How much does it cost a family to raise a child from an early age through college graduation? This is probably a question that many people find hard to calculate and answer offhand.

　　For seven or eight consecutive years, Wei Yi of the China Institute for Educational Finance Research at Peking University has been studying this issue.

　　Getting this straight is not easy, because it requires rigorous data as support, and obtaining the data requires rigorous and standardized sample surveys. Once a survey is involved, it requires funding and manpower support behind it—and not a small amount. Of course, the main reason is that China is too large: from north to south, from east to west, between cities and between urban and rural areas, and across different provinces, there are great differences.

　　After years of research, the institute now has some results. In 2019 it released the "2017 China Education Finance Household Survey Report"; in January this year, the "China Education Finance Household Survey Report (2021)" was released, based on analysis of the 2019 China Education Finance Household Survey and its follow-up tracking data.

　　Through individual, scattered data points, commonalities are identified, and then an overall picture emerges. Some of the conclusions match most people's feelings, while others differ from people's everyday impressions—but data does not lie.

　　4% was reached, and household investment in education also increased

　　That national fiscal spending on education should account for 4% of GDP was a social issue for many years. In the more than a decade that followed, China's public fiscal input into education continued to increase, and in 2012 the 4% target was achieved.

　　Driven by mobile internet technology, online education companies and edtech companies developed rapidly in 2015–2016. Some traditional tutoring companies not only achieved scale but also expanded from first- and second-tier cities to third- and fourth-tier cities.

　　Researchers observed that while fiscal spending on education grew rapidly, official statistics showed that the share of non-fiscal education spending in GDP had been declining. At the same time, household investment in children's education continued to increase.

　　"Through personal experience, media reports and public discussion, everyone has felt the burden of family education—not only the financial burden, but also the burden of time and energy brought by intensive parenting. At the time I wanted to understand household investment in education." Wei Yi, who had just returned from Michigan State University in the United States with a PhD in philosophy to work in China, saw UNESCO data showing that household investment in education actually constitutes a large part of total education investment. Only by considering government public input, input from other institutions, and private household input at the same time can one have a more comprehensive understanding of total education investment—but "more than 60% of countries lack data on private education expenditure, especially household education expenditure data."

　　She learned that South Korea had conducted related surveys, and the United States had similar ones. Some other countries, in order to fully understand total investment in education, tried to integrate education-related data scattered across different departments and institutions. For example, France began building a national education account as early as the 1970s, integrating data related to education input and expenditure scattered across various institutions—including central and local government education fiscal input data, school financial data, household survey data, institutional account data and corporate tax data—under a unified framework.

　　Household surveys are the best choice for understanding private household investment in education while also carrying sufficient information.

　　So the China Institute for Educational Finance Research at Peking University, in cooperation with the China Household Finance Survey and Research Center at Southwestern University of Finance and Economics, carried out the "China Education Finance Household Survey," conducting door-to-door surveys in all provinces except Tibet, Xinjiang, Hong Kong, Macao and Taiwan, covering more than 40,000 households. The survey included educational opportunities, in-school and out-of-school education expenditure, and government subsidies received (tuition waivers, scholarships and grants, etc.).

　　The survey found that in the 2018–2019 school year, average household education expenditure nationwide was 11,300 yuan, with families spending an average of 8,139 yuan per child. Average household education expenditure in urban areas was 14,200 yuan, 1.7 times that of rural households (8,205 yuan).

　　For one child from three years of preschool through college graduation, a family's education expenditure is about 233,000 yuan. Among them, the 20% of households with the lowest spending level spent about 180,000 yuan, the middle 20% spent about 224,000 yuan, and the top 20% spent about 424,000 yuan.

　　In this study, household education expenditure includes, for the in-school part, tuition, school uniform fees, meal fees, exam fees, accommodation fees, physical examination fees, etc.; the out-of-school part includes out-of-school internship costs, interest classes, online education products and services, etc.

