Save Area Based Measures

save_absms.R

library(tidyverse)
library(tidycensus)
library(magrittr)

variables_dict <-
  tibble::tribble(
  ~var,            ~desc,
  "B03002_001",  'total_race',    "total pop for race/non-hispanic estimate",
  "B03002_003",  'white_nh',      "white non-hispanic pop",
  "B25014_001",  'total_crowd',   "total for occupants by room by owner/renter",
  "B25014_005",  'crowd1_own',    "owner occupied occupants by room - 1.01-1.5",
  "B25014_006",  'crowd2_own',    "owner occupied occupants by room - 1.51-2",
  "B25014_007",  'crowd3_own',    "owner occupied occupants by room - 2.01+",
  "B25014_011",  'crowd1_rent',   "renter occupied occupants by room - 1.01-1.5",
  "B25014_012",  'crowd2_rent',   "renter occupied occupants by room - 1.51-2",
  "B25014_013",  'crowd3_rent',   "renter occupied occupants by room - 2.01+",
  "B05010_001",  'total_pov',     "total for poverty estimate",
  "B05010_002",  'under_pov',     "under poverty line",
  "B19001_001",  'hhinc_total',   "total pop for household income estimates",
  "B19001A_014", 'hhinc_w_1',     "white n.h. pop with household income $100 000 to $124 999",
  "B19001A_015", 'hhinc_w_2',     "white n.h. pop with household income $125k-149 999k",
  "B19001A_016", 'hhinc_w_3',     "white n.h. pop with household income $150k-199 999k",
  "B19001A_017", 'hhinc_w_4',     "white n.h. pop with household income $200k+",
  "B19001B_002", 'hhinc_b_1',     "black pop with household income <$10k",
  "B19001B_003", 'hhinc_b_2',     "black pop with household income $10k-14 999k",
  "B19001B_004", 'hhinc_b_3',     "black pop with household income $15k-19 999k",
  "B19001B_005", 'hhinc_b_4',     "black pop with household income $20k-24 999k",
  "B11017_001",  'total_multigen',"total homes for multigenerational estimate",
  "B11017_002",  'multigen_hhs',  "multigenerational households"
 )

# create a named vector, rename_vars, which has elements that are the
# acs variables we request and convenient, human readable names.
#
# used below twice: to request the data, and to rename the data
rename_vars <- setNames(variables_dict$orig_var, variables_dict$var)

# request acs data
us_county_data <- get_acs(geography = "county",
  variables = variables_dict$orig_var)

# pivot to a wide format for renaming, dropping the margin of error data
us_county_data_pivotted  <- us_county_data %>% as.data.frame() %>% select(-moe) %>%
  pivot_wider(names_from = variable, values_from = estimate)

# rename the columns using our rename_vars
us_county_data_pivotted %<>% rename(!!rename_vars)

# construct our area-based-socioeconomic/demographic-measures
us_county_data_pivotted %<>% mutate(
  # pct people of color
  pct_poc = (total_race - white_nh) / total_race * 100,
  # pct crowding
  pct_crowd = (crowd1_own + crowd2_own + crowd3_own + crowd1_rent + crowd2_rent + crowd3_rent) / total_crowd * 100,
  # pct poverty
  pct_pov = under_pov / total_pov * 100,
  # racialized economic segregation
  wnhb_inc_ice = ((hhinc_w_1 + hhinc_w_2 + hhinc_w_3 + hhinc_w_4) -
    (hhinc_b_1 + hhinc_b_2 + hhinc_b_3 + hhinc_b_4)) / hhinc_total,
  # percentage of households which are multigenerational
  pct_multigen = multigen_hhs / total_multigen * 100
)

# save data
saveRDS(
  us_county_data_pivotted,
  'us_county_absms.rds'
)