China’s birth rate has witnessed a rapid decline from 21.06‰ in 1990 to 12.07‰ in 2015, especially after the year 2000, and has been in a very low level for a long time since then. Accompanied by the decline, the number of primary schools in China has also been dramatically reduced from 553622 in 2000 to 228585 in 2012, which is known as the “Compulsory School Merger Program”, starting from 2001 and ending in 2012. The number of schools decreased by 51.9%, while the number of primary school students decreased by only 25%, which led to the longer distance to school and the rise in the ratio of students to teachers. It means that education costs will increase sharply, which depresses the demand for fertility. This paper aims to explore whether the Program will lead to the decline of fertility.

Based on the fact that the Program leads to the rise of parenting costs and the ideas that the rising cost suppresses the demand for fertility, this paper uses the cross-sectional DID method and the instrumental variable method to match the rural household data of CHIP2013 with the primary school data of 2000−2012 based on prefecture-level cities, and provides the evidence of the effect size and how the Program affecting fertility. The results show that, from 2000 to 2012, if the change percentage of primary school merger exceeds the change percentage of students by 0.3（0.5）units, the average inhibitory effect of the policy on the number of children in the family is about 0.26（0.43）. It is found that during the period of 1999−2013, if the change percentage of the number of primary schools per year exceeds the change percentage of students by −0.03, the Program can explain about 13% of the decline in the birth rate during the sample period.

This paper further shows the mechanism behind the negative effect. The rising ratio of students to teachers increases the intensity of family competition for educational resources, and the longer distance to school increases parenting costs. The Program further restrains fertility by increasing the ratio of students to teachers and the ratio of students to schools. In the areas with higher population density, because of the stronger competition among people for educational resources, the negative effect of the Program is even greater. In the areas with more paved roads, the negative effect of the Program is relatively low, since the convenient transportation can partially “weaken” the rising cost caused by the long distance to school.

This paper may be the first one to explore, identify, and estimate the effect size and the mechanism of the Program’s effect on fertility. Many people believe that the decline in fertility has caused school merger. However, this paper provides empirical evidence for the decline of fertility caused by the Program, which can also be used as evidence for the increase of family compulsory education costs and the reduction of fertility. The policy suggestions of this paper are that: on the one hand, the government can improve and solve the problem of the accessibility of primary school compulsory education, and “compress” the time and space distance to school; on the other hand, we should increase the number of full-time teachers and reduce the class size, reduce the family’s competition for educational resources, reduce the cost of compulsory education, and then “release” the fertility demand.