Data and code to accompany paper "Polling bias and undecided voter allocations: US Presidential elections, 2004 - 2016"

Cite as: Bon, J. J., Ballard, T. and Baffour, B. (2019), Polling bias and undecided voter allocations: US presidential elections, 2004–2016. Journal of the Royal Statistical Society, Series A (Statistics in Society), 182(2): 467-493. doi:10.1111/rssa.12414.

Link to paper on journal website

Link to paper on arXiv

• Corresponding Author: Joshua J Bon
• Web: http://joshuabon.com
• Github: https://github.com/bonStats/undecided-voters-us-pres-elections/

Directory description

• Top: contains all .R code for running models and reproducing plots and tables in the paper
• data/: Contain the state-level polling and voting data
• stan_models/: contains .stan code that define (and estimate by HMC) the models
• fitted_models/: Folder for fitted .stan models and summary outputs from those models
• eda/: Contains example(s) of exploratory data analysis, including Figure 1 in the paper.

The fitted_models/ folder may be empty due to large size of files. Run the models and posterior calculations to populate.

Data description

Two data sets are in the data/ directory. Please cite the above paper if using the dataset(s).

Election results: us-pres-state-voting-2004-2016.*

This data contains the election results for the 2004, 2008, 2012, and 2016 US presidential election by state. It is in both .csv and .rds (tibble) format. It has columns:

• state: State names and Washington D.C. (e.g. "washington-d-c")
• year: Presidential election year: 2004, 2008, 2012, 2016
• state_year: Concatenation of state and year: (e.g. washington-d-c_2016)
• state_year_id: Unique integer ids for state_year
• Dem_vote: Vote percentage won by Democratic candidate (0-100)
• Rep_vote: Vote percentage won by Republican candidate (0-100)
• short_state: Two character state id (e.g. DC)
• result_margin6: Category for margin of voting result. Strong Dem. win (margin > 6%), Strong Rep. win (margin > 6%), or close margin (margin < 6%)
• year_id: Unique integer ids for year

Pre-election polls: us-pres-state-polling-2004-2016.*

This data contains the election polls for the 2004, 2008, 2012, and 2016 US presidential election by state. It is in both .csv and .rds (tibble) format. It has columns:

• Dem_poll: Polled percentage support for Democratic candidate (0-100)
• Rep_poll: Polled percentage support for Republican candidate (0-100)
• Undecided: Polled percentage of undecided voters (0-100 and NA)
• sample_size: Reported sample size of poll
• mean_days_to_election: Number of days until election, measured as mean of start and end date of poll
• start_days_to_election: Number of days until election, measured from start date of poll
• end_days_to_election: Number of days until election, measured from end date of poll
• state: State names and Washington D.C. (e.g. "washington-d-c")
• year: Presidential election year: 2004, 2008, 2012, 2016
• state_year: Concatenation of state and year: (e.g. washington-d-c_2016)
• pollster: Original name of polling agency or agencies
• state_year_id: Unique integer ids for state_year
• pollster2: Cleaned name of polling agency or agencies
• year_id: Unique integer ids for year
• result_margin6: Category for margin of voting result. Strong Dem. win (margin > 6%), Strong Rep. win (margin > 6%), or close margin (margin < 6%)
• rmargin_year: result_margin6 concatenated with year
• rmargin_year_id: Unique integer ids for rmargin_year
• pollster_grp: Further cleaned and grouped polling agencies or institutes
• pollster_id: Unique integer ids for pollster_grp

R code description

• state-polls-original-model.R: Fit original SRGG model
• state-polls-extended-model-proportionate.R: Fit extended SRGG model with baseline proportionate split of undecided voters
• state-polls-extended-model-even.R: Fit extended SRGG model with baseline even split of undecided voters
• posterior-calcs.R: Calculate additional posterior quantities from the model
• paper-outputs.R: Reproduce all plots and tables for the paper

Session Info

sessionInfo()
#> R version 3.5.1 (2018-07-02)
#> Platform: x86_64-apple-darwin15.6.0 (64-bit)
#> Running under: macOS High Sierra 10.13.6
#> 
#> Matrix products: default
#> BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
#> LAPACK: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRlapack.dylib
#> 
#> locale:
#> [1] en_AU.UTF-8/en_AU.UTF-8/en_AU.UTF-8/C/en_AU.UTF-8/en_AU.UTF-8
#> 
#> attached base packages:
#> [1] parallel  stats     graphics  grDevices utils     datasets  methods   base     
#> 
#> other attached packages:
#>  [1] shinystan_2.5.0    shiny_1.1.0        gtools_3.8.1       plyr_1.8.4        
#>  [5] rstan_2.17.3       StanHeaders_2.17.2 rv_2.3.2           stringr_1.3.1     
#>  [9] scales_1.0.0       ggplot2_3.0.0      bindrcpp_0.2.2     dplyr_0.7.6       
#> 
#> loaded via a namespace (and not attached):
#>  [1]  Rcpp_0.12.18      lattice_0.20-35   zoo_1.8-4         assertthat_0.2.0  digest_0.6.16
#>  [6]  utf8_1.1.4        mime_0.5          R6_2.2.2          ggridges_0.5.0    stats4_3.5.1
#>  [11] colourpicker_1.0  pillar_1.3.0      rlang_0.2.2       lazyeval_0.2.1    miniUI_0.1.1.1
#>  [16] rstudioapi_0.7    DT_0.4            shinythemes_1.1.1 shinyjs_1.0       devtools_1.13.6
#>  [21] readr_1.1.1       htmlwidgets_1.2   igraph_1.2.2      munsell_0.5.0     compiler_3.5.1
#>  [26] httpuv_1.4.5      pkgconfig_2.0.2   base64enc_0.1-3   htmltools_0.3.6   tidyselect_0.2.4 
#>  [31] tibble_1.4.2      gridExtra_2.3     threejs_0.3.1     fansi_0.3.0       crayon_1.3.4     
#>  [36] withr_2.1.2       later_0.7.4       grid_3.5.1        xtable_1.8-3      gtable_0.2.0     
#>  [41] magrittr_1.5      cli_1.0.0         stringi_1.2.4     reshape2_1.4.3    promises_1.0.1   
#>  [46] dygraphs_1.1.1.6  xts_0.11-1        tools_3.5.1       glue_1.3.0        markdown_0.8     
#>  [51] purrr_0.2.5       hms_0.4.2         crosstalk_1.0.0   rsconnect_0.8.8   yaml_2.2.0       
#>  [56] inline_0.3.15     colorspace_1.3-2  bayesplot_1.6.0   memoise_1.1.0     bindr_0.1.1