ct_selection_untitled-FF5A6079.R

gg1 <- cces %>% 
  filter(year == 2008 | year == 2024) %>% 
  mutate(race = frcode(race == 1 ~ "White", 
                       race != 1 ~ "Non-White")) %>% 
  filter(pid7 == 1 | pid7 == 2 | pid7 == 3) %>% 
  cces_attend(pew_attendance) %>% 
  group_by(year, race) %>% 
  ct(att, wt = weight, show_na = FALSE)


gg1 %>% 
  mutate(lab = round(pct, 2)) %>% 
  mutate(year = as.factor(year)) %>% 
  ggplot(., aes(x = fct_rev(race), y = pct, fill = fct_rev(att))) +
  geom_col(color = "black") + 
  coord_flip() +
  facet_wrap(~ year, ncol =1, strip.position = "top") +
  theme_rb() +
  scale_fill_manual(values = c(moma.colors("OKeeffe", 6))) +
  theme(legend.position = "bottom") +
  scale_y_continuous(labels = percent) +
  theme(strip.text.y.left = element_text(angle=0)) +
  guides(fill = guide_legend(reverse=T, nrow = 1)) +
  # theme(axis.title.y=element_blank(), axis.text.y=element_blank(), axis.ticks.y=element_blank()) +
  theme(panel.grid.minor.y=element_blank(), panel.grid.major.y=element_blank()) +
  geom_text(aes(label = ifelse(pct >.05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "black") +
  geom_text(aes(label = ifelse(att == "Seldom" & pct > .05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "white") +
  geom_text(aes(label = ifelse(att == "Yearly" & pct > .05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "white") +
  theme(plot.title = element_text(size = 16)) +
  theme(strip.text.y.left = element_text(angle = 0, hjust = 1)) +
  labs(x = "", y = "", title = "Religious Attendance of Democrats by Race, 2008 vs 2024", caption = "@ryanburge\nData: Cooperative Election Study, 2008-2024") 
save("attend_democrats_race_compare.png", wd = 9, ht = 4.5)


gg1 <- cces %>% 
  filter(year == 2008 | year == 2024) %>% 
  mutate(race = frcode(race == 1 ~ "White", 
                       race != 1 ~ "Non-White")) %>% 
  filter(pid7 == 1 | pid7 == 2 | pid7 == 3) %>% 
  mutate(relig = frcode(religion == 1 ~ "Protestant",
                        religion == 2 ~ "Catholic",
                        religion >= 3 & religion <= 8 ~ "Other World Religions", 
                        religion == 9 | religion == 10 ~ "Atheist/Agnostic", 
                        religion == 11 ~ "Nothing in Part.")) %>% 
  group_by(year, race) %>% 
  ct(relig, wt = weight, show_na = FALSE)


colors <- c(
  "Protestant" = "#1f78b4",        # Blue
  "Catholic" = "#33a02c",          # Green
  "Other World Religions" = "#e31a1c", # Red
  "Atheist/Agnostic" = "#ff7f00",  # Orange
  "Nothing in Part." = "#6a3d9a"   # Purple
)


gg1 %>% 
  mutate(lab = round(pct, 2)) %>% 
  mutate(year = as.factor(year)) %>% 
  ggplot(., aes(x = fct_rev(race), y = pct, fill = fct_rev(relig))) +
  geom_col(color = "black") + 
  coord_flip() +
  facet_wrap(~ year, ncol =1, strip.position = "top") +
  theme_rb() +
  scale_fill_manual(values = colors) +
  theme(legend.position = "bottom") +
  scale_y_continuous(labels = percent) +
  theme(strip.text.y.left = element_text(angle=0)) +
  guides(fill = guide_legend(reverse=T, nrow = 1)) +
  # theme(axis.title.y=element_blank(), axis.text.y=element_blank(), axis.ticks.y=element_blank()) +
  theme(panel.grid.minor.y=element_blank(), panel.grid.major.y=element_blank()) +
  geom_text(aes(label = ifelse(pct >.05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 9, family = "font", color = "black") +
  geom_text(aes(label = ifelse(relig == "Nothing in Part." & pct > .05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 9, family = "font", color = "white") +
  geom_text(aes(label = ifelse(relig == "Protestant" & pct > .05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 9, family = "font", color = "white") +
  theme(plot.title = element_text(size = 16)) +
  theme(strip.text.y.left = element_text(angle = 0, hjust = 1)) +
  labs(x = "", y = "", title = "Religious Composition of Democrats by Race, 2008 vs 2024", caption = "@ryanburge\nData: Cooperative Election Study, 2008-2024") 
save("relig_democrats_race_compare.png", wd = 9, ht = 4.5)




