democracy_index.qmd

democracy_index.qmd

```{r}
setwd(here::here('democracy_index'))
library(tidyverse)
library(shadowtext)
library(ggiraph)
library(ggtext)
library(gdtools)
register_gfont("Roboto Mono")
addGFontHtmlDependency(family = "Roboto Mono")
register_gfont("Open Sans")
addGFontHtmlDependency(family = "Open Sans")
```

Load Data

```{r}
# https://github.com/xmarquez/democracyData
scores <- democracyData::eiu |> 
  select(eiu_name = eiu_country, score = eiu, year, full_name = extended_country_name)
```


```{r}
# Latest scores extracted manually from
# https://www.eiu.com/n/campaigns/democracy-index-2022/
scores_2022 <- tibble(
  input = read_lines('manual_2022.txt'),
  score = parse_number(input), # grabs first number which luckily is the score,
  country = str_extract(input, '[a-zA-Zô’ ]+') # Extra characters for Côte d’Ivoire
) |> 
  mutate(country = str_sub(country, end = -2)) |> 
  select(country, score) 
```

Correct some parsing  and naming issues

```{r}
## Correcting the following parsing errors
# Warning: 2 parsing failures.                                                                         
# row col expected                                              actual
#  44  -- a number Timor-Leste 7.06 44 -1 9.58 5.93 5.56 6.88 7.35    
# 141  -- a number Guinea-Bissau 2.56 140= -2 4.00 0.00 3.33 3.13 2.35

scores_2022[44, ] <- tibble(country = 'Timor Leste', score = 7.06) # Spelled without dash in other data
scores_2022[141, ] <- tibble(country = 'Guinea-Bissau', score = 2.56)
scores_2022 <- scores_2022 |> 
  mutate(country = if_else(country == 'Eswatini', 'eSwatini', country))
```

Next, fix more naming issues in order to merge the two data sets.

```{r}
non_compliant_eiu_names <- scores_2022 |> 
  filter(!(country %in% scores$eiu_name)) |> 
  pull(country)

replace_dict <- scores |> 
  filter(full_name %in% non_compliant_eiu_names, year == 2006) |> 
  select(eiu_name, non_compliant_name = full_name) 

replacement_vector <- replace_dict$eiu_name
names(replacement_vector) <- paste0("^", replace_dict$non_compliant_name, "$")

str_replace_all(scores_2022$country, replacement_vector)

scores_2022_adjusted <- scores_2022 |> 
  mutate(eiu_name = str_replace_all(country, replacement_vector)) |> 
  mutate(year = 2022)

all_scores <- bind_rows(scores, scores_2022_adjusted) |> 
  arrange(eiu_name) |> 
  fill('full_name') |> 
  select(-country) |> 
  mutate(
    full_name = case_when(
      full_name == 'Russia (Soviet Union)' ~ 'Russia',
      full_name == 'United States of America' ~ 'United States',
      TRUE ~ full_name
    )
  ) 
```

```{r}
selected_countries <- c('Norway', 'United States', 'India', 'Peru', 'Russia', 'China')

ranked_data <- all_scores |> 
  mutate(
    democracy_status = case_when(
      score >= 8 ~ 'Full democracy',
      score >= 7 ~ 'Flawed democracy2',
      score >= 6 ~ 'Flawed democracy1',
      score >= 5 ~ 'Hybrid regime2',
      score >= 4 ~ 'Hybrid regime1',
      score >= 3 ~ 'Authoritarian regime3',
      score >= 2 ~ 'Authoritarian regime2',
      TRUE ~ 'Authoritarian regime1'
    ),
    color = case_when(
      democracy_status == 'Full democracy'  ~ '#49558f',
      democracy_status == 'Flawed democracy2'~ '#799ec2',
      democracy_status == 'Flawed democracy1'~ '#afc5da',
      democracy_status == 'Hybrid regime2'~ '#ffe4af',
      democracy_status == 'Hybrid regime1'~ '#ffce6e',
      democracy_status == 'Authoritarian regime3'~ '#ffa9a6',
      democracy_status == 'Authoritarian regime2'~ '#f6423c',
      TRUE ~ '#bf0f09',
    ),
    text_color = if_else(color == '#49558f', '#E5E5E5', 'grey10'),
    full_name = str_remove_all(full_name, "'")
  ) |> 
  group_by(year) |> 
  mutate(rank = row_number(desc(score))) 
```

