|
View: |
Part 1: Document Description
|
|
Citation |
|
|---|---|
|
Title: |
Replication data for "Evaluating Bias and Noise Induced by the U.S. Census Bureau's Privacy Protection Methods" |
|
Identification Number: |
doi:10.7910/DVN/TMIN3H |
|
Distributor: |
Harvard Dataverse |
|
Date of Distribution: |
2024-02-01 |
|
Version: |
3 |
|
Bibliographic Citation: |
Kenny, Christopher; McCartan, Cory; Kuriwaki, Shiro; Simko, Tyler; Imai, Kosuke, 2024, "Replication data for "Evaluating Bias and Noise Induced by the U.S. Census Bureau's Privacy Protection Methods"", https://doi.org/10.7910/DVN/TMIN3H, Harvard Dataverse, V3 |
|
Citation |
|
|
Title: |
Replication data for "Evaluating Bias and Noise Induced by the U.S. Census Bureau's Privacy Protection Methods" |
|
Identification Number: |
doi:10.7910/DVN/TMIN3H |
|
Authoring Entity: |
Kenny, Christopher (Harvard University) |
|
McCartan, Cory (New York University) |
|
|
Kuriwaki, Shiro (Yale University) |
|
|
Simko, Tyler (Harvard University) |
|
|
Imai, Kosuke (Harvard University) |
|
|
Distributor: |
Harvard Dataverse |
|
Access Authority: |
Kenny, Christopher |
|
Depositor: |
Kenny, Christopher |
|
Date of Deposit: |
2024-01-29 |
|
Holdings Information: |
https://doi.org/10.7910/DVN/TMIN3H |
|
Study Scope |
|
|
Keywords: |
Law, Social Sciences |
|
Abstract: |
The United States Census Bureau faces a difficult trade-off between the accuracy of Census statistics and the protection of individual information. We conduct the first independent evaluation of bias and noise induced by the Bureau's two main disclosure avoidance systems: the TopDown algorithm employed for the 2020 Census and the swapping algorithm implemented for the three previous Censuses. Our evaluation leverages the Noisy Measure File (NMF) as well as two independent runs of the TopDown algorithm applied to the 2010 decennial Census. We find that the NMF contains too much noise to be directly useful, especially for Hispanic and multiracial populations. TopDown's post-processing dramatically reduces the NMF noise and produces data whose accuracy is similar to that of swapping. While the estimated errors for both TopDown and swapping algorithms are generally no greater than other sources of Census error, they can be relatively substantial for geographies with small total populations. |
|
Methodology and Processing |
|
|
Sources Statement |
|
|
Data Access |
|
|
Other Study Description Materials |
|
|
Related Studies |
|
|
Abowd et al., "2010 Census Production Settings Redistricting Data (P.L. 94-171) Demonstration Noisy Measurement File (2023-04-03)", Harvard Dataverse. https://doi.org/10.7910/DVN/1OR2A6 |
|
|
Related Publications |
|
|
Citation |
|
|
Title: |
Evaluating bias and noise induced by the U.S. Census Bureau’s privacy protection methods, <b><i>Science Advances</i></b>, 2024. Vol 10, Issue 18. |
|
Identification Number: |
10.1126/sciadv.adl2524 |
|
Bibliographic Citation: |
Evaluating bias and noise induced by the U.S. Census Bureau’s privacy protection methods, <b><i>Science Advances</i></b>, 2024. Vol 10, Issue 18. |
|
Citation |
|
|
Title: |
“Evaluating Bias and Noise Induced by the U.S. Census Bureau’s Privacy Protection Methods.” arXiv, October 7, 2023. https://doi.org/10.48550/arXiv.2306.07521. |
|
Identification Number: |
2306.07521 |
|
Bibliographic Citation: |
“Evaluating Bias and Noise Induced by the U.S. Census Bureau’s Privacy Protection Methods.” arXiv, October 7, 2023. https://doi.org/10.48550/arXiv.2306.07521. |
|
Label: |
00_README.md |
|
Text: |
Top-level README |
|
Notes: |
text/markdown |
|
Label: |
d_natl.rds |
|
Text: |
Cleaned rectangular dataset for most figures and tables |
|
Notes: |
application/octet-stream |
|
Label: |
00_run_r_geo.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
00_run_r_proc.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
README.md |
|
Notes: |
text/markdown |
|
Label: |
build_aian.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
build_gaf.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
gaf.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
README.md |
|
Notes: |
text/markdown |
|
Label: |
01_build_nmf_all.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
02_build_state_dbs.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
build_db_de.sql |
|
Notes: |
application/x-sql |
|
Label: |
build_duckdb.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
build_nmf.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
build_pl.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
build_ppmf21.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
build_ppmf23.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
fix_nmf_mis.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
nmf.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
nmf_baf.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
nmf_make_agg.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
nmf_stats.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
nmf_workflow.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
pl.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
README.md |
|
Text: |
Readme for R_proc |
|
Notes: |
text/markdown |
|
Label: |
utils.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
00_setup.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
01_create_db.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
02_fig-01_pop-error.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
03_fig-02_racepop-rmse.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
04_fig-03-04_pop-bias-rmse_by-totpop.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
05_fig-05_noisy-reg_lpme.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
06-figs-bias-rmse_appendix.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
07_tda_corr.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
README.md |
|
Notes: |
text/markdown |
|
Label: |
bafs.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
build_db_parq.sql |
|
Notes: |
application/x-sql |
|
Label: |
census_api.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
consts.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
db.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
geo.R |
|
Notes: |
type/x-r-syntax |
|
Label: |
stats.R |
|
Notes: |
type/x-r-syntax |