## Load needed packages
library(readxl)
library(tidyverse)
library(rworldmap)
library(ggthemes)

## Input the data
d <- read_excel("All_data_FIW_2013-2021-tidy.xlsx", sheet = 2)

Section 1: Introducing the Freedom House Data Project

Freedom House is run by its’ President, Michael J. Abramowitz. Other than him, there is a non-partisan Board of Trustees that helps to run the more formal affairs of the group. The board of trustees is made up of business and labor leaders. The research itself is conducted by around 150 staff members around the world, operating out of offices in New York, Washington, and internationally.

The aims of this group are simple. These leaders share a common belief that promoting democracy and human rights is a necessary and just cause, with the potential to help secure international peace. They do not simply create data packs and release them, they also are involved directly in supporting “frontline defenders of freedom”.

This project is important because it highlights the state of freedom and human rights internationally. It allows the reality to be brought forward so that more people can get involved and find ways they too can contribute to spreading freedom. Not only does this organization spread awareness, they also help in the form of advocacy. They advocate for policy choices that support freedom and democracy. They also advocate for groups who are fighting for freedom, especially in close proximity to authoritarian regimes.

Section 2: Analyzing the Methodology

The methodology used by the experts in Freedom House has many strengths. The questions in order to determine a country or territory’s ratings are very detailed, with many factors going into each one. This allows for an extreme amount of clarification in order to give a country an accurate score in each area (1). The process for ratings could potentially introduce bias, however it seems extremely thorough. This lends itself to the idea that Freedom House’s scores are rigorously judged and likely very accurate. An important strength among their system is that they acknowledge that legal rights are not always guarantees of actual rights (2). That authoritarian regimes often disregard their own laws altogether. This means that they look into actual civil rights as exercised by the populations of each area.

Some of the weaknesses of the methodology lie in the way research itself is conducted. Freedom House typically only changes scores in the presence of what they call a “signal event” (3). This is an event that demonstrates a real-world development changing the conditions of a country. This leads to a lot of potentially missed data. Another issue present with Freedom House’s methodology is that the research itself does not seem to always be entirely well-defined. The sources themselves are vague, such as “on-the-ground research” and “individual professional contacts” (4). This leaves much to be wondered about how exactly Freedom House collects some of its’ data. There could also be bias introduced in the release of their scores due to the review meetings that must be attended. If the experts have a staunch bias for an undisclosed reason, they may be unwilling to verify some of the information that the research demonstrates.

Section 3: Analyzing the World in 2021

The visuals below list the current ratings for countries as of the 2021 data provided by Freedom House. There is a 7 point scale for the political rights and the civil liberties sections. The Freedom Statuses maintain 3 categories, free, partially free, and not free. The data demonstrates that when it comes to political rights and civil liberties ratings, there is a larger amount of countries that fall into the lower ratings. In the case of freedom statuses, there is a larger amount of countries fitting into the free category, however, there is only a slight difference between the partially free and not free categories.

d2021 <- d |>
  filter(Edition == 2021)

first <- ggplot(data = d, aes(x = `CL rating`))
first + geom_bar(fill = "blue3", color = "black", width = .5)+
  labs(x = "CL Rating", y = "Frequency", caption = "Source: Freedom House", title = "Civil Liberties Rating's for 2021" )+
  theme_bw()

first2 <- ggplot(data = d, aes(x = `PR rating`))
first2 + geom_bar(fill = "blue3", color = "black", width = .5)+
  labs(x = "PR Rating", y = "Frequency", caption = "Source: Freedom House", title = "Political Rights Rating's for 2021" )+
  theme_bw()

first3 <- ggplot(data = d, aes(x = `Status`))
first3 + geom_bar(fill = "blue3", color = "black", width = .5)+
  labs(x = "Status", y = "Frequency", caption = "Source: Freedom House", title = "Freedom Statuses for 2021" )+
  theme_bw()

The visuals below represent several histograms of the current distribution of different variables. First there is the Political Rights scores, then the Civil Liberties scores, and finally the Total scores. The Political Rights scores are a total of all of the points that a coutry or region earned in that section of the analysis, the same for the Civil Liberties scores.

d2021 <- d |>
  filter(Edition == 2021)

first4 <- ggplot(data = d, aes(x = `PR`))
first4 + geom_histogram(binwidth = 1,fill = "blue3", color = "black")+
  labs(x = "PR", y = "Frequency", caption = "Source: Freedom House", title = "Political Rights Scores Frequency Distribution 2021" )+
  theme_bw()

