根据《中国统计年鉴-2020》，我国总人口从1949年的5.4亿增长至1970年的8.3亿，平均每年增加约1500万。人口迅速扩张与当时的经济增长、资源承载能力不相匹配。为了抑制人口快速增长的趋势，1979年国家出台了独生子女政策（One-child Policy，简写为OCP），在城镇地区普遍实施，长达30余年。根据2015年1%人口抽样调查资料，2010-2015年的修正的总和生育率均值仅为1.18，远远低于维持人口数量不增不减的自然更替水平2.1。在出生率下降、死亡率上升的情况下，我国人口自然增长率急剧下滑，或将面临负增长。生育率的下降，一方面通过老龄化挤占家庭资源而抑制经济增长；另一方面通过全部家庭资源向少数小孩聚集、增加人力资本积累而促进经济增长。小孩数量的变化对人力资本积累的影响方向是决定经济发展的关键所在。尽管在计生政策限制小孩数量的条件下，家庭不能自主选择小孩数量和小孩质量。然而，计生政策的实施因职业、城乡而异，不同经济背景的家庭面临计生政策干预的严厉程度不一。基于此，本文以我国为例，将不同收入阶层的家庭纳入分析框架，放松家庭资源约束、借贷约束，分析家庭收入阶层对数量-质量替代效应的影响；放松父母对不同性别小孩一视同仁的假设，分析数量-质量替代效应的性别差异。从“双独二孩”“单独二孩”到“全面二孩”，在生育政策不断放松的背景下，考察小孩数量对质量的影响具有重要现实意义。

美国经济学家Becker于20世纪60年代提出数量-质量替代理论，没有考虑男孩偏好。原因有二：一是，在发达的北美国家，当时的总和生育率约为3.7，相对较高；二是，发达国家更偏好小孩性别的多样性，不存在性别偏好。然而，作为一个深受儒家思想影响和农耕文明发达的国家，男性在传宗接代、父母养老、市场劳动参与等方面具有比较优势，中国的男孩偏好观念较为强烈。在计生政策促使生育率探底的背景下，有男孩偏好的家庭借助胎儿性别鉴定技术，满足生育男孩的意愿，从而家庭内部的小孩性别结构向男孩倾斜。宏观是微观的总和，所有个体按照性别、年龄加总求和，汇集成地区层面的出生性别比失衡。根据第六次人口普查数据，我国0-4岁小孩的性别比高达1.20（根据联合国相关规定，出生性别比的正常区间为1.02-1.07），这意味着在未来的婚姻市场上，将近16.7%的青年男性无法匹配婚姻伴侣，从而可能引发一系列严重的社会犯罪和伦理道德问题。性别结构作为一个地区层面的宏观概念，大多数文献以地区作为研究对象，从宏观角度探讨性别失衡问题，无法度量男孩偏好，也难以探寻其失衡背后的微观动因。在人口性别结构扭曲的背景下，本文从微观家庭视角出发，探寻家庭对小孩进行性别选择的文化、经济成因。在此基础上，以性别比失衡的青年个体作为研究对象，本文进一步分析性别失衡与两性态度的因果关系，形成“父代性别态度→家庭内部的子代性别结构→地区层面的子代性别失衡→子代个体性别态度”的逻辑链条。这些研究丰富了性别失衡成因和后果等相关文献，具有一定理论价值。

家庭作为理性的经济主体，在子女数量、质量、性别之间进行选择，优化生育决策以达到效用最大化目标。首先，本文回答的问题是在我国计生政策背景下，家庭无法自主选择小孩数量，那么数量-质量替代关系是否存在？放松家庭资源约束、借贷约束，家庭经济背景对数量-质量替代有怎样的影响？放松对不同性别小孩一视同仁的假设，数量-质量替代是否存在性别差异？基于中国教育追踪调查（CEPS 2013-2014年）数据，以认知能力得分、兴趣辅导班参与及支出作为质量指标，本文运用OLS和MLE计量方法研究上述问题。研究发现：（1）数量-质量替代关系在各收入阶层均成立，即小孩数量与认知能力得分、兴趣辅导班参与及支出显著负相关，且经济中等阶层的替代效应最大，经济困难阶层和经济富裕阶层的替代效应相对较小。（2）尽管经济中等阶层小孩质量随着同胞数量的增加而下降更快，但控制个体、家庭、班级、学校特征后，家庭所处收入阶层仍然与小孩质量正相关，表明良好的家庭经济背景有助于小孩质量的提高。（3）就性别异质性来看，同胞数量增加对女孩的负面影响更大。（4）就计生政策实施的严格程度来看，受独生子女政策严格限制的家庭，同胞数量增加对个体有更大负面影响。（5）以分区县分年级的家庭生养小孩数量均值作为小孩数量的工具变量进行内生性讨论，以家长职业代理收入阶层进行稳健性检验，结论相似。

