Comparing the states by partisanship and other development metrics

partisan_indexing.R

library(tidyverse)
library(janitor)
library(magrittr)
library(campfin)
library(rvest)

url_edu <- "https://en.wikipedia.org/wiki/List_of_U.S._states_and_territories_by_educational_attainment"
url_life <- "https://en.wikipedia.org/wiki/List_of_U.S._states_and_territories_by_life_expectancy"
url_gdp <- "https://en.wikipedia.org/wiki/List_of_U.S._states_by_GDP_per_capita"
url_hdi <- "https://en.wikipedia.org/wiki/List_of_U.S._states_and_territories_by_Human_Development_Index"
url_income <- "https://en.wikipedia.org/wiki/List_of_U.S._states_and_territories_by_income"
url_pvi <- "https://raw.githubusercontent.com/fivethirtyeight/data/master/partisan-lean/fivethirtyeight_partisan_lean_STATES.csv"

parse_percent <- function(x) {
  parse_number(x)/100
}

edu <-
  read_html(url_edu) %>%
  html_node(".wikitable") %>%
  html_table(fill = TRUE) %>%
  clean_names("snake") %>%
  select(-contains("rank")) %>%
  as_tibble() %>%
  set_names(c("state", "high_school", "bach_degree", "grad_degree")) %>%
  mutate_at(vars(-state), parse_percent) %>%
  mutate_at(vars(state), abbrev_state)

life <-
  read_html(url_life) %>%
  html_node(".wikitable") %>%
  html_table(fill = TRUE) %>%
  clean_names() %>%
  na_if("–") %>%
  select(2, 3) %>%
  as_tibble() %>%
  set_names(c("state", "life_expect")) %>%
  mutate_at(vars(state), abbrev_state)

gdp <-
  read_html(url_gdp) %>%
  html_nodes(".wikitable") %>%
  html_table(fill = TRUE) %>%
  extract2(2) %>%
  select(state = `State`, gdp = `2018`) %>%
  as_tibble() %>%
  mutate_at(vars(-state), parse_number) %>%
  mutate_at(vars(state), abbrev_state)

hdi <-
  read_html(url_hdi) %>%
  html_node(".wikitable") %>%
  html_table(fill = TRUE) %>%
  select(state = 2, hdi = 3) %>%
  as_tibble() %>%
  mutate_at(vars(state), abbrev_state) %>%
  filter(state %in% state.abb)

income <-
  read_html(url_income) %>%
  html_nodes(".wikitable") %>%
  html_table(fill = TRUE) %>%
  extract2(2) %>%
  select(state = 2, ind_income = 3, house_income = 4) %>%
  as_tibble() %>%
  mutate_at(vars(-state), parse_number) %>%
  mutate_at(vars(state), abbrev_state)

pvi <-
  read_csv(url_pvi) %>%
  separate(
    col = pvi_538,
    into = c("party", "lean"),
    convert = TRUE
  ) %>%
  mutate(
    state = abbrev_state(state),
    lean = if_else(party == "D", lean * -1L, lean)
  )

left_join(pvi, hdi) %>%
  ggplot(aes(x = lean, y = hdi)) +
  geom_smooth(method = "lm", color = "black", se = FALSE) +
  geom_label(aes(label = state, fill = lean), size = 5) +
  scale_fill_distiller(palette = "RdBu", guide = FALSE) +
  labs(
    x = "Partisan Lean Index (538)",
    y = "Human Development Index (UNDP)"
  )

ggsave("~/Pictures/hdi_pvi.png", dpi = "retina", height = 5, width = 9)

left_join(pvi, edu) %>%
  pivot_longer(cols = c("high_school", "bach_degree", "grad_degree")) %>%
  mutate(name = as_factor(name)) %>%
  ggplot(aes(x = lean, y = value)) +
  facet_wrap(~name, scales = "free_y") +
  geom_smooth(method = "lm", color = "black", se = FALSE) +
  geom_label(aes(label = state, fill = lean), size = 5) +
  scale_fill_distiller(palette = "RdBu", guide = FALSE) +
  scale_y_continuous(labels = scales::percent) +
  labs(
    x = "Partisan Lean Index (538)",
    y = "Percent Attained"
  )

ggsave("~/Pictures/pvi_devel.png", dpi = "retina", height = 5, width = 9)


left_join(pvi, life) %>%
  ggplot(aes(x = lean, y = life_expect)) +
  geom_smooth(method = "lm", color = "black", se = FALSE) +
  geom_label(aes(label = state, fill = lean), size = 5) +
  scale_fill_distiller(palette = "RdBu", guide = FALSE) +
  labs(
    x = "Partisan Lean Index (538)",
    y = "Life Expectancy"
  )

ggsave("~/Pictures/pvi_life.png", dpi = "retina", height = 5, width = 9)


left_join(pvi, gdp) %>%
  ggplot(aes(x = lean, y = gdp)) +
  geom_smooth(method = "lm", color = "black", se = FALSE) +
  geom_label(aes(label = state, fill = lean), size = 5) +
  scale_fill_distiller(palette = "RdBu", guide = FALSE) +
  scale_y_continuous(labels = scales::dollar) +
  labs(
    x = "Partisan Lean Index (538)",
    y = "GDP Per Capita"
  )

ggsave("~/Pictures/pvi_gdp.png", dpi = "retina", height = 5, width = 9)

left_join(pvi, income) %>%
  pivot_longer(cols = c("ind_income", "house_income")) %>%
  ggplot(aes(x = lean, y = value)) +
  facet_wrap(~name) +
  geom_smooth(method = "lm", color = "black", se = FALSE) +
  geom_label(aes(label = state, fill = lean), size = 5) +
  scale_fill_distiller(palette = "RdBu", guide = FALSE) +
  scale_y_continuous(labels = scales::dollar) +
  labs(
    x = "Partisan Lean Index (538)",
    y = "Income"
  )

ggsave("~/Pictures/pvi_income.png", dpi = "retina", height = 5, width = 9)

url_charity <- "https://www.forbes.com/sites/karstenstrauss/2017/12/04/the-most-and-least-charitable-states-in-the-u-s-in-2017/#4d6150720703"

charity <-
  read_html(url_charity) %>%
  html_node("table") %>%
  html_table(fill = TRUE) %>%
  select(state = State, score = `Total Score`) %>%
  as_tibble() %>%
  mutate_at(vars(state), abbrev_state)

left_join(pvi, charity) %>%
  ggplot(aes(x = lean, y = score)) +
  geom_smooth(method = "lm", color = "black", se = FALSE) +
  geom_label(aes(label = state, fill = lean), size = 5) +
  scale_fill_distiller(palette = "RdBu", guide = FALSE) +
  labs(
    x = "Partisan Lean Index (538)",
    y = "Charitability Score"
  )