Abstract: The relationship between marriage and happiness has always been concerned by the public as well as researchers. Some studies found that marriage could decrease individual happiness, the others revealed that marriage could increase individual happiness. So no clear conclusion has been reached at present. Previous studies mostly used cross-section data, so this study aimed at using the tracking data of China Family Panel Studies (CFPS) to explore the influence of individual marital status change on happiness. To control the influence of other confounding factors, the methods of Differences-in-Differences (DID) and Propensity Score Matchings (PSM) were adopted in this study. In real life, whether an individual marries or not is not at random, because it may be affected by other covariable factors, such as income, education level, and region. The idea of PSM matching is to convert multiple covariables that need to be matched into an index—propensity score, namely the probability that an individual becomes an experimental group, which in this study is the probability that the marital status changes from unmarried to married. Then, the control group and the experimental group were matched according to different matching methods (such as one-to-one matching, nearest-neighbor matching, K-nearest neighbor matching, kernel-matching, radius matching, etc.). After that, the difference in happiness between the experimental group and the control group was compared by the method of DID, and the pure effect of the experimental treatment (from unmarried to married) was obtained, to reach a conclusion. This study adopted PSM-DID, which not only avoided self-selection bias by means of PSM, but also solved the problems of missing variables and time effect using DID. Besides, it could also answer more scientifically and reasonably whether marriage may lead to the improvement of happiness. The result showed that: (1) the marital status transition (from unmarried to married) could improve individual happiness significantly, which is moderated by gender. To be more specific, happiness increased more in men than women if changing from unmarried to married, of which the explanation mechanism may be related to the health and economic problems brought by marriage. Besides, intimate relationshihp could make people feel more supportive and help improve the physical and mental health, thus affecting individual happiness, which may be moderated by the quality of marriage. What’s more, marriage may also bring more economic benefits for married men than for unmarried men, such as higher wages, and this economic well-being may also contribute to higher happiness to some extent; (2) Increased happiness of men is significantly higher than that of women, the reason may be related to the traditional marriage culture of China and the division of labor between men and women in family structure, which may be the important reasons why change of happiness in marriage is significantly higher in men than in women. According to the traditional view of marriage, men are symbol of power and dominate the family, which could make them be more satisfied about present life after marriage. Therefore, based on the tracking samples in big data for the first time, this study answered the influence of marriage in enhancing happiness, as well as the moderating effect of gender in the relationship between marriage and happiness.

Key words: marital status, happiness, Propensity Score Matching, Differences-in-Differences, gender difference

摘要： 婚姻与幸福的关系一直备受学界和公众关注，但尚无明确定论。本研究采用中国家庭追踪调查（CFPS）两次追踪到的24172人数据，利用双重差分倾向得分匹配法（PSM-DID）探讨婚姻与幸福感因果机制。结果发现婚姻状态转变（未婚→已婚）能明显提升个体幸福感，但提升效应受性别调节，即婚姻对男性幸福感回报明显高于女性。结论：婚姻带来的社会支持，经济增益等可能是婚姻促进幸福感的主因，而传统婚姻文化及婚后男女在家庭结构或功能上的分工差异，可能是婚姻对男性幸福感的回报明显高于女性的重要原因。

关键词: 婚姻状况转变, 幸福感, 双重差分法, 倾向得分匹配法, 性别差异