Page 10 - IJPS-4-2

Child trafficking in China

of them were found between May 2016 and May 2018. Among all the solved cases, the top four categories were running
away from home (55.9%, 1,705 cases), getting lost (13.9%, 424 cases), accidental death (4.6%, 140 cases), and being
trafficked (1.6%, 48 cases) (Mo, 2018). Roughly <30% of total reported cases, 250 per year were child trafficking even if
we assume the worst case scenario that all the rest were child trafficking cases.
Based on the number of cases uploaded to the websites of CJO and Open Law and the official gazettes of the SPC, the
percentage of the uploaded among sentenced cases for crimes of infringing on the rights of the person and the democratic
rights of citizens including murder, homicide, rape, and human trafficking was about 60%. If there is no substantial
discrepancy on the pattern of selective uploading across provinces, the number of cases that we have was about one-third
to half of total number of cases.
In contrast to the international definition of trafficking that focuses on the purpose of exploitation, child trafficking in
the Chinese Judicial System emphasizes the illegal trafficking process and the purpose of profit making. Therefore, we
acknowledge that a direct comparison of findings between the international child trafficking studies and those using data
from the Chinese Judicial System might not be suitable. Furthermore, we acknowledge that sentencing data acquired from
the CJO may be incomplete due to the interpretations of the SPC (2013; 2016), which states that all sentencing documents
are required to be uploaded, except for cases with classified information, individual privacy concerns, defendants under
18 years old, those closed for mediation, and other “unpublishable” situations, thus leading to suspicions that the uploaded
sentencing documents might be subjective to selection or purposeful deletion (Liebman, Roberts, Stern et al., 2017).
Unfortunately, there are no available sources (e.g. official gazettes, news reports, and research articles) that could be used
to cross-reference and evaluate the pattern and extent of uploading or deleting case information so far. In addition to the
possible selective uploading or deleting of documents, it should be cautioned that not all incidents of trafficking have been
identified by the police, and of those offenders who have been caught, not all have gone through the legal system. Still,
despite the limited access to child trafficking data, the information garnered from the CJO sentencing documents may add
some much-needed insight into the current trafficking issues in China.
2.2. Measurements

As a structured document, each of the sentencing reports has three distinct parts: A caption, main body, and conclusion.
The caption contains the parties, the court, and the case number. The main body gives a description of acknowledged facts
by the court, for example, information about the offender(s) and the victim(s), characteristics of the offense, such as time
and location, and court processes (e.g. legal representation and appeal). The dispositional decisions, such as articles cited
and sentences, are listed in the conclusion. We manually coded all the information on offenders, victims, offenses, and
victim-offender relationships. For example, gender of victim (male), age of victim in years, whether the victim is sold by
guardian(s), final price paid by the buyer in thousand Chinese yuan, whether the case is reported to police, whether the
purpose of trafficking is adoption, the number of offenders involved, and the proportion of male offenders involved, were
coded.
For each of the cases, geographic information on both the origin and destination of trafficking routes was also recorded
at both the provincial and city levels, which were used to match local socioeconomic indices, such as the urban–rural
average income ratio (URAIR), the SRB, the rate of illiteracy, and the level of legal adoption. For provincial socioeconomic
characteristics, the URAIR and the percentage of rural illiteracy were chosen as the measures of socioeconomic
development. The ratio of the summed number of adoptions over 2013-2016 versus the average total population was used
to access the level of legal adoption, and the SRB was used as the level of son preference.
Except for the level of adoption that were from the Statistical Yearbook of the China Civil Affairs (Ministry of Civil
Affairs, 2013; 2014; 2015; 2016), all other measures were from the latest census. Regional hotspots, as identified by the
heat maps, were also controlled using dummy variables for regions: Central East (Shandong, Henan, and Hebei Provinces),
East (Jiangsu and Zhejiang Provinces), Southeast (Guangdong and Fujian Provinces), and Southwest (Sichuan, Guizhou,
and Yunan Provinces).

2.3. Analytical methods
We provided regular descriptive analyses to explore the link between socioeconomic factors and characteristics of child
trafficking, such as the gender of the victim. To examine the geographic pattern of child trafficking, heat maps were
generated to see the areas where trafficking origins and destinations were mostly concentrated, and trafficking networks
were plotted using R packages such as recharts (Xie, 2017) and igraph (Csardi and Nepusz, 2006). We also estimated

4 International Journal of Population Studies | 2018, Volume 4, Issue 2