　　Based on the survey results, Wei Yi estimated that the overall scale of household education expenditure across all schooling stages nationwide is about 2,163.21 billion yuan, equivalent to 2.4% of 2018 GDP. The "China Education Finance Statistical Yearbook 2019" shows that in 2018 total national education funding was 4,614.300 billion yuan, of which national fiscal education funding was 3,699.577 billion yuan, accounting for 4.11% of GDP.

　　Analyzed from the perspective of family burden, the lowest household education burden rate across schooling stages is at the compulsory education stage, followed by the preschool stage. The household burden rate rises at the high school and college stages. By the college stage, the household education burden rate rises, and "the burden of preschool education is close to that of high school." The study found that for rural households, education expenditure for each college student accounts for 35% of total household expenditure.

　　 In 2018, China carried out governance of off-campus training institutions, mainly targeting six types of problems: safety hazards, operating without licenses, teaching beyond the syllabus, exam-oriented tendencies, schools teaching from a non-zero starting point, and teachers not teaching in class but teaching after class.

　　After this, the average participation rate in off-campus tutoring and interest classes among primary and secondary school students declined. At the same time, another feature emerged: a trend of polarization in out-of-school investment among different groups. Students whose families had more economic and socio-cultural resources not only participated more in off-campus training, but their families' investment also continued to increase.

　　Some extracurricular training institutions also began to manufacture anxiety, which the media called an education "arms race."

　　In February 2021, on Chinese New Year's Eve, CCTV's Spring Festival Gala—surprisingly—had several training institutions as major sponsors, with Yuanfudao's advertisements bombarding viewers until midnight. Ironically, that day's program also included a sketch, "Such Parents," satirizing parents frantically signing their children up for extracurricular tutoring classes.

　　In July 2021, the "double reduction" policy was implemented. Wei Yi and her colleagues also conducted follow-up research: "We found that spending on academic extracurricular training was declining, while spending on interest-based classes rose slightly." This conclusion is fairly consistent with the findings of other education researchers.

　　The more educated the mother, the more she "chickens" her kids

　　From several years of survey data, Wei Yi observed characteristics of Chinese households' investment in education. For example, the more educated the parents—especially the mother—the higher the household's investment in children's education.

　　Looking back at earlier education research around the world, many countries showed a "son preference" over past decades, but with declining fertility trends in various countries, some studies found this "preference" disappearing. In China, affected by the one-child policy for many years, nuclear families consisting of a couple and an only child dominated in cities. According to the "China Family Development Report" released by the former National Health and Family Planning Commission in 2015, households with three or fewer members accounted for 70.2%, while nuclear families consisting of a couple and an only child accounted for 64.3%.

　　The study found that among one-child families, an only daughter could obtain more out-of-school education opportunities and investment, especially in non-academic interest-expansion areas.

　　"It should be said that in education investment, the gender difference is not significant. High-income families invest somewhat more in girls' out-of-school education—piano, vocal music, fine arts, language training and other arts-interest categories—girls get somewhat more," Wei Yi said. Another reason, she said, is that girls are more "obedient," understand their parents' thinking, and are willing to meet their parents' expectations.

　　Low-income and high-income families also differ in their investment in children's education. Low-income families tend to choose "conditionally 'chickening' their kids." First, low-income families are more sensitive to academic performance: the better the child's grades, the more likely the family is to increase education investment. Second, low-income families are more sensitive to opportunity cost. As the schooling stage rises, family investment increases substantially, especially at the high school stage.

　　On the other hand, high-income families tend to "unconditionally 'chicken' their kids." Wei Yi observed that "compared with low-income families, in high-income families the relationship between children's academic performance and parents' education level, on the one hand, and family investment, on the other, is relatively weak." Especially regarding academic performance: whether a child's class ranking was good or bad did not cause the family to invest more or less in their education. Second, high-income families are insensitive to opportunity cost and tend to substitute quality for quantity and out-of-school for in-school. For high-income families, one-child families—especially those with an only daughter—are more likely to choose to "chicken" their kids.

　　Overall, compared with middle- and high-income families, the investment model chosen by low-income families is more oriented toward advancing to the next level of schooling.

　　China's demographic data show that from 1995 to 2019, the age at which Chinese women and men had children gradually increased. The average age at which women had children rose from 26 to 30, while for men it rose from 27 to 32.