gg1 <- cces %>% 
  filter(year == 2008 | year == 2024) %>% 
  cces_race(race) %>% 
  filter(pid7 == 1 | pid7 == 2 | pid7 == 3) %>% 
  mutate(relig = frcode(religion == 1 ~ "Protestant",
                        religion == 2 ~ "Catholic",
                        religion >= 3 & religion <= 8 ~ "Other World Religions", 
                        religion == 9 | religion == 10 ~ "Atheist/Agnostic", 
                        religion == 11 ~ "Nothing in Part.")) %>% 
  group_by(year, race) %>% 
  ct(relig, wt = weight, show_na = FALSE)



colors <- c(
  "Protestant" = "#1f78b4",        # Blue
  "Catholic" = "#33a02c",          # Green
  "Other World Religions" = "#e31a1c", # Red
  "Atheist/Agnostic" = "#ff7f00",  # Orange
  "Nothing in Part." = "#6a3d9a"   # Purple
)

gg1 %>% 
  mutate(lab = round(pct, 2)) %>% 
  mutate(year = as.factor(year)) %>% 
  ggplot(., aes(x = fct_rev(year), y = pct, fill = fct_rev(relig))) +
  geom_col(color = "black") + 
  coord_flip() +
  facet_wrap(~ race, ncol =1, strip.position = "top") +
  theme_rb() +
  scale_fill_manual(values = colors) +
  theme(legend.position = "bottom") +
  scale_y_continuous(labels = percent) +
  theme(strip.text.y.left = element_text(angle=0)) +
  guides(fill = guide_legend(reverse=T, nrow = 1)) +
  # theme(axis.title.y=element_blank(), axis.text.y=element_blank(), axis.ticks.y=element_blank()) +
  theme(panel.grid.minor.y=element_blank(), panel.grid.major.y=element_blank()) +
  geom_text(aes(label = ifelse(pct >.05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "black") +
  geom_text(aes(label = ifelse(relig == "Nothing in Part." & pct > .05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "white") +
  geom_text(aes(label = ifelse(relig == "Protestant" & pct > .05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "white") +
  theme(plot.title = element_text(size = 16)) +
  theme(strip.text.y.left = element_text(angle = 0, hjust = 1)) +
  labs(x = "", y = "", title = "Religious Composition of Democrats by Race, 2008 vs 2024", caption = "@ryanburge\nData: Cooperative Election Study, 2008-2024") 
save("relig_democrats_race_compare_allraces.png", wd = 9, ht = 7.5)




gg2 <- cces %>% 
  mutate(race = frcode(race == 1 ~ "White", 
                       race != 1 ~ "Non-White")) %>% 
  filter(pid7 == 1 | pid7 == 2 | pid7 == 3) %>% 
  mutate(vimp = case_when(pew_importance == 1 ~ 1, 
                          pew_importance == 2 | pew_importance == 3 | pew_importance == 4 ~ 0)) %>% 
  group_by(year, race) %>% 
  mean_ci(vimp, wt = weight, ci = .84) %>% 
  filter(race != "NA") %>% 
  select(race, year, imp = mean)


gg3 <- cces %>% 
  filter(year >= 2008) %>% 
  mutate(race = frcode(race == 1 ~ "White", 
                       race != 1 ~ "Non-White")) %>% 
  filter(pid7 == 1 | pid7 == 2 | pid7 == 3) %>% 
  mutate(wk = case_when(pew_attendance == 1 | pew_attendance == 2 ~ 1, 
                        pew_attendance <= 6 ~ 0)) %>% 
  filter(pid3 != "NA") %>% 
  group_by(race, year) %>% 
  mean_ci(wk, wt = weight, ci = .84) %>% 
  select(race, year, wk = mean)


graph <- left_join(gg2, gg3) %>% na.omit() 


library(dplyr)
library(ggplot2)
library(grid)

