Obtaining 2010 US Decenial Census Response Rates with 2010 boundaries

Question

This question may be out-of-scope, but I'm not sure where else to ask it.

I am trying to obtain response rates to the 2010 US census with 2010 boundaries. The census provides an overview of getting response rates here but it seems to use 2020 boundaries. I tried using this call to obtain all census tracts in Ohio: https://api.census.gov/data/2010/dec/responserate?get=NAME,GEO_ID,FSRR2010&for=tract:*&in=state:39

Is there a way to get the response rates using 2010 boundaries? Also, can this be done in tidycensus?

Answer 1

Here's how you'd do it in tidycensus:

library(tidycensus)

ohio_response <- get_decennial(
  geography = "tract",
  variables = "FSRR2010",
  state = "OH",
  sumfile = "responserate",
  year = 2010
)

# A tibble: 3,157 × 4
   GEOID       NAME                                     variable value
   <chr>       <chr>                                    <chr>    <dbl>
 1 39001770100 Census Tract 7701, Adams County, Ohio    FSRR2010  71.5
 2 39001770200 Census Tract 7702, Adams County, Ohio    FSRR2010  60.9
 3 39001770301 Census Tract 7703.01, Adams County, Ohio FSRR2010  69  
 4 39001770302 Census Tract 7703.02, Adams County, Ohio FSRR2010  64.4
 5 39001770400 Census Tract 7704, Adams County, Ohio    FSRR2010  72.1
 6 39001770500 Census Tract 7705, Adams County, Ohio    FSRR2010  66.6
 7 39001770600 Census Tract 7706, Adams County, Ohio    FSRR2010  63.1
 8 39003010100 Census Tract 101, Allen County, Ohio     FSRR2010  83.1
 9 39003010200 Census Tract 102, Allen County, Ohio     FSRR2010  79.4
10 39003010300 Census Tract 103, Allen County, Ohio     FSRR2010  79.4
# … with 3,147 more rows

You are correct however that the rates are aggregated to 2020 Census tracts. Unless Census fixes (I assume that was you posting on the Census Slack channel?), you would have to join to shapes from tigris (as geometry = TRUE won't work correctly here) then interpolate back to 2010 tracts using a method like those described here.