Visualise trendlines with "open" datapoints: support for government measures against income differences, 2002-2008

d_ess_w1w9_gincdif_w_subset.csv

vis_trendline_open_point_ess_incdiff_ess_2002_2018.r

# ################################################################## #
# Trendlines with 'open' datapoints: support for government measures #
#  against income differences, using ESS 2002-2018 data              #
# ################################################################## #

library(readr)
library(here)
library(dplyr)
library(ggplot2)
library(hrbrthemes)
library(ggtext)
library(ggrepel)

# read CSV-file with survey-weighted responses for selected countries
d_essw1w9_incomediff_w <- read_csv(
  file = here::here('d_ess_w1w9_gincdif_w_subset.csv'), 
  col_types = cols(
    essround = col_factor(),
    essround_year = col_factor(),
    country_iso2c = col_character(),
    country_name_en = col_character(),
    gincdif = col_character(),
    n = col_double(),
    n_se = col_double(),
    pct = col_double() ))  


d_essw1w9_incomediff_w <- d_essw1w9_incomediff_w  %>%
  # sum the two agree-categories by country
  filter(gincdif %in% c('Agree strongly', 'Agree')) %>%
  select(-gincdif) %>%
  group_by(essround_year, country_name_en) %>%
  tally(pct, name = 'pct_agree') %>%
  # create a label combining country name + rounded percentage agreeing
  mutate(
    country_label = if_else(
      essround_year == '2018', 
      paste(country_name_en, 
            as.character(scales::percent(pct_agree, accuracy = 1))), NA_character_))


# From https://stackoverflow.com/a/52251548/125085 :
# To obtain an "open point", you need a plot symbol with a
#  separate definable fill and stroke => 
#  You can check what these are with example("points") 
#  and going to the 3rd plot."

p_essw1w9_incomediff_w <- d_essw1w9_incomediff_w %>%
  ggplot(aes(x = essround_year, y = pct_agree, 
             group = country_name_en, color = country_name_en, 
             label = country_label)) + # specify label for geom_label_repel()
  geom_line() +
  geom_point(
    # leaving color unspecified here uses the color specified in aes() above,
    #  so it will be the same as the line color.
    size = 2.0,       # select an appropriate size for the point
    shape = 21,       # required: select a plot symbol with a separate fill and stroke
    stroke = 0.5,     # specify a narrow stroke around the point
    fill = "white") + # fill the point with white / "leave open"
  geom_label_repel(
    size = 4,    
    nudge_x = 0, xlim = 9.2,   # tweak horizontal positioning
    segment.colour = 'grey80', # connect point and label with grey line
    na.rm = TRUE, # ignore all points with no label, i.e. only label last point 2018
    direction = 'both') + # labels can move horizontal & vertically
  scale_x_discrete(
    limits = c('2002', '2004', '2006', '2008', '2010', 
               '2012', '2014', '2016', '2018', 
               '2020', '2022'), # manually specify to increase plot range past 2018
    breaks = c('2002', '2008', '2014', '2018')) + # only label selected years on x-axis
  scale_y_continuous(labels = scales::percent_format(accuracy = 1)) +
  theme_ipsum_rc() + # select theme (hrbrthemes)
  theme(
    # remove vertical grid lines
    panel.grid.major.x = element_blank(), panel.grid.minor.x = element_blank(),
    legend.position = "none",
    plot.subtitle = element_markdown()) + # enable markdown for subtitle italics (ggtext)
  labs(x = NULL, y = NULL,
       title = 'Support for government action against income inequality is increasing',
       subtitle = 'Percentage of respondents that (strongly) agrees with the statement "*The government should take<br>measures to reduce differences in income levels*", 2002 – 2018".',
       caption = 'Data: European Social Survey, wave 1 – wave 9, n = 116.444 • Maarten Hermans (@hermansm).')

ggsave(here::here('p_trendline_open_point_ess_incdiff_2002_2018_en.png'), 
       p_essw1w9_incomediff_w, width = 8, height = 6)

rm(d_essw1w9_incomediff_w, p_essw1w9_incomediff_w)