1 Installation and Activation

We reconstruct here the analysis for the article:
“National Leadership in Time of A Pandemic: Mechanisms that Create a Gendered Gap” by Freund and Shomer

To conduct the reconstruction, follow these steps:

  1. The analysis was conducted using R version 4.1.3 (2022-03-10) and RStudio 2022.07.2+576 release for Windows.
  2. Extract the contents of the provided reconstruction zip file into a directory on your PC.
  3. Make the directory the RStudio working directory (usung the Files menu or the R command setwd).
  4. The file wl_load_packages.R contains the list of required R packages.
  5. The data directory includes the research data file wldb_weekly.rds as well as additional support tables.
  6. Open the file wl_reconstruct.Rmd in RStudio.
  7. Select the Knit Document command from the File menu to run the reconstruction.
    • All graphs and plots will be saved in the plots sub-directory.
    • All tables will be saved in LaTeX format in the tables sub-directory.
    • The reconstruction results are provided in the HTML file wl_reconstruct.html.
    • For reference, the original reconstruction results file wl_reconstruct_orginal.html can be used.
  8. When reviewing the reconstruction results HTML file:
    • To view the code, click on the ‘Code’ buttons on the right side of the page.
    • To view the interpretation of the regression results, click on the ‘Interpretation’ button just below the regression table.

1.1 Initialization

load external packages and Source helper functions Note: All routines starting with ’

source("wl_load_packages.R")
source("wl_utils.R")
source("wl_graph.R") #wl_ggsave()
source("wl_regression.R")
source("wl_variables.R") # wl_desc_summary() 
source("wl_align_waves.R")
source("wl_countries.R")
source("wl_smartAgg.R") # Aggregate multiple columns using different functions
source("wl_wldb.R") # wl_db()
source("wl_delay.R") #wl_delay_fields()
source("wl_pal.R") #wl_pal_color()
source("wl_reg_report.R") #wl_reg_interpret()
source("wl_learning_curve.R")
source("wl_polls_ml_olm.R")
source("wl_public_support.R")
source("wl_leaders.R") # wl_leaders_si_table
source("wl_data_dictionary.R") # wl_si_data_sources
source("wl_coef_summary.R") # wl_coef_and_ci_table, wl_coef_and_ci_plot
source("wl_timing.R") # wl_ld_prepare_cox

options(dplyr.summarise.inform = FALSE)

Set the default theme & font for plots and tables

loadfonts(device = "win", quiet=TRUE)
theme_set(theme_bw() +
            theme(text = element_text(family = "LM Roman 10", size = 12)))

Load the data

wl_get_wldb("Weekly")
wl_get_data_dictionary()
wl_delay_fields()
# Set scandinavia and Northern EU fields
wl_geography_fields()
# Add order information for the dependent variables 
# (for continuous learning curve analysis)
wl_learning_order()

