Date of Award


Degree Name

Doctor of Philosophy



First Advisor

Wei-Chiao Huang, Ph.D.

Second Advisor

Jean Kimmel, Ph.D.

Third Advisor

Kevin H. Lee, Ph.D.


China is currently facing the issue of low fertility rates. In three chapters, I discuss the reasons for the low fertility rate in China from three different microeconomics perspectives, by using econometric methods including machine learning tools.

Chapter One considers the effects of income on fertility rate. There is a widely accepted view that the improvement of the education level has led to a decline in fertility. I critique this view and propose a new perspective: the childbirth choice of a family member is not only related to the education level, but also to the differences in education level between family members. Differences in education level among family members lead to a clear family division of labor, resulting in increased family income and fertility rates, thereby improving overall family welfare. My empirical analysis finds that differences in education level among family members do indeed have a significant positive impact on family fertility rates. Besides, such a positive impact is related to the gender differences in families and the overall difference in education levels between families.

Chapter Two considers the effects of income on fertility rate. The correlation between income and fertility rate has always been controversial. Most research shows that income and fertility rates are simply negatively related, while others indicate a more complicated relationship. I first try to identify two different types of income, each affecting fertility rates differently. I also consider the income from assets like real estate, and the other variables related to income such as “Hukou” and job type. My finding is that the negative correlation between income and fertility rate indeed exists, but it is only in the sample of people who are engaged in agriculture jobs. There is no statistical correlation between income and fertility rate for those engaged in other jobs. Compared to income, there is a very clear positive correlation between house size and the number of children.

Chapter Three considers the effects of social activities on fertility rate. I discuss the distinction between social and non-social activities and their potential impact on fertility rates. An individual’s time opportunity costs are limited, and they must be allocated between social and non-social activities. The growth of non-social activities in modern society, driven by new entertainment products and lifestyles, has reduced the time for social activities, leading to a decline in fertility rates. In the empirical section, I attempt to use machine learning methods to help identify effective variables from a range of factors, specifically including two models using shrinkage methods like Ridge and Lasso. I find that certain activities, such as using the Internet, have a negative impact on fertility rates, especially some activities like having a cell phone or surfing the internet have a significant negative impact on the number of children.

Access Setting

Dissertation-Open Access

Included in

Economics Commons