2020-2035年是实现教育现代化目标的关键十五年。本文收集了2000-2019年间的年龄别人口、出生率和死亡率数据，利用队列要素法对2020-2035间基础教育学龄人口规模进行了推演。考虑到基础教育阶段，尤其是义务教育阶段，大部分学生可能存在就近上学的偏好，以及城乡之间可能存在的义务教育阶段的教育质量差距，因此笔者认为有必要进一步预测学龄人口城乡分布，以提供更加精细的教育规划制定依据。

在队列要素模型预测期参数设定上。对于分年龄生育率的预测，笔者首先使用GM（1，1）模型预测总和生育率，再根据各分布模型的简易程度和对实际数据的拟合效果，选择使用对数正态分布，基于历史数据利用最小二乘法估计分布中的参数，将预测的总和生育率数据代入估计得到的对数正态分布模型即可计算出预测期的分年龄生育率。对于分年龄死亡率的预测，则使用了当前国际上普遍使用的Lee-Carter模型，基于历史数据使用最小二乘法估计出了模型的部分参数，并基于单位根检验和信息准则选择合适的ARIMA(p,d,q)模型预测2020-2035年间时变因子的变化，最后代入模型中计算分年龄死亡率。

将对2020-2035年分城镇、分年龄生育率和和分城镇、分年龄、分性别死亡率的预测值代入队列要素法的模型设定中，并取6-11岁为小学阶段，12-14岁为初中阶段，15-17岁为高中阶段，即可推算出2020-2035年基础教育学龄人口数。进一步对基础教育在校生城镇化率进行预测，结果发现：一、2020-2035年间，全国基础教育阶段在校生数将经历一个短期（4年）缓慢上升到长期（11年）下降过程，且城区在校生总量超过镇区和农村在校生数量。二、我国基础教育总体进入以城区教育为主体的时代。三、农村在校生数量逐年下降，其中小学减幅最大。四、小学阶段在校生数量先上升后下降，2023年达到峰值，在空间分布上城区小学在校生数量占比大于镇区大于农村地区。五、初中阶段在校生总量先上升后下降，2029年达到峰值。空间分布上城区初中在校生数量2028年超过镇区，农村初中在校生数量仍相对较小。六、高中阶段在校生总量先上升后下降，2032年达到峰值，空间分布上城区在校生规模大于县镇大于农村。笔者认为应以长期学生总量下降的变化趋势作为发展的契机，不断地深化教育教学的改革，不仅在数量上满足人民群众对教育资源的需求，更在教育质量上不断符合经济社会发展的需要。

We collected age-specific population, birth rate and mortality data from 2000 to 2019, and mainly used the cohort-component method to deduce the scale and urban-rural distribution of the fundamental education population (including primary, junior high and high school) in the key fifteen years of the future 2020-2035 educational modernization plan. CCM model (CCM is short for cohort-component method) derives the future age- and sex-specific population based on historical age- and sex-specific population data,  projections of future age-specific fertility rate, birth-sex ratio of newborn babies, and age- and sex- specific mortality rate. Considering that at the fundamental education stage, especially the compulsory education stage, most students may prefer going to school near to where they are resident in, and there may be a gap in the quality of education between urban and rural areas, we believe that it is necessary to make projections about urban-rural distribution of school-age population to provide a more refined basis for educational planning.

For the projection of age-specific fertility rate, we first used the GM (1,1) model to predict the total fertility rate, and then choose the lognormal distribution according to the simplicity of each distribution model and the fitting effect to the historical data. The parameters in the distribution model were estimated by the least square method. Substituting the predicted total fertility rate data into the estimated lognormal distribution model, we calculated the age-specific fertility rate in the prediction period. For the prediction of age-specific mortality, we used the Lee Carter model which is widely used in mortality projection all over the world. Based on the historical data, some parameters of the model were estimated by the least square method, and an appropriate ARIMA (P, D, q) model was selected based on the unit root test and information criteria to predict the time-varying factors from 2020 to 2035. Finally, substituting those parameters and time-varying factors into Lee Cater model, we calculated the age-specific mortality.

Taking 6-11 years old as primary school stage, 12-14 years old as junior middle school stage and 15-17 years old as senior high school stage, the school-age population of fundamental education in 2020-2035 can be calculated. Further predicting the urbanization rate of students in fundamental education, the results show that: First, from 2020 to 2035, the number of students in fundamental education in China will experience a short-term (4 years) slow rise and then a long-term (11 years) decline. Secondly, the total number of students in urban areas will exceed the number of students in towns and rural villages. Third, the number of students in rural area will decrease year by year, especially in the primary school. Forth, the number of primary school students will first increase and then decrease, reaching a peak in 2023, and the proportion of primary school students in urban areas is greater than that in towns and rural areas. Fifth, the total number of students in junior middle school will first increase and then decrease, reaching a peak in 2029, and the number of junior middle school students in urban areas will exceed that in towns in 2028, and the number of junior middle school students in rural areas will remain relatively small. Sixth, the total number of students in senior high school will first increase and then decrease, reaching the peak in 2032, and the scale of students in urban areas is larger than that in counties or towns, and larger than that in rural areas. We should take the long-term downward trend of the total number of students as an opportunity for education development, and constantly deepen the reform of education and teaching, so as to meet the people's demand for educational resources in quantity and meet the demands of economic and social development to education quality.