Census Commute Shares in TidyCensus

census_commute.R

# Create census commute share in Portland
if(!require(pacman)) install.packages('pacman')
pacman::p_load(dplyr, tidycensus)

# https://data.census.gov/cedsci/table?q=means%20transportation&g=1600000US4159000&tid=ACSDT1Y2021.B08006
trans_table <- 'B08006' 

# note, no data for 2020, the 2005 data is corrupt :
#   the non-auto counts are off
years <- c(2006:2019, 2021)

# download data in a loop over each year
acs_raw <- lapply(years, function(year) get_acs('place'
  , state = '41', table = trans_table
  , year = year, survey = 'acs1', output = 'wide') %>% 
    mutate(Year = year)
)

# variable definitions here: https://api.census.gov/data/2019/acs/acs1/variables.html
acs <- acs_raw %>% 
  bind_rows() %>% 
  transmute(Name = NAME, Year, Total = B08006_001E, Auto = B08006_002E
    , Public_Transit = B08006_008E, Bike = B08006_014E
    , Walk = B08006_015E, Other = B08006_016E
    , Work_From_Home = B08006_017E)

# verify that the sum of categories equal the total
all.equal(acs[[3]], rowSums(acs[4:9]))

# output to a csv
write.csv(acs, 'census_commute_city.csv', row.names = FALSE)

census_commute_city.csv