line chart
timeseries
Author

Ana Luisa Bodevan

Published

November 25, 2025

This week challenge dataset in on World Bank SPI. Check the TidyTuesday GitHub repo for the data.

1. SETUP

1.1 Load libraries and data

Code 
library(pacman)
pacman :: p_load(tidytuesdayR, tidyverse, dplyr, janitor, ggtext, showtext, scales, glue, ggrepel)

tuesdata <- tidytuesdayR::tt_load('2025-11-25')

spi_indicators <- tuesdata$spi_indicators |>
  janitor :: clean_names()

rm(tuesdata)

2. DATA WRANGLING

Code 
spi <- spi_indicators |>
  mutate(
    year = as.integer(year),
    income = factor(
      income,
      levels = c(
        "Low income",
        "Lower middle income",
        "Upper middle income",
        "High income"
      )
    )
  ) |>
  filter(!is.na(overall_score),
  !is.na(income))

spi_summary <- spi |>
  group_by(income, year) |>
  summarise(
    median = median(overall_score, na.rm = TRUE),
    mean = mean(overall_score, na.rm = TRUE),
    p25 = quantile(overall_score, .25, na.rm = TRUE),
    p75 = quantile(overall_score, .75, na.rm = TRUE),
    .groups = "drop"
  )
Code 
ggplot(spi_summary, aes(x = year, y = median)) +
  geom_ribbon(aes(ymin = p25, ymax = p75), alpha = 0.2, fill = "#95a5a6") +
  geom_line(linewidth = 1.4, color = "black") +
  facet_wrap(~ income, scales = "free_y") +
  theme_minimal()

Now for some more detailed statistics and call outs:

Code 
latest_year <- max(spi_summary$year, na.rm = TRUE)

gap_df <- spi_summary |>
  filter(year == latest_year, income %in% c("High income", "Low income")) |>
  select(income, median) |>
  tidyr::pivot_wider(names_from = income, values_from = median) |>
  mutate(gap = `High income` - `Low income`)

gap_value <- round(gap_df$gap, 1)

3. PLOT

3.1 Fonts and colors

Code 
font_add_google("Source Sans Pro", "source_sans")
showtext_auto()

title <- "Statistical Performance Has Improved, but Inequalities Persist"
subtitle <- glue(
  "The SPI was created to help countries assess and improve the performance of their statistical systems.\nDivergence between median and mean scores reveals that underperforming outliers drag down averages across income levels."
)
caption <- "Data: World Bank | #TidyTuesday | Graphic: Ana Bodevan"

wb_colors <- c(
  "Low income" = "#C03728",
  "Lower middle income" = "#E27A3F",
  "Upper middle income" = "#3A87B8",
  "High income" = "#2A6F2A"
)

3.2 Theme

Code 
theme_worldbank <- function(base_size = 13, base_family = "source_sans") {
  theme_minimal(base_size = base_size, base_family = base_family) %+replace%
    theme(
      text = element_text(family = base_family, color = "#1A1A1A"),

      # ---- Title + Subtitle Left Aligned ----
      plot.title.position = "plot",
      plot.title = element_text(
        size = base_size * 1.8,
        face = "bold",
        hjust = 0,
        margin = margin(b = 6)
      ),
      plot.subtitle = element_text(
        size = base_size * 1.2,
        hjust = 0,
        margin = margin(b = 12)
      ),

      # ---- Axis + Grid ----
      axis.title = element_text(size = base_size * 1.1),
      axis.text = element_text(size = base_size * 0.9),
      panel.grid.major.x = element_blank(),
      panel.grid.major.y = element_line(color = "#D8D8D8", linewidth = 0.3),
      panel.grid.minor = element_blank(),

      # ---- Facet titles (strip) ----
      strip.text = element_text(
        size = base_size * 1.2,
        face = "bold",
        margin = margin(t = 8, b = 8)
      ),

      # ---- Layout + Padding ----
      plot.margin = margin(20, 20, 20, 20),
      legend.position = "none",
      plot.caption.position = "plot"
    )
}

3.3 Plot

Code 
ggplot(spi_summary, aes(x = year)) +

  # IQR ribbon (soft World Bank-style shading)
  geom_ribbon(
    aes(ymin = p25, ymax = p75, fill = income),
    alpha = 0.12
  ) +

  # Median (solid)
  geom_line(
    aes(y = median, color = income),
    linewidth = 1.3
  ) +

  # Mean (dashed)
  geom_line(
    aes(y = mean, color = income),
    linewidth = 1,
    linetype = "dashed",
    alpha = 0.85
  ) +

  # Manual colors, legend removed above
  scale_color_manual(values = wb_colors) +
  scale_fill_manual(values = wb_colors) +

  # 2×2 layout (side-by-side)
  facet_wrap(~income, ncol = 2, scales = "free_y") +

  labs(
    title = title,
    subtitle = subtitle,
    caption = caption,
    x = "Year",
    y = "Overall SPI Score"
  ) +

  theme_worldbank()