# Compute predicted fit lines with imp sequence going from max to min (for arrow direction)
fit_lines <- graph %>%
  group_by(race) %>%
  do({
    mod <- lm(wk ~ imp, data = .)
    imp_seq <- seq(max(.$imp), min(.$imp), length.out = 100)  # reverse seq
    tibble(imp = imp_seq, wk = predict(mod, newdata = data.frame(imp = imp_seq)))
  })

pres_years <- c(2008, 2012, 2016, 2020, 2024)


graph %>% 
  ggplot(aes(x = imp, y = wk, color = race, group = race)) +
  geom_point(stroke = 1, shape = 21, show.legend = FALSE) +
  geom_path(data = fit_lines, aes(x = imp, y = wk, color = race), 
            arrow = arrow(length = unit(0.15, "inches"), type = "closed"), size = 1) +
  theme_rb() +
  geom_text_repel(
    data = filter(graph, year %in% pres_years),
    aes(label = year),
    family = "font",
    size = 5
  ) +
  add_text(y = .225, x = .225, word = "White", sz = 10) +
  add_text(y = .35, x = .45, word = "Non-White", sz = 10) +
  
  scale_color_calc() +
  scale_x_continuous(labels = scales::percent_format(), breaks = c(.25, .35, .45, .55)) +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, .5)) +
  labs(
    x = "Share Saying Religion is Very Important",
    y = "Share Attending Services Weekly",
    title = "Religion Importance and Religious Attendance among Democrats by Race", 
    caption = "@ryanburge\nData: Cooperative Election Study, 2008-2024"
  )

save("scatter_imp_att_race_dems.png")


gg1 <- cces %>% 
  filter(year == 2008 | year == 2024) %>% 
  mutate(race = frcode(race == 1 ~ "White", 
                       race != 1 ~ "Non-White")) %>% 
  filter(pid7 == 1 | pid7 == 2 | pid7 == 3) %>% 
  mutate(rel = frcode(religion <= 8 ~ "Religious", 
                      religion == 9 | religion == 10 | religion == 11 ~ "Non-Religious")) %>% 
  mutate(id5 = frcode(ideo5 == 1 ~ "Very Liberal", 
                      ideo5 == 2 ~ "Liberal", 
                      ideo5 == 3 ~ "Moderate", 
                      ideo5 == 4 ~ "Conservative", 
                      ideo5 == 5 ~ "Very Conservative")) %>% 
  group_by(year, race, rel) %>% 
  ct(id5, wt = weight, show_na = FALSE)
  

library(ggplot2)
library(scales)    # for percent formatting
library(forcats)   # for factor reordering

# Reorder id5 from Very Liberal to Very Conservative
gg1 <- gg1 %>% 
  filter(rel != 'NA') %>% 
  mutate(id5 = factor(id5, levels = c("Very Liberal", "Liberal", "Moderate", "Conservative", "Very Conservative")))

colors <- c(
  "Very Liberal"      = "#2166ac",  # Deep blue
  "Liberal"           = "#67a9cf",  # Lighter blue
  "Moderate"          = "#f7f7f7",  # Light gray/neutral
  "Conservative"      = "#f4a582",  # Light red/orange
  "Very Conservative" = "#b2182b"   # Deep red
)


