Democracy-Analysis-South-America-Africa

The data analysis of the situation of democracy in South America and Africa

This project aims to analyse and generate insight into the general situation of democracy in the South American and African countries. To achieve this, it uses the Python libraries of pandas, numpy, matplotlib and the Tableau for further visualizations. After having done with the coding part, the backstory of the current democracy scores of each country were explained according to their own political and historical events. Here are the specifics questions defined for the project:

• What are the average democracy scores of South American and African countries during COVID-19 pandemic?
• What is the trend of change in democracy index during this period?

• How can we compare all of countries in general (i.e., create a general map to show the democracy score of each country)
• Which countries are remarkable in the sense of having full democracies and authoritarian regimes?

The dataset was taken from is based on the index of the Economist Intelligence Unit, EIU. The general structure of the dataset is as follows:

• Entity: The name of the countries
• Code: The standard ISO code name of the countries
• Year: Respective years
• Democracy score: The score showing the condition of democracy in each country, ranging from 0-10.

Below are the general steps taken for the data analysis. A few example codes are given to have an idea of what has been done. To look at the full codes, you can refer to the notebook.

1. Importing the necessary packages:

import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
%matplotlib inline

2. First quick glance at the dataset:

df.head()
df.info()
df.value_counts()

3. Since the focus is on South American and African countries only, making a list for each of them and also dropping the "Code" column since Tableau can generate an automatic map using only the name of the countries:

south_american_countries = [
    "Argentina", "Bolivia", "Brazil", "Chile", "Colombia", "Ecuador",
    "Guyana", "Paraguay", "Peru", "Suriname", "Uruguay", "Venezuela"
]
df_filtered = df[df['Entity'].isin(south_american_countries)]
.
.
.
df_filtered.drop(columns=['Code'], inplace=True)
df_filtered.head()

4. Filtering the years to see if there is a trend of decline during COVID-19 pandemic:

year_filtered_df = df_filtered[(df_filtered['Year'] >= 2019) & (df_filtered['Year'] <= 2023)]
year_filtered_df.head(20)

5. Looking at the democracy change throughout the years:

year_filtered_df['Democracy change'] = year_filtered_df.groupby('Entity')['Democracy score'].diff()
.
.
.
average_change = year_filtered_df.groupby('Entity')['Democracy change'].mean().round(2)

6. Making the exact same thing for the African countries except that for the sake of simplicity of preliminary exploration, I decided to focus only on countries having suffered from a coup or military junta in the last five years, leaving the full comparison of South American and African countries for Tableau.

african_countries = ['Mali', 'Burkina Faso', 'Niger', 'Chad', 'Guiena', 'Gabon', 'Sudan', 'Democratic Republic of Congo', 'Central African Republic']
.
.
.

4. Gaining insights:

   1. The condition of democracy in South American countries in terms of score (table)

   2. The condition of democracy in selected African countries in terms of score (table)

   3. The trend of these during the COVID-19 pandemic (line plot)

   .
   .
   .
   bottom_updated_pivot = bottom_updated_df.pivot(index='Year', columns='Entity', values = 'Democracy score')
   for country in bottom_updated_pivot.columns:
   plt.plot(bottom_updated_pivot.index, bottom_updated_pivot[country], marker='o', label=country)
   
   # Add title and labels
   plt.title("Democracy Scores of Countries Over Time", fontsize=14)
   plt.xlabel("Year", fontsize=12)
   plt.ylabel("Democracy Score", fontsize=12)
   plt.xticks(bottom_updated_pivot.index, rotation=45)
   plt.legend(title="Country")
   plt.grid(True, linestyle='--', alpha=0.6)
   
   # Show the plot
   plt.tight_layout()
   plt.show()

5. Further visualizations with Tableau

1. How can we compare all of countries in general (i.e., create a general map to show the democracy score of each country)

2. Which countries are remarkable in the sense of having full democracies and authoritarian regimes?

For the detailed expression, please refer to this blog post or Tableau dashboard.