This repository contains code and data to estimate expected deaths and excess deaths in the United States in 2020 and 2021 by county and month. This model estimates that there were about 467602 excess deaths in the United States in 2020 (excluding January and February) and 417400 in 2021 (excluding December), for a total of 885002.
Estimated excess deaths are available at
/blob/main/results/united_states_county_monthly_excess_deaths_estimates.csv.
Excess mortality for the months in which total deaths exceeded expected deaths are highlighted in blue.
-
CDC WONDER all-cause, all-ages deaths by county and month
https://wonder.cdc.gov/ucd-icd10.html
These public-use historical mortality data were obtained for years 2011-2020 and used to estimate the expected deaths model. 2020 data were also compared against estimated expected deaths to calculate excess deaths. Because these data have an unusual and specific license, they are not included as part of the repository, but can be freely downloaded from https://wonder.cdc.gov/ucd-icd10.html.
Note: County-months wither fewer than 10 deaths are censored in the source data.
-
CDC WONDER provisional and partial all-cause deaths and deaths involving COVID-19 by county and month
https://wonder.cdc.gov/mcd-icd10-provisional.html
Example query: https://wonder.cdc.gov/controller/saved/D176/D273F449
These public-use provisional and partial mortality data were obtained for 2021 and and compared to estimated expected deaths to calculate excess deaths. Because these data have an unusual and specific license, they are not included as part of the repository, but can be freely downloaded from https://wonder.cdc.gov/ucd-icd10.html.
Note: Values for county-months with fewer than 10 deaths are suppressed in these data.
-
US Census Population Estimates
Population estimates for 2011-2020 were retrieved from the US Census Bureau Population Estimates, because these data are not yet included in the Census Bureau Data API. 2020 population estimates were used for 2021.
Note: This product uses Census Bureau data but is not endorsed or certified by the Census Bureau.
The excess deaths model used began as an adaptation of The Economist’s
excess mortality
model,
but has since diverged. For details on model selection and performance,
see
/blob/main/results/README.md.
Given the large number of counties in the United States (over 3000), a linear mixed model with county, county set, and census division as random grouping factors was used to make estimation tractable. These random grouping factors enable each county, county set, and state to have their own intercepts in the final model.
More precisely, total deaths per day was regressed on:
-
county population (z-scored)
-
years since 2011
-
month of the year (fixed grouping factor)
-
county (random grouping factor nested within county set)
-
county set (random grouping factor nested within state)
-
state (random grouping factor)
This model can be expressed by the equation:
The model object is available at
/blob/main/results/united_states_county_monthly_model.RDS.
The model estimates for each observation, including fitted values and
residuals, are available at
/blob/main/results/united_states_county_monthly_fitted_deaths_per_day_estimate.csv.
Note that these estimates are only y-hat directly from the initial
model, not predicted values on the full dataset, which are
available
here..
For more details on model selection and performance, see
/blob/main/results/README.md.
Because the COVID-19 pandemic only began in the United States partway through March, 2020, we can evaluate model performance by examining concordance of predicted and observed deaths in the counties that were unaffected by COVID-19 in March, 2020.
Observed and expected deaths in March, 2020 are highly correlated, r = 0.998. As can be seen from the following scatterplot, total deaths (y-axis) tended to exceed expected deaths (x-axis). Unsurprisingly, some of the counties that were hardest hit early in the pandemic are among those with total deaths that most diverge from expected deaths.
-
Excess deaths estimates and fitted LMM are available in
/results/. -
Code to fit the model and estimate excess deaths are available in
/code/. -
Data used in modeling are available in
/data/, when possible.
The code contained in this repository are available under the MIT License, and the data generated by this code are licensed under the Creative Commons Attribution 4.0 International License.
sessionInfo()
## R version 4.0.4 (2021-02-15)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows 10 x64 (build 19042)
##
## Matrix products: default
##
## locale:
## [1] LC_COLLATE=English_United States.1252
## [2] LC_CTYPE=English_United States.1252
## [3] LC_MONETARY=English_United States.1252
## [4] LC_NUMERIC=C
## [5] LC_TIME=English_United States.1252
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] lmerTest_3.1-3 lme4_1.1-27.1 Matrix_1.3-2
## [4] texPreview_1.5 tinytex_0.34 knitr_1.31
## [7] equatiomatic_0.3.0.9000 lubridate_1.7.10 data.table_1.14.2
## [10] ggh4x_0.2.0 forcats_0.5.1 stringr_1.4.0
## [13] dplyr_1.0.5 purrr_0.3.4 readr_2.0.2
## [16] tidyr_1.1.3 tibble_3.1.0 ggplot2_3.3.5
## [19] tidyverse_1.3.1
##
## loaded via a namespace (and not attached):
## [1] rdocsyntax_0.4.1.9000 minqa_1.2.4 colorspace_2.0-0
## [4] ellipsis_0.3.1 rprojroot_2.0.2 base64enc_0.1-3
## [7] fs_1.5.0 rstudioapi_0.13 farver_2.1.0
## [10] ggrepel_0.9.1 fansi_0.4.2 xml2_1.3.2
## [13] splines_4.0.4 cachem_1.0.4 jsonlite_1.7.2
## [16] nloptr_1.2.2.2 broom_0.7.9 dbplyr_2.1.1
## [19] png_0.1-7 broom.mixed_0.2.6 shiny_1.6.0
## [22] clipr_0.7.1 compiler_4.0.4 httr_1.4.2
## [25] backports_1.2.1 assertthat_0.2.1 fastmap_1.1.0
## [28] cli_3.0.1 svgPanZoom_0.3.4 later_1.2.0
## [31] htmltools_0.5.1.1 tools_4.0.4 coda_0.19-4
## [34] gtable_0.3.0 glue_1.4.2 reshape2_1.4.4
## [37] V8_3.4.0 Rcpp_1.0.7 cellranger_1.1.0
## [40] vctrs_0.3.7 pdftools_2.3.1 nlme_3.1-152
## [43] xfun_0.26 rvest_1.0.0 mime_0.10
## [46] lifecycle_1.0.0 MASS_7.3-53 scales_1.1.1
## [49] hms_1.0.0 promises_1.2.0.1 rematch2_2.1.2
## [52] TMB_1.7.20 yaml_2.2.1 curl_4.3
## [55] memoise_2.0.0 stringi_1.5.3 highr_0.8
## [58] desc_1.3.0 boot_1.3-26 rlang_0.4.10
## [61] pkgconfig_2.0.3 evaluate_0.14 lattice_0.20-41
## [64] labeling_0.4.2 tidyselect_1.1.0 here_1.0.1
## [67] plyr_1.8.6 magrittr_2.0.1 R6_2.5.0
## [70] magick_2.7.3 generics_0.1.0 DBI_1.1.1
## [73] pillar_1.6.0 haven_2.3.1 whisker_0.4
## [76] withr_2.4.2 mgcv_1.8-33 modelr_0.1.8
## [79] crayon_1.4.1 utf8_1.2.1 tzdb_0.1.2
## [82] rmarkdown_2.7 grid_4.0.4 readxl_1.3.1
## [85] qpdf_1.1 reprex_2.0.0 digest_0.6.27
## [88] xtable_1.8-4 httpuv_1.6.0 numDeriv_2016.8-1.1
## [91] details_0.2.1 munsell_0.5.0 viridisLite_0.4.0
## [94] askpass_1.1