gg1 %>% 
  mutate(lab = round(pct, 2)) %>% 
  mutate(year = as.factor(year)) %>% 
  ggplot(., aes(x = fct_rev(year), y = pct, fill = fct_rev(id5))) +
  geom_col(color = "black") + 
  coord_flip() +
  facet_grid(rel ~ race) +
  theme_rb() +
  scale_fill_manual(values = colors) +
  theme(legend.position = "bottom") +
  scale_y_continuous(labels = percent) +
  theme(strip.text.y.left = element_text(angle=0)) +
  guides(fill = guide_legend(reverse=T, nrow = 1)) +
  # theme(axis.title.y=element_blank(), axis.text.y=element_blank(), axis.ticks.y=element_blank()) +
  theme(panel.grid.minor.y=element_blank(), panel.grid.major.y=element_blank()) +
  geom_text(aes(label = ifelse(pct >.075, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "black") +
  geom_text(aes(label = ifelse(pct >.075 & id5 == "Very Liberal", paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "white") +
  # geom_text(aes(label = ifelse(att == "Yearly" & pct > .05, paste0(lab*100, '%'), '')), position = position_stack(vjust = 0.5), size = 8, family = "font", color = "white") +
  theme(plot.title = element_text(size = 16)) +
  theme(strip.text.y.left = element_text(angle = 0, hjust = 1)) +
  labs(x = "", y = "", title = "The Political Ideology of Democrats by Race and Religion", caption = "@ryanburge\nData: Cooperative Election Study, 2008-2024") 
save("relig_democrats_race_compare_allracesnn.png", wd = 10, ht = 4.5)






dvs_summary <- cces24 %>%
  filter(pid7 %in% 1:3) %>% 
  mutate(religion = religpew) %>% 
  mutate(
    race = frcode(race == 1 ~ "White", race != 1 ~ "Non-White"),
    rel = frcode(religion <= 8 ~ "Religious", 
                 religion == 9 | religion == 10 | religion == 11 ~ "Non-Religious"),
    cc444a = ifelse(CC24_444a == 1, 1, 0),
    cc444b = ifelse(CC24_444b == 1, 1, 0),
    cc444c = ifelse(CC24_444c == 1, 1, 0),
    cc444d = ifelse(CC24_444d == 1, 1, 0),
    cc444e = ifelse(CC24_444e == 1, 1, 0),
    cc444f = ifelse(CC24_444f == 1, 1, 0)
  ) %>%
  select(commonweight, race, rel, starts_with("cc444")) %>%
  pivot_longer(
    cols = starts_with("cc444"),
    names_to = "item",
    values_to = "support"
  ) %>%
  group_by(race, rel, item) %>%
  mean_ci(support, wt = commonweight) %>%
  ungroup() %>%
  mutate(
    type = case_when(
      item == "cc444a" ~ "Ban gender transition\nfor minors",
      item == "cc444b" ~ "Parental consent\nfor pronoun changes",
      item == "cc444c" ~ "Ban abortion\npills by mail",
      item == "cc444d" ~ "Ban out-of-state\nabortions",
      item == "cc444e" ~ "Require age\nchecks for porn",
      item == "cc444f" ~ "Support school\nvouchers"
    )
  )


dvs_summary <- dvs_summary %>%
  filter(race != "NA") %>% 
  filter(rel != "NA") %>% 
  mutate(race_rel = paste(race, rel, sep = " \n"))


colors <- c(
  "White \nReligious"     = "#2c7fb8",  # Medium blue
  "White \nNon-Religious" = "#7fcdbb",  # Light teal
  "Non-White \nReligious" = "#e08214",  # Bright orange
  "Non-White \nNon-Religious" = "#fdb863" # Light orange/yellow
)


ggplot(dvs_summary, aes(x = fct_rev(race_rel), y = mean, fill = race_rel)) +
  geom_col(color = "black") +
  facet_wrap(~ type, ncol = 2) +
  scale_y_continuous(labels = scales::percent_format(accuracy = 1)) +
  labs(
    x = "",
    y = "",
    fill = "Group",
    title = "Support for Policies by Combined Race & Religion Groups among Democrats",
    caption = "@ryanburge | Data: Cooperative Election Study 2024"
  ) +
  theme_rb() +
  scale_fill_manual(values = colors) +
  lab_bar(above = TRUE, type = mean, pos = .065, sz = 9) 
save("issue_compares_dem_race_relig.png", ht = 11, wd = 7.5)