在计生政策限制小孩数量的背景下，有男孩偏好的家庭进行性别选择，以满足生养男孩的愿望，从而出生性别比失衡。我国作为世界上性别比失衡最严重的国家，引发国内外学者关注，但大多文献基于地区层面探讨性别失衡问题，难以挖掘其背后的微观动因，且无法度量地区层面的男孩偏好。本文基于2000年和2010年中国妇女社会地位调查数据，以微观家庭作为研究对象，运用Tobit及Logit计量方法，分析家庭对小孩进行性别选择的文化、经济成因。研究发现：（1）文化观念因素对小孩性别结构的影响主要表现在冠姓权和财产继承方面，持传统文化观念的家庭，小孩性别结构向男孩倾斜，表现为男孩数量占比上升，多孩家庭中最后一胎是男孩、一孩家庭中生养独生子的几率比增加。（2）家庭经济因素对小孩性别结构的影响主要表现在“养老”方面。有养老金、退休金的家庭，通过削弱“养儿防老”依赖，使小孩性别结构向女孩倾斜。然而，“是否有医疗保险”对小孩性别结构没有显著影响。（3）按照妻子出生年份，划分“不受计生政策影响”“受计生政策影响”两个子样本重新推算模型，发现在受计生政策影响的子样本中，文化观念和家庭经济变量的系数绝对值更大、显著水平更高，即在受计生政策影响的子样本中，性别比失衡更为严重。

家庭内部向男孩倾斜的性别结构组成地区层面的子代性别结构失衡，从而影响青年个体的两性态度。基于中国综合社会调查（CGSS 2010、2012、2013、2015年）微观数据和2010年第六次人口普查报告的性别比数据，本文按照受访者出生年份区分父代和子代，按照相关设问区分市场劳动和家务劳动两个维度的性别态度，运用OLS和2SLS估计方法，考察子代性别比、父代性别态度和子代性别态度之间的关系。研究发现：（1）父母的性别态度影响小孩的性别结构。性别态度，尤其是市场劳动维度的性别态度越传统，期望的和实际的男孩数量占比越高，子代性别比越高。（2）子代性别比对其性别态度的影响存在两种相反的传导机制：性别比失衡通过增强男性主义文化而使态度观念更倾向于传统，通过增强婚姻家庭中的女性议价能力而削弱传统的性别不平等态度。无论是对于市场劳动维度还是家务劳动维度的性别态度，性别比的系数均为正，表明性别比与性别态度正相关，即性别比越高，性别态度越传统，男性主义文化发挥主导作用。比较性别比对两个维度性别态度影响的大小，发现性别比对市场劳动维度的性别态度有更大且更显著的影响；对家务劳动维度性别态度的影响为正，但系数值较小且不显著或显著水平较低。这表明对于家务劳动维度的性别态度，性别比失衡通过增强女性在婚姻家庭内部的议价能力对传统性别态度产生负向影响，抵消了部分男性主义文化的正向影响。（3）父代与子代之间的性别态度正相关，即性别态度存在代际传递性。同一维度的性别态度在代际之间的传递性更强，而不同维度的性别态度在代际之间的相关性较弱。（4）本文进一步基于2005年1％人口抽样调查的20％样本量测算分省分年龄组的独生子女家庭数量占比，作为性别比的工具变量，运用2SLS计量方法重新推算模型，结论高度一致。

在回顾相关文献和理论的基础上，本文进行了实证研究，进而从降低家庭生育和抚养成本、削弱男孩偏好、提高教育水平、挖掘人口红利四个方面提出促使生育率回升、性别比回落、人力资本积累的政策建议。

本文的创新点主要表现在以下方面：基于微观家庭视角，将子女数量、质量、性别选择纳入家庭生育决策的同一逻辑框架；放松相关假设前提，拓展了数量-质量替代理论；通过分析第四、五、六次人口普查数据，以更准确的第六次普数据测算性别比，并辨析性别比的工具变量，进行实证分析和因果检验。

