Adolescent romantic involvement is among the primary concerns of parents and school authorities in China. Adolescent romance is regarded as being related to various negative outcomes, especially poor academic performance. As a result, Chinese parents and teachers frequently interfere in adolescent romantic relationships. However, few studies have examined the incidence, nature, and social processes of adolescent romantic relationships in China because of a lack of suitable data. This study provides the first national estimates of the prevalence and contextual determinants of adolescent heterosexual romantic involvement in China by using data from two nationally representative surveys, namely the China Education Panel Survey (CEPS) and the China Family Panel Studies (CFPS).

This study adopts two theoretical perspectives—social bonding theory and social learning theory—to examine adolescent romantic relationships in the context of their social appropriateness. According to social bonding theory (Hirschi, 1971), an adolescent’s bonds with family, peers, and school authorities are key sources of informal social control that discourage involvement in romantic relationships. Specifically, parental guidance and supervision are conducive to the internalization of conventional values and norms about adolescent romance. Because acceptance by peers is crucial, adolescents may modify their behavior to match that of their peers. Therefore, the opposition of peers may dissuade adolescents from romantic involvement. With respect to informal social control from school, a strict school climate may be negatively associated with students’ romantic involvement. We therefore expect that the disapproval (approval) of family members and peers decreases (increases) an adolescent’s likelihood of romantic involvement. We also predict that a stricter school climate is associated with a lower likelihood of adolescent romantic involvement.

The effectiveness of social bonds relies on the assumption that the attitudes and behaviors of family, peers, and school authorities reinforce conventional values and norms. In reality, the attitudes of peers toward dating during adolescence and their own romantic behaviors vary widely. However, the conservative values of parents and school authorities toward adolescent romance are undoubtable. Therefore, social bonding theory underestimates bonding to figures who reinforce unconventional values and norms. Therefore, social learning theory must be incorporated to account for variations in the behaviors modeled by peers. Social learning theory predicts that adolescents imitate behaviors modeled by those to whom they are bonded (Elliott et al., 1985). Thus, we expect that peers’ involvement in romantic relationships increases the likelihood of an adolescent’s involvement in romantic relationships.

This study primarily analyzes data from the CEPS, an ongoing large-scale and nationally representative longitudinal survey. Two waves of CEPS data are currently publicly available: The baseline data and the first set of follow-up data. The baseline survey, which was launched in the 2013 to 2014 academic year, interviewed students in grades 7 and 9. The first follow-up survey, which was conducted in the 2014 to 2015 academic year, tracked the original cohort of seventh graders in the baseline survey. This study only analyzed data from the participants in the first follow-up survey (i.e., the students in grade 8) because variables measuring adolescent romance were only examined during that wave.

Although the first follow-up survey of the CEPS provides valuable information on adolescent romance, its focus on 8th graders limits the generalizability of its findings in terms of the entire adolescent population. Moreover, its cross-sectional nature prevents us from establishing rigorous causal conclusions. We therefore supplement the analysis by employing data from the CFPS, an almost nationally representative family panel survey, with interviews administered biennially since 2010. Five follow-up surveys have been completed at the time of this study (in 2012, 2014, 2016, 2018, and 2020). During the first three waves of the study, questions concerning romantic involvement were posed only for adolescents aged 10 to 15 years; from 2016 onwards, questions concerning romantic involvement were extended to cover the unmarried population aged 10 to 40 years. The World Health Organization defines an adolescent as an individual aged between 10 and 19 years. Because sample weights are not available in the 2020 wave, we use data from the CFPS 2016 and 2018 to compute the prevalence of romantic involvement in the entire adolescent population. To assess the robustness of the effect of peers on adolescent romantic involvement as estimated using the CEPS data, we employ the longitudinal nature of the CFPS. To implement a robustness check, we restrict our analysis to a sample of adolescents aged from 10 to 15, used the 2010 wave of the CFPS as our baseline population, and track the adolescents’ romantic involvement in 2016. These restrictions are required because (1) the CFPS only collected information on adolescent romance in the 2010 baseline survey, and (2) the 2016 wave of CFPS covered a wider population when collecting information on romantic involvement, and a small proportion of the original adolescents aged 10 to 15 in the baseline survey were found to be married in the 2016 wave.

In the CEPS, measures of romantic involvement can be classified into three categories: (1) having romantic feelings for a schoolmate or friend of the opposite sex; (2) having ever been in a relationship; and (3) having engaged in intimate behavior with someone of the opposite sex (e.g., handholding, kissing, and other intimate physical contact). The measurement of romantic involvement in the CFPS is relatively simple. During the first three waves of the study, respondents were asked, “Do you have a boyfriend or girlfriend?”; in the latter three waves, respondents were asked, “Have you had a boyfriend or girlfriend in the past 2 years?”

We consider two sets of contextual determinants of romantic involvement: one relates to social bonding theory, and the other relates to social learning theory. In the CEPS, both sets of contextual determinants are available. Social bonding theory comprises the determinants of (1) parental monitoring of adolescent friendships, (2) the attitudes of teachers and elder family members toward romantic relationships among junior high students, (3) the overall attitude of peers toward romantic relationships among junior high students, and (4) the school climate. The determinants of social learning theory include the relationship status of an adolescent’s five closest friends. With respect to contextual determinants, the CFPS only collected information on the relationship status of the closest male or female friend in the 2010 baseline wave.

This study analyzes data and tests the research hypothesis by employing descriptive statistics, the Pearson correlation coefficient, and binary logistic regression.

The descriptive results reveal that, in 2018, approximately 8% of Chinese adolescents had engaged in a heterosexual romantic relationship during the preceding 2 years. Although intimate physical contact accompanied romantic relationships among Chinese adolescents, it was generally limited to handholding; kissing and other intimate physical contact were rare.

The results of multivariate analyses indicate that the supervision of parents and school authorities have little influence on the initiation of adolescent romance. Parental monitoring and the disapproval of teachers and elder family members cannot prevent adolescents from initiating a romantic relationship or from engaging in intimate behavior with someone of the opposite sex. Moreover, school climate is not associated with adolescent romantic involvement. Nevertheless, the attitudes and behaviors of peers crucially influence adolescent romance in China. Specifically, the objections (envy) of peers to adolescent romance decreases (increases) the likelihood of engaging in romantic relationships. Having close friends who are in a romantic relationship increases the likelihood of an adolescent becoming romantically involved.