The data here is skewed left, meaning that there is a higher frequency of countries that had higher PR scores.

d2021 <- d |>
  filter(Edition == 2021)

first5 <- ggplot(data = d, aes(x = `CL`))
first5 + geom_histogram(binwidth = 1, fill = "blue3", color = "black")+
  labs(x = "CL", y = "Frequency", caption = "Source: Freedom House", title = "Civil Liberties Scores Frequency Distribution 2021" )+
  theme_bw()

The data here is also skewed left demonstrating higher frequency of countries with higher CL scores.

first6 <- ggplot(data = d, aes(x = `Total`))
first6 + geom_histogram(binwidth = 1, fill = "blue3", color = "black")+
  labs(x = "Total", y = "Frequency", caption = "Source: Freedom House", title = "Total Scores Frequency Distribution 2021" )+
  theme_bw()

The data here feels more balanced, yet it still is skewed left. The Total scores still lean toward the higher end, though there is a significant chunk around the 20-30 range, as well as another around the 60 range.

The Below maps represent the current status of the world.

d_2021 <- d |> filter(Edition == 2021)

map1 <- joinCountryData2Map(dF = d_2021,
joinCode = "NAME",
nameJoinColumn = "Country/Territory")
## 194 codes from your data successfully matched countries in the map
## 16 codes from your data failed to match with a country code in the map
## 49 codes from the map weren't represented in your data
mapCountryData(map1,
nameColumnToPlot = "CL rating",
mapTitle="Civil Liberties in 2021 (FH)",
catMethod = "categorical",
colourPalette = "heat")

This shows the distribution of Civil Liberty ratings across the world for each country.

d_2021 <- d |> filter(Edition == 2021)

map1 <- joinCountryData2Map(dF = d_2021,
joinCode = "NAME",
nameJoinColumn = "Country/Territory")
## 194 codes from your data successfully matched countries in the map
## 16 codes from your data failed to match with a country code in the map
## 49 codes from the map weren't represented in your data
mapCountryData(map1,
nameColumnToPlot = "PR rating",
mapTitle="Political Rights in 2021 (FH)",
catMethod = "categorical",
colourPalette = "heat")

This map demonstrates the distribution of the Political Rights ratings in 2021 across the world.

map1 <- joinCountryData2Map(dF = d_2021,
joinCode = "NAME",
nameJoinColumn = "Country/Territory")
## 194 codes from your data successfully matched countries in the map
## 16 codes from your data failed to match with a country code in the map
## 49 codes from the map weren't represented in your data
mapCountryData(map1,
nameColumnToPlot = "Status",
mapTitle="Freedom Statuses in 2021 (FH)",
catMethod = "categorical",
colourPalette = "heat")

This final map demonstrates the Freedom Statuses of countries around the globe. When viewing it in this regard there seems to be an extremely high concentration of the ‘Not Free’ categorization.

Section 4: Analyzing the World Across Time

d_summary <- d |>
  group_by(Edition) |>
    summarize(
      Mean = mean(Total),
      pct25 = quantile(Total, probs = .25),
      Median = median(Total),
      pct75 = quantile(Total, probs = .75)
    )

d |>
  ggplot(aes(x = factor(Edition), y = Total)) +
  geom_boxplot() +
  labs(x = "Edition", title = "Total Scores Distribution Since 2013")+
  theme_bw()

This series of box plots shows how the Total scores distributions have changed since 2013. There is a slight shift downward overall by 2021.

d_summary <- d |>
  group_by(Edition) |>
    summarize(
      Mean = mean(Total),
      pct25 = quantile(Total, probs = .25),
      Median = median(Total),
      pct75 = quantile(Total, probs = .75)
    )

d |>
  ggplot(aes(x = factor(Edition), y = CL)) +
  geom_boxplot() +
  labs(x = "Edition", y = "Civil Liberties", title = "Civil Liberties Distribution Since 2013") + 
  theme_bw()

This set of box plots shows the distribution of the Civil Liberties scores sine 2013. This set of scores also follows a similar pattern, with a downward trend over the last few years.

d_summary <- d |>
  group_by(Edition) |>
    summarize(
      Mean = mean(Total),
      pct25 = quantile(Total, probs = .25),
      Median = median(Total),
      pct75 = quantile(Total, probs = .75)
    )

d |>
  ggplot(aes(x = factor(Edition), y = PR)) +
  geom_boxplot() +
  labs(x = "Edition", y = "Political Rights", title = "Political Rights Distribution Since 2013") + 
  theme_bw()

