Data publication and documentation for a project on states' reallocation of prisoner populations in the U.S. 2020 Census.
To understand the differences between the original census counts, states' department of corrections counts, and the ammended counts for states that reallocate prisoner residences to addresses before incarceration.
Guiding Questions: Why are there gaps between the U.S. Census and DOC population counts for incarcerated facilities? How pervasive is this issue?
Questions to Answer:
- How often do these gaps occur?
- How large are the discrepancies between the Census and DOC Numbers?
- Where is the breakdown in process?
- Are there specific facility types that more often have these issues such as the size, location, demographics of a facility?
- What is the effect?
- Who is harmed?
- Exploring_DOC_numbers_vs_Census.ipynb looks at the U.S. Census 2020 PL 94-171 data summary files for states at the block level compared to the Department of Corrections' Counts
- census_vs_adjusted_incarcerated_populations.ipynb merges the U.S. Census 2020 PL 94-171 data summary files for states at the block level with the states' adjusted population files into one dataframe per state
- CO-Identifying-Census-Blocks.ipynb is a notebook used for identifying census blocks in Colorado where DOC has incarcerated populations in facilities.
- NJ-adjusted-census-merged.ipynb merges New Jersey's 21 adjusted datasets for each county into one dataset for importing into the ccensus_vs_adjusted_incarcerated_populations notebook.
- VA-SQLite-adjusted-census-exploration.ipynb is an exploration of Virginia's Adjusted Census files using SQLite to identify important tables for importing into the census_vs_adjusted_incarcerated_populations notebook.
- Redistricting Data Hub: Many state census and ammended census files were downloaded with the hub's API
- Individual state redistricting sites
TBD
TBD