Introduction

The educational advantage is the primary channel to increase human capital. At the micro level, does the increase in education level increase the population’s income to the same extent?

This paper selects data on the education index as well as the income index for six provinces or municipalities directly under the central government of China, namely Beijing, Guangzhou, Shanghai, Jiangsu, Yunnan, and Gansu, in 2019, and observe the years of schooling children aged six and the mean year’s schooling population aged 25+ for residents of these cities.

Setup

import numpy as np import pandas as pd import matplotlib.pyplot as plt import seaborn as sns import plotly.express as px import plotly.graph_objects as go !pip install statsmodels import statsmodels.api as sm import statsmodels.formula.api as smf import geopandas as gpd !pip install stargazer from stargazer.stargazer import Stargazer, LineLocation

Import data

df = pd.read_csv('/work/dataset.csv') df

df.info()

df_definition = pd.read_csv('/work/SHDI-SGDI-5.0-Vardescription.csv') df_definition

Prepare data

df1 = df[df.level != 'National'] df1

df.columns

df["year"]= df["year"].astype(str)

df2019 = df[df['year'] == '2019']

df2019['continent'].unique()

df2019['country'].unique()

df2019['region'].unique()

Descriptive statistics

df1.loc[:, df1.columns != 'year'].describe().round(2)

df2019.describe().round(2)

df2019.query("country == 'China'").describe().round(2)

df_2019 = df1[['country','year','region','sgdi','shdi','shdif','shdim', 'healthindex','healthindexf','healthindexm','incindex', 'incindexf','incindexm','edindex','edindexf','edindexm', 'esch','eschf','eschm','msch','mschf','mschm','lifexp', 'lifexpf','lifexpm','gnic','gnicf','gnicm','lgnic','lgnicf', 'lgnicm','pop','mfsel']].query("year == 2019") df_2019

df_cn = df2019[['country','year','region','sgdi','shdi','shdif','shdim', 'healthindex','healthindexf','healthindexm','incindex', 'incindexf','incindexm','edindex','edindexf','edindexm', 'esch','eschf','eschm','msch','mschf','mschm','lifexp', 'lifexpf','lifexpm','gnic','gnicf','gnicm','lgnic','lgnicf', 'lgnicm','pop','mfsel']].query("country == 'China'") df_cn

df_cn[['country','year','region','incindex','edindex','msch','esch']].query("region==['Beijing','Shanghai','Guangdong','Jiangsu','Yunnan','Gansu']")

Exploratory data analysis

px.scatter(df_cn, x ='edindex', y ='incindex', log_x = False, log_y = False, hover_data= ['region','msch','esch'], trendline = "ols", trendline_options=dict(log_x = False, log_y = False), marginal_x = "box", marginal_y = "box", )

Regression analysis

smf.ols(formula='incindex ~ edindex', data=df_cn).fit().summary()

mod1 = smf.ols(formula='incindex ~ edindex', data = df_cn).fit() Stargazer([mod1])

Conclusion

The inequality of income distribution due to education in China is still relatively high in terms of unfair and unreasonable components.

We should focus on enhancing the inequality of educational opportunities.

We should promote the equal distribution of primary education and strengthen the policy support for compulsory education in poor areas.