Election Voter turnout is a vital metric in evaluating the health of a democracy. Best Democracy stands on the people who have the power to vote and allows its people to leverage their power by casting their votes. Understanding who votes and who doesn’t can reveal significant insights into political engagement, national interests, social inequality, and the effectiveness of electoral systems and their functioning in that particular country using R Programming Language.

Objectives and Goals

The primary objective of this article is to visualize voter turnout data to identify trends and patterns over time and across different States in India. By doing so, we aim to understand the factors influencing voter turnout and make informed predictions about future electoral participation. India consists of a national election where each states are allocated a certain number of seats and the party which gets the majority in them gets into power in the center. In our dataset, we have the information on the voting details for each state.

Dataset Explanation

The provided dataset contains information about candidates in Indian national elections which particularly focuses on the voter turnout and other key details. Here are the relevant columns explained:

• st_name: The name of the state or union territory where the election took place.
• year: The year in which the election was held.
• pc_no: The parliamentary constituency number.
• pc_name: The name of the parliamentary constituency.
• pc_type: The type of parliamentary constituency (e.g., General, Reserved).
• cand_name: The name of the candidate who contested the election.
• cand_sex: The gender of the candidate.
• partyname: The full name of the political party to which the candidate belongs.
• partyabbre: The abbreviation of the political partys in india.
• totvotpoll: The total number of votes polled (i.e., the number of votes received by the candidate).
• electors: The total number of registered electors in the constituency.

Dataset Link: Election Voter

Load the Dataset and data preprocessing

First, let’s load the data from a CSV file and clean it by removing any undefined data, filling missing value and remove the outliers.

# Load necessary libraries
library(ggplot2)
library(dplyr)
library(readr)

# Load the dataset
dataset_path <- "path/to/indianelection.csv"
voter_data <- read_csv(dataset_path)

# Display the structure and first few rows of the dataset
str(voter_data)
head(voter_data)

# Data cleaning: Remove any rows with missing values and ensure consistency 
voter_data_clean <- voter_data %>%
  filter(!is.na(totvotpoll), !is.na(electors)) %>%
  mutate(partyname = ifelse(partyname == "Indian Natioanl Congress (I)",
                            "Indian National Congress", partyname))

# Calculate voter turnout percentage
voter_data_clean <- voter_data_clean %>%
  mutate(turnout_percentage = (totvotpoll / electors) * 100)

Output:

                    st_name year pc_no                   pc_name pc_type
1 Andaman & Nicobar Islands 1977     1 Andaman & Nicobar Islands     GEN
2 Andaman & Nicobar Islands 1977     1 Andaman & Nicobar Islands     GEN
3 Andaman & Nicobar Islands 1980     1 Andaman & Nicobar Islands     GEN
4 Andaman & Nicobar Islands 1980     1 Andaman & Nicobar Islands     GEN
5 Andaman & Nicobar Islands 1980     1 Andaman & Nicobar Islands     GEN
6 Andaman & Nicobar Islands 1980     1 Andaman & Nicobar Islands     GEN
          cand_name cand_sex                partyname partyabbre totvotpoll electors
1       K.R. Ganesh        M             Independents        IND      25168    85308
2 Manoranjan Bhakta        M Indian National Congress        INC      35400    85308
3   Ramesh Mazumdar        M             Independents        IND        109    96084
4     Alagiri Swamy        M             Independents        IND        125    96084
5       Kannu Chemy        M             Independents        IND        405    96084
6         K.N. Raju        M             Independents        IND        470    96084

Perform Exploratory Data Analysis (EDA)

We’ll start with some basic exploratory data analysis to understand the data better. Exploratory Data Analysis or EDA is a statistical approach or technique for analyzing data sets to summarize their important and main characteristics generally by using some visual aids. EDA Gives a descriptive statistical analysis of the dataset. We are going to use the EDA to summarize the indian election dataset statistically.The below code performs the EDA and outputs the Summary Statistics of the dataset.

# Summary statistics of the dataset
summary(voter_data_clean)

# Number of unique parties
unique_parties <- voter_data_clean %>%
  select(partyname) %>%
  distinct()
nrow(unique_parties)

Output:

           st_name           year          pc_no                pc_name     
 Uttar Pradesh :14791   Min.   :1977   Min.   : 1.00   Belgaum      :  567  
 Bihar         : 7727   1st Qu.:1989   1st Qu.: 7.00   Nalgonda     :  563  
 Maharashtra   : 6458   Median :1996   Median :18.00   East Delhi   :  434  
 Tamil Nadu    : 5309   Mean   :1997   Mean   :22.31   Chandni Chowk:  344  
 Andhra Pradesh: 5236   3rd Qu.:2004   3rd Qu.:33.00   Lucknow      :  319  
 Madhya Pradesh: 5196   Max.   :2014   Max.   :85.00   Outer Delhi  :  319  
 (Other)       :28364                                  (Other)      :70535  
 pc_type                 cand_name     cand_sex      partyname        
    : 8070   None Of The Above:  543   F   : 3648   Length:73081      
 GEN:54862   Ashok Kumar      :   87   M   :68885   Class :character  
 SC : 7293   Om Prakash       :   78   NULL:  542   Mode  :character  
 SC :   15   Raj Kumar        :   57   O   :    6                     
 ST : 2841   Ram Singh        :   54                                  
             Rajesh Kumar     :   51                                  
             (Other)          :72211                                  
   partyabbre      totvotpoll        electors       turnout_percentage
 IND    :41127   Min.   :     0   Min.   :  19471   Min.   : 0.00000  
 INC    : 4800   1st Qu.:   872   1st Qu.: 912985   1st Qu.: 0.07908  
 BJP    : 3350   Median :  2743   Median :1099503   Median : 0.25292  
 BSP    : 2624   Mean   : 49835   Mean   :1122277   Mean   : 4.83608  
 SP     : 1057   3rd Qu.: 19185   3rd Qu.:1329086   3rd Qu.: 1.86813  
 JD     :  943   Max.   :863358   Max.   :3368399   Max.   :68.27193  
 (Other):19180                                                        