According to China Statistical Yearbook in 2020, the total population has increased from 540 million in 1949 to 830 million in 2019. The rapid population expansion did not match with the economic growth rate and resource carrying capacity at that time. In order to curb the trend of rapid population growth, the One-child Policy is introduced in 1979, which has been widely implemented in urban areas for more than 30 years. According to the 1% population sample survey in 2015, the average value of adjusted fertility rate from 2010 to 2015 was only 1.18, which is lower than the natural replacement level (2.1) that maintains the population constant. With the birth rate falling and the death rate rising, the natural population growth rate has fallen sharply. China's population may face negative growth. The decline in the fertility rate, on the one hand, suppresses economic growth by squeezing family resources through aging; on the other hand, promotes economic growth by gathering all family resources to a few children and increasing the human capital accumulation. The influence direction of the children’s quantity on the human capital accumulation is the key to determining economic development. Under the background of the family planning policy restricting the number of children born, the families cannot choose the quantity and quality of children. However, the implementation of the family planning policy varies among different occupations, between urban and rural areas, meaning families with different economic backgrounds are subject to the varying severity of family planning policy interventions. Based on this, the paper takes China as a case study, incorporating all families from different income class into the analysis framework, relaxes family resource constraints and the assumption that parents treat children of different genders equally to analyze the impact of family income class on quantity-quality trade-off and gender differences in substitution effects. From "both only-child couples can have two children" to "separate only-child couples can have two children" and then to “all couples can have two children”, in the context of the continuous relaxation of the family planning policy, it is of great practical significance to examine the impact of the quantity of children on individual quality.

Becker, the American economist, put forward the quantity-quality trade-off in the 1960s, without considering the boy preference. There are two reasons: first, in the developed North American countries, the total fertility rate at that time was about 3.7, which was relatively high; second, the developed countries prefer the diversity of children's gender to a certain gender. However, as a country deeply immersed by Confucian culture and with a well-developed agricultural civilization, males have comparative advantages in inheriting the family name, raising children to provide against old age and participating in market labor. As a result, Chinese nation’s boy preference is deeply rooted. In the context of the family planning policy prompting the total fertility rate to the lowest point, families with boy preference use fetal sex identification technique to satisfy the willingness to have at least one son, so that the gender structure of children within the family is inclined to boys. The macro is the sum of the micro. All individuals are summed up by gender and age, subsequently sex ratio imbalance at birth at the regional level. According to the Sixth Population Census, we calculated the sex ratio of children between 0 and 4 years old. It is as high as 1.20 (while the normal range of sex ratio at birth is 1.02-1.07 defined by United Nations), meaning that nearly 16.7% of young men cannot match partners in the future marriage market, which will lead to a series of serious social crimes and ethical problems. Gender structure is a macro concept at the regional level. Most of the existing literature takes regions as the research object, discussing the related issue from a macro perspective, which is impossible to measure boy preference, and difficult to explore the micro-motivation buried deeply. In the context of the distorted gender structure, this paper attempts to explore the cultural and economic causes for children’s gender selection from the micro-family perspective. Based on this, taking young individuals with unbalanced sex ratio as the research object, this paper further analyzes the causal relationship between sex ratio imbalance and gender role attitude, forming a logical chain of "parental gender role attitude → gender structure of the offspring within the family → offspring sex ratio imbalance at the regional level → offspring individual gender attitude”. These studies have enriched the literature on the causes and consequences of sex ratio imbalance and have certain theoretical value.

As a rational economic entity, every family chooses the children’s quantity, quality and gender to optimize the fertility decision-making for the goal of maximizing utility. First of all, the question answered in this paper is that when families cannot independently choose the children’s quantity because of family planning policy, does the quantity-quality trade-off exist? Relaxing family resource constraints and borrowing constraints, how does family economic background affect quantity-quality substitution effect? Relaxing the assumption that the children of different genders are treated equally, is there gender difference in substitution effect? Based on the China Educational Panel Survey (CEPS in 2013 and 2014), regarding cognitive ability scores, interest tutoring class participation and expenditure as quality indicators, the paper uses OLS and MLE estimation methods to explore the above issues. The results show that: (1) The quantity-quality trade off holds in different income classes. That is, the size of siblings has a significantly negative effect on individual’s cognitive scores as well as the tutoring class participation and expenditure. The middle-income class bears the largest trade-off effect, but those families from low- or high-income classes bear less. (2) Although the quality declines faster with the growing number of siblings for those middle-income families, individual’s cognitive scores, as well as the tutoring class participation and expenditure, have a positive correlation with income class after controlling for personal, family, class and school characteristics. (3) In terms of gender heterogeneity, an increase in the quantity of siblings has a greater negative impact on girls. (4) In terms of the severity of family planning policy, the children from families subjected to a severer One-child Policy intervention suffer a greater negative impact due to an increase in the quantity of siblings. (5) The causality is identified with the 2SLS-IV estimation, using the average number of children raised by every family at the county (or district) and grade levels as the instrumental variable of children quantity. This paper also carries out the robustness check using parental occupations as the income class indicator. All conclusions are similar to the benchmark regression.

