Determinants Of Age At First Marriage For Women And International Comparisons

In 2007, Chinese Ministry of Education acknowledged the term "leftover women" as one of the new words in Chinese dictionary, arousing widespread concern in China. In recent years, the trend of delayed marriage for single women exists not only in China, but also many places around the international community.Marriage is not just about personal happiness and social harmony; it will also affect the population structure and national economic growth. From the perspective of an individual, late marriage brings reduced life satisfaction for women, causing personal psychological problems. From the perspective of the whole society, late marriage leads to infertility problems for women, thereby reducing the proportion of young adults in the labor force, and thus brings a series of problems of an aging society. Therefore, the problem of late marriage is worth studying.This paper has established econometric models to compare age at first marriage for women internationally. It attempts to explain the reasons of late marriage in an all-round quantitative sense. It adopts singulate mean age at first marriage (SMAM) as the dependent variable. In terms of independent variables, this paper includes the following dimensions like average life expectancy, female education attainment, economic and political empowerment, religion and etc.By studying the results of the previous researches on the age at first marriage for women, and by taking the accessibility of data into general consideration, all the independent variables used in this paper includes life expectancy at birth, female enrollment in tertiary education, GNI per capita, female labor force participation rate, economic participation and opportunity (EPO), political empowerment (PE) and religious beliefs. Among them, EPO and PE are the sub indices taken from the Global Gender Gap Index.Specifically, this study has conducted two regression analyses using the databank above-one panel data test and one cross-sectional data test. In the panel data analysis, dependent variable comes from SMAM for women in 218 countries collected in five periods, namely around 1970, around 1985, around 1995, around 2005, and latest; independent variables cover life expectancy at birth, female enrollment in tertiary education, GNI per capita, female labor force participation rate. The results show that there is sufficient evidence to prove that average life expectancy, female enrollment in tertiary education and GNI per capita have a significant positive effect on the delaying women’s marriage timing. These three variables combined can explain 66.16% of the variation in SMAM. This result is in line with expectations. The female labor force participation rate does not pass the significance test, which may be explained by the following reasons. First, it is likely that women’s performance in the labor market does not play an important role in deciding their marriage timing. Second, there may be correlations between female labor force participation rate and other independent variables. Third, the number of observation in this study is too small-only 237 observations for panel data analysis.In the cross-sectional regression based on the year 2010, the model adds other independent variables like EPO, PE and religion, in addition to three variables that passed the above panel data test. The regression results show that none of the newly added variables is significant. To explain, variables that are reflective of the gender gap-EPO and PE-may have multicollinearity problems with other variables (e.g. GNI per capita). Failure to pass the significance test may also be caused by too few observations, thus there is sufficient evidence to support the argument. Religion contains seven specific religious categories, each represented by a dummy variable. Contrary to expectations, none of the dummies passes the test. The reason may lie in the inappropriateness of data processing methods for the religion variable.Based on the findings and potential explanations for these findings, the paper acknowledges limitations of this study and proposes measures to improve this study accordingly. Finally, this paper puts forward some suggestions to accept and cope with the late marriage phenomenon from personal as well as social perspectives.