的挤入效应，且小学阶段的挤入效应大于初中阶段；小学阶段短期表现为挤出效应，而长期则为挤入效应；初中阶段长短期均表现为挤入效应。异质性分析结果显示，小学阶段农村的挤入效应大于城市，而初中阶段则相反；该影响在短期的城乡差异较小，而在长期较明显。学校性质、学生成绩、家庭学生数和地区生产总值有较大可能对挤入效应的城乡差异存在影响。

Basic education is very important for both individuals and society. From the perspective of macroeconomic policies, whether the increase in public education expenditure can increase the education that individuals obtain in the same direction is a great concern. Learning from literatures, this article adopts a combination of theoretical and empirical analysis. The theoretical part is based on the "crowd-out effect" theory of public goods, and maximize utility to deduce the personal-education-expenditures elasticity of public education expenditures. For individuals whose utility function is in the form of Cobb-Douglas, increasing public education funds for improving the quality of school education will keep their expenditures on compulsory education unchanged, while expenditures on autonomous education and total education expenditure will decrease. The empirical part uses the average education expenditure data on county-level from 2016 to 2018 and CHIP2018 micro data to perform OLS regression and Heckman two-order regression on the basic model and the lag model. It suggests that public education expenditure has a significant crowd-in effect on family education expenditure and the effect at the primary school stage is greater than that at the junior high school stage, that primary school stage shows a crowd-out effect in the short-term and a crowd-in effect in the long-term, that the junior high school stage shows a crowd-in effect in both short and long term. According to the heterogeneity analysis, the crowd-in effect in rural areas at the primary school level is greater than that in the cities, while the opposite is true at the junior high school level, and the differences of the impact is smaller in the short-term between urban and rural areas, but more obvious in the long-term. The type of the school, student’s performance, the number of students in the family, and regional GDP are most likely to have an impact on the differences of the crowd-in effect between urban and rural areas.