Overview

The entire paper was written in R Markdown, which was then converted into
LaTeX for finer formatting. An R Markdown file is a plain text file that
embeds R code into Markdown text that can be directly compiled (i.e., “knitted”)
into reports with publication quality tables and figures in Word, PDF, or HTML
format.

Among the five tables and four figures presented in the main text, all tables
and one figure (Figure 1), which were based on the China Family Panel Studies
(CFPS) data, can be readily replicated. Unfortunately, Figures 2-4 cannot be
replicated because they require data source that is not publicly available.

Data

The 2010 wave of the CFPS data can be requested from the project’s official web
site (http://www.isss.edu.cn/cfps/EN/). Unfortunately, the 1982 one-per-thousand
fertility survey, which were carried out by the Chinese government some 30
years, is not publicly available.

Program Files

A total number of six program files of two different types were provided,
including:

1.  read_CFPS.do. This is the Stata do file that reads in the official 2010
    CFPS data (CFPS_2010_adult.dta), does some preliminary cleaning and
    recoding, and produces the final data set (cfpsdata_fm.dta) that is used
    in all the other program files.

2.  Tables.Rmd. This is the R Markdown source file that produces all five
    tables reported in the manuscript.

3.  Fig1.Rmd. This is the R Markdown source file that produces Figure 1 in
    the manuscript.

4.  Fig2. Rmd. This is the R Markdown source file that produces Figure 2 in
    the manuscript. Unfortunately, it requires the 1982 one-per-thousand
    fertility survey, which cannot be made publicly available.

5.  Fig3.Rmd. This is the R Markdown source file that produces Figure 3 in
    the manuscript. Unfortunately, it requires the 1982 one-per-thousand
    fertility survey, which cannot be made publicly available.

6.  Fig4.Rmd. This is the R Markdown source file that produces Figure 4 in
    the manuscript. Unfortunately, it requires the 1982 one-per-thousand
    fertility survey, which cannot be made publicly available.