[1] 1424

Visualize Voter Turnout Trends

We’ll create visualizations to explore voter turnout trends over time and by party. For this purpose we use certain R packages that helps in great visualizations. Examples include ggplot2,ggmap,leaflet,plotly and etc.

1. Voter Turnout Over Time

A versatile package used in R for creating visually appealing and customizable graphics.

# Aggregate data by year
turnout_by_year <- voter_data_clean %>%
  group_by(year) %>%
  summarise(total_votes = sum(totvotpoll), total_electors = sum(electors)) %>%
  mutate(turnout_percentage = (total_votes / total_electors) * 100)

# Plot voter turnout percentage over time
ggplot(turnout_by_year, aes(x = year, y = turnout_percentage)) +
  geom_line() +
  geom_point() +
  labs(title = "Voter Turnout Percentage Over Time",
       x = "Year",
       y = "Turnout Percentage") +
  theme_minimal()

Output:

Election Voter Turnout Visualization

By executing the above code, we understand that the total number of unique parties that are contesting in the election in india amounts.

2. Gender Distribution of Candidates

A bar plot to visualize the gender distribution of candidates.

# Aggregate data by gender
gender_distribution <- voter_data_clean %>%
  group_by(cand_sex) %>%
  summarise(count = n())

# Plot gender distribution of candidates
ggplot(gender_distribution, aes(x = cand_sex, y = count, fill = cand_sex)) +
  geom_bar(stat = "identity") +
  labs(title = "Gender Distribution of Candidates",
       x = "Gender",
       y = "Number of Candidates") +
  theme_minimal() +
  scale_fill_manual(values = c("M" = "blue", "F" = "pink"))

Output:

Election Voter Turnout Visualization

The bar plot analysis of the gender distribution of the candidates shouts out that there is a clear majority of male caditates those who are contesting in the election than the female candidates who are fewer in number. This suggests that more female candidates has to participate in the election contest and involve more in politics.

3. Bar Plot of Total Votes Polled by Year

We will create a Bar Plot of Total Votes Polled by Year.

# Bar plot of total votes polled by year
votes_by_year <- voter_data %>%
  group_by(year) %>%
  summarize(TotalVotes = sum(totvotpoll))

p1 <- ggplot(votes_by_year, aes(x = year, y = TotalVotes)) +
  geom_bar(stat = "identity", fill = "blue") +
  theme_minimal() +
  labs(title = "Total Votes Polled by Year", x = "Year", y = "Total Votes")

# Display plot
print(p1)

Output:

Election Voter Turnout Visualization

4.Creating Dashboard with Shiny

In R programming, we can create interactive dashboards using the shiny R package by creating a shiny app and maintaining the appropriate folder structure. Below code gives you the interactive dashboard that visualizes the voter turnout percentage for every political party in India.

# Shiny app for interactive dashboard
ui <- fluidPage(
  titlePanel("Voter Turnout Analysis"),
  sidebarLayout(
    sidebarPanel(
      selectInput("year", "Select Year", choices = unique(voter_data_clean$year)),
      selectInput("party", "Select Party", choices = unique(voter_data_clean$partyname))
    ),
    mainPanel(
      plotOutput("turnoutPlot"),
      plotOutput("partyPlot")
    )
  )
)

server <- function(input, output) {
  filtered_data <- reactive({
    voter_data_clean %>%
      filter(year == input$year)
  })
  
  filtered_party_data <- reactive({
    voter_data_clean %>%
      filter(partyname == input$party)
  })
  
  output$turnoutPlot <- renderPlot({
    ggplot(filtered_data(), aes(x = pc_name, y = turnout_percentage)) +
      geom_bar(stat = "identity", fill = "steelblue") +
      labs(title = paste("Voter Turnout in", input$year),
           x = "Parliamentary Constituency",
           y = "Turnout Percentage") +
      theme_minimal()
  })
  
  output$partyPlot <- renderPlot({
    ggplot(filtered_party_data(), aes(x = year, y = turnout_percentage, color = pc_name)) +
      geom_line() +
      geom_point() +
      labs(title = paste("Turnout Percentage for", input$party, "Over Years"),
           x = "Year",
           y = "Turnout Percentage") +
      theme_minimal()
  })
}

shinyApp(ui = ui, server = server)

Output:

Conclusion

Our analysis of the Indian national election dataset interestingly reveals several trends:

1. Voter turnout has varied over the years, with noticeable peaks and troughs.
2. Certain parties show higher voter turnout percentages, which could be due to their popularity or campaign strategies.
3. Predictions suggest that voter turnout may continue to follow current trends, though this is subject to various factors not accounted for in our simple model.