This series of box plots demonstrates the distribution of the Political Rights scores since 2013. similarly to the last set of box plots, there has been a noticeable shift downward by 2021.

d_summary <- d |>
  group_by(Edition) |>
    summarize(
      Mean = mean(Total),
      pct25 = quantile(Total, probs = .25),
      Median = median(Total),
      pct75 = quantile(Total, probs = .75)
    )

 
  p <- ggplot(data = d_summary, aes(x = Edition))

p +
  geom_line(aes(y=pct25), color = "red", linetype = "dotdash") +
  geom_line(aes(y=Median), color = "blue") +
  geom_line(aes(y=pct75), color = "red", linetype = "dotdash")+
  coord_cartesian(ylim = c(0,100)) +
  theme_bw() +
  scale_x_continuous(breaks = 2013:2021) +
  labs(x = "", y = "Total Scores") +
  annotate("text", x = 2014, y = 95, label = "75th pct", size = 3, color = "red")

This line plot isolates the 75th and 25th percentiles, as well as the median values of the Total scores since 2013. This is useful, while the box plots demonstrate outliers and show the values extremely well, this line plot gives a more clear look at the trends of the data.

# Find the top questions
top <- d |>
  filter(Edition == 2013) |>
  arrange(Total) |>
  select(`Country/Territory`)|>
  slice(200:211)

  top
## # A tibble: 10 x 1
##    `Country/Territory`
##    <chr>              
##  1 Denmark            
##  2 Liechtenstein      
##  3 Barbados           
##  4 Netherlands        
##  5 Finland            
##  6 Iceland            
##  7 Luxembourg         
##  8 Norway             
##  9 San Marino         
## 10 Sweden
  d |>
  filter(`Country/Territory` %in%top[[1]]) |>
  ggplot(aes(x = Edition, y = Total, color = `Country/Territory`)) +
  geom_line() +
    labs(title = "Top Countries Total Scores Over Time")+
    theme_bw()

This line graph demonstrates the Total scores of the 10 countries with the highest scores in 2013. It is important to note these countries represent what were the highest Total freedom ratings as of 2013. Whatever policy and cultural choices that made them this way are likely able to be replicated.

  # Find the bottom questions

bottom <- d |>
  filter(Edition == 2013) |>
  arrange(Total) |>
  select(`Country/Territory`)|>
  slice(1:10)

  bottom
## # A tibble: 10 x 1
##    `Country/Territory`
##    <chr>              
##  1 Tibet              
##  2 Somalia            
##  3 Eritrea            
##  4 North Korea        
##  5 Uzbekistan         
##  6 South Ossetia      
##  7 Syria              
##  8 Western Sahara     
##  9 Sudan              
## 10 Turkmenistan
d |>
  filter(`Country/Territory` %in%bottom[[1]]) |>
  ggplot(aes(x = Edition, y = Total, color = `Country/Territory`)) +
  geom_line() +
  labs(title = "Bottom Countries Total Scores Over Time") +
  theme_bw()

Similarly to the previous plot, this chart shows how the 10 countries with the lowest Total scores as of 2013 have changed since. The conditions within each country are important to note, because like the top 10 countries, the policies, attitudes, and events are replicable and thus preventable.

Section 5: Conclusions

Overall, the aims and goals of Freedom House are noble and important. Without the knowledge of the state of freedom and democracy in the world, it would be all too easy to slip into authoritarian control worldwide. Their methodology seems strong, though not as intense as it should be. The lack of consistent change to scores is worrisome to say the least. This likely has much to do with staffing and workload issues. Due to the majority of their funding coming from the U.S. State Department, they likely can only afford so many staff members and analysts to conduct research. If this were to be expanded upon in the future, they could potentially begin to more consistently adjust the scores of each country and region.

The data provided by Freedom House does not show a significant change anywhere, and this is likely due to how Freedom House chooses to typically only change their scores after significant events. Despite this, there has been slight declines in most scores since 2013. This data project highlights a decline in overall democracy since 2013. This has not been a rapid decline, but a gradual one. That almost seems more dangerous. A rapid decline would be clear, it would ignite change. A gradual change may, and does, go unnoticed by the general public. The hopes of combating a gradual change rely on convincing people it is a significant enough danger that it will actually affect them. Once they are at that point it is even more difficult to inspire them to do something if they only barely feel the effects of this shift. This project should be used as a warning and tool in order to inspire a marked shift back towards democracy. If it is not effective in this, I fear the change will continue.