2 Descriptive Information

2.1 Country Leaders table

country_leaders.tex country_leaders.rtf

wl_leaders_si_table("tables/country_leaders.tex")
#> <div id="mipvvheyho" style="padding-left:0px;padding-right:0px;padding-top:10px;padding-bottom:10px;overflow-x:auto;overflow-y:auto;width:auto;height:auto;">
#>   <style>#mipvvheyho table {
#>   font-family: system-ui, 'Segoe UI', Roboto, Helvetica, Arial, sans-serif, 'Apple Color Emoji', 'Segoe UI Emoji', 'Segoe UI Symbol', 'Noto Color Emoji';
#>   -webkit-font-smoothing: antialiased;
#>   -moz-osx-font-smoothing: grayscale;
#> }
#> 
#> #mipvvheyho thead, #mipvvheyho tbody, #mipvvheyho tfoot, #mipvvheyho tr, #mipvvheyho td, #mipvvheyho th {
#>   border-style: none;
#> }
#> 
#> #mipvvheyho p {
#>   margin: 0;
#>   padding: 0;
#> }
#> 
#> #mipvvheyho .gt_table {
#>   display: table;
#>   border-collapse: collapse;
#>   line-height: normal;
#>   margin-left: auto;
#>   margin-right: auto;
#>   color: #333333;
#>   font-size: 14px;
#>   font-weight: normal;
#>   font-style: normal;
#>   background-color: #FFFFFF;
#>   width: auto;
#>   border-top-style: solid;
#>   border-top-width: 0px;
#>   border-top-color: #A8A8A8;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 0px;
#>   border-bottom-color: #A8A8A8;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_caption {
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#> }
#> 
#> #mipvvheyho .gt_title {
#>   color: #333333;
#>   font-size: 125%;
#>   font-weight: initial;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-color: #FFFFFF;
#>   border-bottom-width: 0;
#> }
#> 
#> #mipvvheyho .gt_subtitle {
#>   color: #333333;
#>   font-size: 85%;
#>   font-weight: initial;
#>   padding-top: 3px;
#>   padding-bottom: 5px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-color: #FFFFFF;
#>   border-top-width: 0;
#> }
#> 
#> #mipvvheyho .gt_heading {
#>   background-color: #FFFFFF;
#>   text-align: center;
#>   border-bottom-color: #FFFFFF;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_bottom_border {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_col_headings {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_col_heading {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 6px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   overflow-x: hidden;
#> }
#> 
#> #mipvvheyho .gt_column_spanner_outer {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   padding-top: 0;
#>   padding-bottom: 0;
#>   padding-left: 4px;
#>   padding-right: 4px;
#> }
#> 
#> #mipvvheyho .gt_column_spanner_outer:first-child {
#>   padding-left: 0;
#> }
#> 
#> #mipvvheyho .gt_column_spanner_outer:last-child {
#>   padding-right: 0;
#> }
#> 
#> #mipvvheyho .gt_column_spanner {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 5px;
#>   overflow-x: hidden;
#>   display: inline-block;
#>   width: 100%;
#> }
#> 
#> #mipvvheyho .gt_spanner_row {
#>   border-bottom-style: hidden;
#> }
#> 
#> #mipvvheyho .gt_group_heading {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   text-align: left;
#> }
#> 
#> #mipvvheyho .gt_empty_group_heading {
#>   padding: 0.5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: middle;
#> }
#> 
#> #mipvvheyho .gt_from_md > :first-child {
#>   margin-top: 0;
#> }
#> 
#> #mipvvheyho .gt_from_md > :last-child {
#>   margin-bottom: 0;
#> }
#> 
#> #mipvvheyho .gt_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   margin: 10px;
#>   border-top-style: solid;
#>   border-top-width: 1px;
#>   border-top-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   overflow-x: hidden;
#> }
#> 
#> #mipvvheyho .gt_stub {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #mipvvheyho .gt_stub_row_group {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   vertical-align: top;
#> }
#> 
#> #mipvvheyho .gt_row_group_first td {
#>   border-top-width: 2px;
#> }
#> 
#> #mipvvheyho .gt_row_group_first th {
#>   border-top-width: 2px;
#> }
#> 
#> #mipvvheyho .gt_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #mipvvheyho .gt_first_summary_row {
#>   border-top-style: solid;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_first_summary_row.thick {
#>   border-top-width: 2px;
#> }
#> 
#> #mipvvheyho .gt_last_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_grand_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #mipvvheyho .gt_first_grand_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-style: double;
#>   border-top-width: 6px;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_last_grand_summary_row_top {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: double;
#>   border-bottom-width: 6px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_striped {
#>   background-color: rgba(128, 128, 128, 0.05);
#> }
#> 
#> #mipvvheyho .gt_table_body {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_footnotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_footnote {
#>   margin: 0px;
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #mipvvheyho .gt_sourcenotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #mipvvheyho .gt_sourcenote {
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #mipvvheyho .gt_left {
#>   text-align: left;
#> }
#> 
#> #mipvvheyho .gt_center {
#>   text-align: center;
#> }
#> 
#> #mipvvheyho .gt_right {
#>   text-align: right;
#>   font-variant-numeric: tabular-nums;
#> }
#> 
#> #mipvvheyho .gt_font_normal {
#>   font-weight: normal;
#> }
#> 
#> #mipvvheyho .gt_font_bold {
#>   font-weight: bold;
#> }
#> 
#> #mipvvheyho .gt_font_italic {
#>   font-style: italic;
#> }
#> 
#> #mipvvheyho .gt_super {
#>   font-size: 65%;
#> }
#> 
#> #mipvvheyho .gt_footnote_marks {
#>   font-size: 75%;
#>   vertical-align: 0.4em;
#>   position: initial;
#> }
#> 
#> #mipvvheyho .gt_asterisk {
#>   font-size: 100%;
#>   vertical-align: 0;
#> }
#> 
#> #mipvvheyho .gt_indent_1 {
#>   text-indent: 5px;
#> }
#> 
#> #mipvvheyho .gt_indent_2 {
#>   text-indent: 10px;
#> }
#> 
#> #mipvvheyho .gt_indent_3 {
#>   text-indent: 15px;
#> }
#> 
#> #mipvvheyho .gt_indent_4 {
#>   text-indent: 20px;
#> }
#> 
#> #mipvvheyho .gt_indent_5 {
#>   text-indent: 25px;
#> }
#> </style>
#>   <table class="gt_table" data-quarto-disable-processing="false" data-quarto-bootstrap="false">
#>   <thead>
#>     <tr class="gt_col_headings">