Compute data for the border between democracy/regime groups

```{r}
border_data <- ranked_data |> 
  filter(!(year %in% c(2007, 2009)), !str_detect(democracy_status, 'Full')) |> 
  group_by(year, democracy_group = str_remove(democracy_status, '[1-3]')) |> 
  summarise(max_rank = max(167 - rank + 1), .groups = 'drop')


compute_step_border <- function(dem_group) {
  border_data_dem_group <- border_data |> 
    filter(democracy_group == dem_group)
  
  tibble(
    democracy_group = dem_group,
    x = c(rbind(border_data_dem_group$year - 0.5, border_data_dem_group$year + 0.5)),
    y = rep(border_data_dem_group$max_rank, each = 2) + 0.5
  ) 
}

border_data_step <- map_dfr(unique(border_data$democracy_group), compute_step_border)
```

Create plot

```{r}
y_pos_democracy_labels <- c(20, 75, 115, 160)

plt <- ranked_data |> 
  ggplot(aes(x = year, y = 167 - rank + 1))  +
  geom_tile_interactive(
    data = ranked_data, 
    aes(data_id = full_name, fill = color), height = 1
  ) +
  geom_step_interactive(
    data = border_data_step,
    aes(x = x, y = y, group = democracy_group, data_id = democracy_group),
    linewidth = 1
  ) +
  geom_text_interactive(
    data = ranked_data |> filter(!(full_name %in% selected_countries), year == 2022),
    aes(data_id = full_name, label = paste0(rank, '. ', full_name)),
    hjust = 0,
    position = position_nudge(x = 0.55),
    alpha = 0,
    size = 5,
    family = "Open Sans"
  ) +
  geom_text_interactive(
    data = ranked_data |> filter((full_name %in% selected_countries), year == 2022),
    aes(data_id = full_name, label = paste0(rank, '. ', full_name) ),
    hjust = 0,
    position = position_nudge(x = 0.55),
    alpha = 1,
    size = 5,
    family = "Open Sans"
  ) +
  geom_text_interactive(
    aes(data_id = full_name, label = score),
    color = ranked_data$text_color,
    alpha = 0,
    size = 3.75,
    hjust = 0.5,
    position = position_nudge(y = -0.8),
    family = "Open Sans",
    fontface = 'bold'
  ) +
  annotate(
    'tile',
    x = 2007,
    y = 1:167,
    fill = 'grey70',
    height = 1.2,
    width = 1.03,
    col = 'grey70'
  ) +
  annotate(
    'tile',
    x = 2009,
    y = 1:167,
    fill = 'grey70',
    height = 1.2,
    width = 1.03,
    col = 'grey70'
  ) +
  annotate(
    'shadowtext',
    x = 2006.5,
    y = y_pos_democracy_labels,
    label = c('Authoritarian regimes', 'Hybrid regimes', 'Flawed democracies', 'Full democracies'),
    hjust = 0,
    size = 6.5,
    fontface = 'bold',
    color = 'black',
    bg.color = 'white',
    family = "Open Sans"
  )  +
  theme_minimal(base_size = 20, base_family = 'Open Sans') +
  theme(
    legend.position = 'none', 
    panel.grid = element_blank(),
    plot.title = element_text(
      family = 'Merriweather',
      face = 'bold',
      size = 32
    ),
    plot.subtitle = element_markdown(lineheight = 1.25),
    plot.caption = element_markdown(family = 'Roboto Mono', size = 14)
  ) +
  coord_cartesian(xlim = c(2005.5, 2027), ylim = c(1, 168.5), expand = FALSE) +
  labs(x = element_blank(), y = element_blank()) +
  scale_x_continuous(
    breaks = 2006:2022,
    labels = c('2006', paste0(0, 7:9), 10:22)
  ) +
  labs(
    title = 'The world’s most, and least, democratic countries in 2022',
    subtitle = 'The Economist Intelligence Unit released their most recent global democracy index<br> <span style= "font-size:16pt;">**10=most democratic**</span>',
    caption = 'Data: EIU | Design: The Economist Graphic Detail | Remake: Albert Rapp (Twitter: @rappa753)'
  ) +
  scale_y_continuous(breaks = NULL) +
  scale_fill_identity()


girafe(
  ggobj = plt,
  options = list(
    opts_hover(
      css = girafe_css(
        css = 'user-select: none;', 
        area = 'height:12px;stroke:black;stroke-width:1px;opacity:1;',
        text = 'fill-opacity:1;'
      )
    ),
    opts_hover_inv(css = girafe_css(
      css = 'user-select: none; opacity:0.3;',
      area = 'height:3px;',
    )),
    opts_sizing(rescale = FALSE)
  ),
  height_svg = 12,
  width_svg = 16
)
```