R/harmonize_vars.R
bl_harmonize_vars.RdCreates estimates for race/ethnicity variables for blocks in 2020, based on
the 2010 Census (bl_download_2010_vars()) and 2019 ACS
(bl_download_acs_vars()) data.
bl_harmonize_vars(block_d, census_d, acs_d)
| block_d | the output of |
|---|---|
| census_d | the output of |
| acs_d | the output of |
A data frame of 2020 block-level estimates for the variables in
census_d.
The racial breakdown of total and voting-age population are proportionally adjusted within block groups to match the ACS block-group percentages. Then these percentages are applied to the newly estimated block populations. Finally, the block race populations are rescaled so that the block populations match the 2020 estimates. This is essentially a 1-round iterative proportional fitting (IPF) procedure.