Child abandonment in China, particularly since the late 1970s, has been under-researched due to a lack of readily available data. Previous studies have relied on limited datasets, hindering a comprehensive understanding of the interplay between socioeconomic factors, cultural norms, and government policies in shaping abandonment behaviors. This study leverages the extensive, self-reported data from the Baby Come Back Home (BCBH) website, a significant online platform for reuniting lost or abandoned children with their families, to overcome these limitations. The BCBH data reveal a wave of child abandonment from the late 1970s to the early 2000s, peaking around 1990 and declining thereafter. Girls constituted a disproportionately large percentage of abandoned children. This phenomenon is linked to China's patriarchal culture and the implementation of the OCP. Economists view the OCP as an implicit consumption tax on additional births, leading to potential consequences like reduced birth rates and increased abandonment. The impact of stricter OCP enforcement on gender-selective abandonment is complex, potentially influenced by the concurrent increase in gender-selective abortions. The study investigates this interaction, considering the moderating role of the deeply rooted son-preference culture, examining how the interplay between stringent OCP implementation and cultural preferences affects abandonment rates differently for boys and girls. The study uses a provincial-level analysis to account for variations in policy stringency and cultural strength. The OCP's implementation varied across provinces, offering a natural experiment to assess its effects. The period after 2012, when the OCP was gradually relaxed, serves as a crucial comparison point.

Existing literature on child abandonment in China is limited by data scarcity. Early studies used small-scale surveys or indirect estimations, providing incomplete insights. These studies highlighted the prevalence of female infanticide and abandonment, linking it to the OCP and deeply ingrained son preference. However, they lacked the scale and granularity necessary to fully grasp the spatial and temporal dynamics of this phenomenon, and they often overlooked the impact on boys. This research aims to bridge this gap by leveraging a unique dataset that offers a more comprehensive picture. Studies examining informal adoption and child trafficking have been more prominent, but they rarely focus on the direct act of abandonment itself. This study aims to directly address this gap by using a large-scale dataset of self-reported cases to provide a comprehensive analysis of abandonment trends. The authors also reference the work of Ebenstein (2010), which highlights the "missing girls" phenomenon in China as a consequence of the one-child policy and son preference. This research builds upon these existing studies by providing a more nuanced understanding of how policy and cultural factors interact to shape abandonment behaviors.

This study employs a novel data collection method, web crawling, to extract information from the BCBH website, the largest online platform in China dedicated to reuniting families. Text mining techniques were used to process the collected data, specifically focusing on cases of child abandonment. Over 107,000 cases spanning 40 years were analyzed. The data were cleaned to exclude cases involving kidnapping, stealing, or trafficking. Cases where the child was over 12 years old were also removed to focus on abandonment rather than runaway cases. Provincial-level data on the stringency of the OCP were obtained from official documents, including birth quotas and penalty rates. The OCP's stringency was calculated as a composite indicator (penalty rate/birth quota). Two measures were used to quantify son-preference culture: the number of genealogies per capita and the proportion of the Han ethnic population. Socioeconomic variables, such as urbanization rates, per capita income, and illiteracy rates, were also incorporated. The study employs a multiple regression model at the provincial level to analyze the relationship between OCP stringency, son-preference culture, and child abandonment. The model included province-level fixed effects, time-fixed effects, and control variables to address heterogeneity and potential confounding factors. The dependent variable was the density of abandoned children (cases weighted by provincial population). Separate regressions were conducted for boys and girls, allowing for a gender-differentiated analysis. The analysis also involved grouping provinces based on OCP strictness and clan culture strength to assess regional variations. The period before and after 2012 (when OCP relaxation began) was compared to assess shifts in abandonment patterns. Robustness checks were conducted using alternative measures of son-preference culture and by removing cases involving children with disabilities. Additionally, the analysis was replicated using the two separate datasets within the BCBH platform (parents searching for children, children searching for parents).

The study's key findings show a significant positive association between stricter OCP implementation and the overall number of child abandonments. This supports Hypothesis 1. The impact of the OCP on abandonment was more pronounced for boys than girls, a finding that contrasts with the commonly held belief that the policy mainly affects girls. The interaction between OCP stringency and son-preference culture was significant in influencing abandonment. In provinces with strong son-preference cultures, the OCP's negative impact on girl abandonment was weaker. However, this doesn't imply that the OCP or son-preference culture reduces sex discrimination against girls. This result is interpreted as a shift from postnatal sex discrimination (abandonment) to prenatal sex selection (abortion), driven by the increased availability and affordability of ultrasound technology for prenatal sex determination. This contrasts with findings from previous studies, which mainly focused on the impact on girls. The study demonstrated that in provinces with strong son-preference culture, the negative impact of the OCP on girl abandonment is weaker. This suggests a substitution effect: when OCP is strict, parents may choose to abort female fetuses rather than abandon girls. The analysis of the period after 2012, when the OCP was relaxed, revealed that the effects of the OCP and its interaction with culture on child abandonment weakened significantly. This suggests the impact of the policy changes. The robustness checks confirmed the main findings, even when using alternative measures of son-preference or removing cases of children with disabilities. The analysis of separate datasets from the BCBH website yielded consistent results.

The findings challenge the prevailing view that the OCP primarily impacts girls. The results suggest that the OCP's impact on abandonment is more complex and depends on interactions with cultural norms. The observed shift from postnatal to prenatal sex discrimination highlights the limitations of focusing solely on abandonment as an indicator of gender inequality. The study provides evidence that the OCP, despite potentially increasing the demand for sons, may simultaneously have inadvertently increased the acceptance of daughters, particularly in regions with more lenient interpretations of the policy, leading to a lesser need for abandonment of girls. This is evidenced by the weaker relationship between stricter OCP policies and the abandonment of girls in provinces with stronger son-preference cultures, where families might opt for gender-selective abortion as an alternative. This underscores the importance of considering the broader context of social, cultural, and policy factors when examining the complex issue of child abandonment.

This study makes several key contributions. It offers a large-scale analysis of child abandonment in China, using a novel data source. It reveals that the OCP's impact on abandonment is more complex than previously understood, impacting both boys and girls differently. The findings highlight the importance of considering the interplay between policy, culture, and technological advancements in shaping gender inequality. Future research could explore the potential substitution effect between gender-selective abortion and abandonment in more depth, using more comprehensive data on abortion rates. Furthermore, exploring the long-term consequences of abandonment on the children involved and the evolving cultural norms around gender preference would also be valuable.

The study relies on self-reported data from the BCBH website, which may suffer from underreporting bias, particularly for abandoned girls. The measure of OCP stringency captures only the monetary penalties and birth quotas, neglecting other potential penalties (e.g., job loss). The study does not directly account for the impact of other socioeconomic changes in China, such as urbanization or marketization, which might have also influenced abandonment rates. Finally, the study does not consider the health status of the abandoned children, which might influence abandonment decisions.