In addition, when the family planning policy restricts the number of children born, families with boy preference make gender selection to meet the desire to have a son, resulting in sex ratio imbalance at birth. As the country with the most serious sex ratio imbalance in the world, China has attracted the attention of scholars at home and abroad. However, most of the existing literature discusses the related issue at the regional level. It is difficult to explore the micro-motivation buried deeply. Meanwhile, it is impossible to measure the boy preference at the regional level. Based on the data of Chinese Women's Social Status Survey in 2000 and 2010, this paper regards micro-family as the research object，uses Tobit and Logit measurement methods, to analyze the cultural and economic reasons for the family's gender selection of children. The results show that: (1) The influence of cultural concept on children’s quantity-gender structure is mainly manifested in two aspects of surname tradition and property inheritance. If parents hold the traditional cultural concept, children’s quantity-gender structure favors boys, meaning the boy(s) share rise, the last child for families with one more children is more likely to be a son, one-child families have an increased odds ratio of boy to girl. (2) The influence of economic factor on children’s quantity-gender structure is mainly manifested in the aspect of "whether parents have pension". Families with pension insurance have a pro-daughter gender structure by reducing the dependence of "raising children to provide against old age". However, the effect of "whether parents have medical insurance" on the children’s quantity-gender structure is not significant. (3) According to wives’ birth years, we divide the two sub-samples of "Not affected by family planning policy" and "Affected by family planning policy" to recalculate the model and find that the coefficients of cultural and economic factor are significant at a higher statistical level with larger absolute values for the sub-sample affected by the family planning policy. That is to say, the sex ratio imbalance of the offspring is more serious for the subsample affected by the family planning policy.

Finally, the gender structure leaning toward boys within families constitutes sex ratio imbalance of offspring at the regional level, which affects the gender role attitudes of young individuals. This paper uses OLS and 2SLS estimation method to verify the relationship between sex ratio of younger generation, gender role attitude of older generation and that of young individuals based on the micro-data of Chinese General Social Survey in 2010, 2012, 2013, 2015 and the sex ratio data reported by the Sixth Population Census. We distinguish older and younger generations according to the interviewee’s birth year, and divides gender role attitude in two dimensions of market labor and housework sharing according to related questions. The results show that: (1) Parents’ gender role attitude affects children’s gender structure. The stronger the parent’s traditional gender role attitude especially in market labor dimension, the higher the desired and actual son’s proportion, which will contribute to the sex ratio imbalance of younger generation. (2) As far as the relationship between sex ratio and gender role attitude for younger generation, there are two opposite transmission mechanisms. The sex ratio imbalance makes gender role attitude more traditional by strengthening the masculine culture, and weakens the traditional gender inequality attitude by enhancing the female bargaining power in marriage and family. Regardless of the gender attitude in dimensions of market labor or housework sharing, the coefficients of the sex ratio are positive, indicating that there is a positive correlation. That is, the higher the sex ratio, the more traditional the individual gender role attitude, indicating that the masculine culture plays an important role. In term of the impacts of sex ratio on gender role attitude in two dimensions, we found that the coefficient of sex ratio is greater and more significant in market labor dimension, but small and insignificant or has a low significance in housework sharing dimension. Because for the latter, the negative influence of female bargaining power in marriage and family offsets the positive influence of masculine culture to some extent. (3) There exists an intergenerational transmission in gender role attitude, especially for attitude in the same dimension. The correlation of gender role attitudes in different dimensions from older generation to younger individuals is less relevant. (4) Based on the 20% sample size of the 1% population sample survey in 2005, this paper calculates the proportion of only-child families by province and by age group as an instrumental variable of sex ratio. Then, using 2SLS estimation method to recalculate the model, the conclusion is consistent.

On the base of reviewing relevant literature and theories, this paper conducts empirical research, and proposes policy recommendations to promote the recovery of fertility rate, the decline of sex ratio, and the accumulation of human capital from four aspects: reducing the cost of family childbirth and raising children, weakening boy preference, improving education level, and tapping the demographic dividend.

The innovations of this article are mainly manifested in the following aspects. Firstly, this paper incorporates the children’s quantity, quality and gender selection into the logical framework of fertility decision-making, from the micro-family perspective. Secondly, relaxing related assumptions and expanding the quantity-quality trade-off theory. Thirdly, conducting empirical analysis and causality testing, by analyzing the fourth, fifth, sixth census data to calculate sex ratio with more accurate sixth census and identifying the instrumental variables of sex ratio.