　　Based on these data, Wei Yi also found an interesting phenomenon: whether parents had children early or late has a certain effect on education investment, but this phenomenon is reversed between urban and rural areas.

　　In urban areas, young parents invest more in their children's education. Within the same age group, the older the parents were when they had children, the more they invested in their children's education. "The results for urban areas indicate that two causes of low fertility—delayed childbearing age and increased investment in children's education—may reinforce each other in the trend of lowering the total fertility rate."

　　In rural areas, the opposite is true: young parents invest less in their children's education. Within the same age group, the older rural parents were when they had children, the less they invested in their children's education. Wei Yi believes this deserves further study.

　　"Of course, survey research results on household education expenditure can only reflect averages and trends; some extreme values can only be explained through case analysis," Wei Yi said.

　　The relationship between education input and returns is complex

　　In July 2021, the "double reduction" policy was implemented. Wei Yi led team members to use a consumer review website and match it with geographic information to count off-campus tutoring institutions one by one for 2018–2019, then conduct a retrospective re-check in 2023, finding that 88% of primary and secondary school academic tutoring institutions had ceased operations.

　　Their survey statistics of students showed that the off-campus academic tutoring rate was 24% in 2019 and fell to 17% by 2022. They also found that after "double reduction," household education investment was not much affected for low-income and high-income families—especially, out-of-school training spending among high-income groups was unaffected—while what changed was middle-income families. It can be said that low-income families face multidimensional challenges.

　　Wei Yi found that most middle-income and below families in China are "very willing" to spend on their children's education.

　　They made a preliminary estimate of household education expenditure as a share of total household expenditure: 14.9% nationwide, 15.8% in rural areas, and 14.1% in urban areas. By region, the share of household education expenditure in the northeast was noticeably higher than in other regions (18.8%), followed by the central and western regions at 15.7% and 14.4% respectively, while the eastern region was lowest at 14%.

　　Looking at the share of out-of-school education expenditure at the preschool and basic education stages, the share of household out-of-school education expenditure at the basic education stage is relatively high—especially at the primary and junior high stages, averaging 26.2% and 18.4% of household per-student education expenditure. Among them, urban areas are far higher than rural areas, at three times the rural level. At the high school stage, the share of out-of-school expenditure drops substantially, averaging 11.7% for regular high schools and 2.6% for secondary vocational schools. Regular high schools still show urban-rural differences: the rural out-of-school expenditure share is 5.5%, and the urban share is three times the rural share.

　　The largest urban-rural difference in household education expenditure across schooling stages is at the primary stage, followed by the preschool and junior high stages; the difference gradually narrows at the high school stage, with secondary vocational schools even showing a reversal between urban and rural areas, while at the college stage the urban-rural difference in household education expenditure is even smaller than at the high school stage.

　　Wei Yi noted that this study did not include economic costs such as income parents lose from caring for children or families buying homes in school districts. In some county towns, "parent villages" have long existed; in some places where "super high schools" are located, parents accompanying their children for schooling have formed a small town, and the names of key universities—"985" and "211"—are often written into the names of local restaurants and supermarkets. On the other hand, in big cities, buying a school-district home appears to be education investment, but sometimes it can yield investment returns.

　　As for how these complex concepts—summarized in economics as "sunk cost" and "opportunity cost"—should be reflected in research on education investment, she thinks a "time use" survey could be done in the future.

　　After several years focused on one research topic, Wei Yi's biggest impression is that Chinese parents are very willing to invest in their children's education and hope for returns, but it is not easy to clarify the relationship between input and returns: "The results of education investment cannot be evaluated in the short term."

　　There have been cases where off-campus training institutions, to let parents see results, bought exam papers from nearby schools so that students could score high, leading parents to mistakenly believe the training had worked.

　　"Unlike other services, where whoever pays enjoys it, education investment is paid for by parents but used by children, and their feelings differ." Wei Yi has fully felt the complexity of education: education results rise in a spiral, and a child's growth is a curve—none of which is easy to evaluate.