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Leader Gender">Leader Gender</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Leader Name">Leader Name</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Start in Position">Start in Position</th>
#>     </tr>
#>   </thead>
#>   <tbody class="gt_table_body">
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Argentina">Argentina</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Argentina  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Argentina  pm_name" class="gt_row gt_left">Alberto Fernandez</td>
#> <td headers="Argentina  pm_start_in_position" class="gt_row gt_left">2019-12-10</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Australia">Australia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Australia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Australia  pm_name" class="gt_row gt_left">Scott Morrison</td>
#> <td headers="Australia  pm_start_in_position" class="gt_row gt_left">2018-08-24</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Austria">Austria</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Austria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Austria  pm_name" class="gt_row gt_left">Sebastian Kurz</td>
#> <td headers="Austria  pm_start_in_position" class="gt_row gt_left">2020-01-07</td></tr>
#>     <tr><td headers="Austria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Austria  pm_name" class="gt_row gt_left">Karl Nehammer</td>
#> <td headers="Austria  pm_start_in_position" class="gt_row gt_left">2021-12-06</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Bahamas">Bahamas</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Bahamas  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bahamas  pm_name" class="gt_row gt_left">Hubert Minnis</td>
#> <td headers="Bahamas  pm_start_in_position" class="gt_row gt_left">2017-05-11</td></tr>
#>     <tr><td headers="Bahamas  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bahamas  pm_name" class="gt_row gt_left">Philip Davis</td>
#> <td headers="Bahamas  pm_start_in_position" class="gt_row gt_left">2021-09-17</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Barbados">Barbados</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Barbados  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Barbados  pm_name" class="gt_row gt_left">Mia Mottley</td>
#> <td headers="Barbados  pm_start_in_position" class="gt_row gt_left">2018-05-25</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Belgium">Belgium</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Belgium  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Belgium  pm_name" class="gt_row gt_left">Sophie Wilmes</td>
#> <td headers="Belgium  pm_start_in_position" class="gt_row gt_left">2019-10-27</td></tr>
#>     <tr><td headers="Belgium  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Belgium  pm_name" class="gt_row gt_left">Alexander De Croo</td>
#> <td headers="Belgium  pm_start_in_position" class="gt_row gt_left">2020-10-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Belize">Belize</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Belize  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Belize  pm_name" class="gt_row gt_left">Dean Barrow</td>
#> <td headers="Belize  pm_start_in_position" class="gt_row gt_left">2008-02-08</td></tr>
#>     <tr><td headers="Belize  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Belize  pm_name" class="gt_row gt_left">Johnny Briceno</td>
#> <td headers="Belize  pm_start_in_position" class="gt_row gt_left">2020-11-12</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Botswana">Botswana</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Botswana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Botswana  pm_name" class="gt_row gt_left">Mokgweetsi Masisi</td>
#> <td headers="Botswana  pm_start_in_position" class="gt_row gt_left">2018-04-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Brazil">Brazil</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Brazil  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Brazil  pm_name" class="gt_row gt_left">Jair Bolsonaro</td>
#> <td headers="Brazil  pm_start_in_position" class="gt_row gt_left">2019-01-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Bulgaria">Bulgaria</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Boyko Borisov</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2017-05-04</td></tr>
#>     <tr><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Stefan Yanev</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2021-05-12</td></tr>
#>     <tr><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Kiril Petkov</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2021-12-13</td></tr>
#>     <tr><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Galab Donev</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2022-08-02</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Canada">Canada</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Canada  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Canada  pm_name" class="gt_row gt_left">Justin Trudeau</td>
#> <td headers="Canada  pm_start_in_position" class="gt_row gt_left">2015-11-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Cape Verde">Cape Verde</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Cape Verde  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Cape Verde  pm_name" class="gt_row gt_left">Ulisses Correia e Silva</td>
#> <td headers="Cape Verde  pm_start_in_position" class="gt_row gt_left">2016-04-22</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Chile">Chile</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Chile  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Chile  pm_name" class="gt_row gt_left">Sebastian Pinera</td>
#> <td headers="Chile  pm_start_in_position" class="gt_row gt_left">2018-03-11</td></tr>
#>     <tr><td headers="Chile  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Chile  pm_name" class="gt_row gt_left">Gabriel Boric</td>
#> <td headers="Chile  pm_start_in_position" class="gt_row gt_left">2022-03-11</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Costa Rica">Costa Rica</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Costa Rica  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Costa Rica  pm_name" class="gt_row gt_left">Carlos Alvarado Quesada</td>
#> <td headers="Costa Rica  pm_start_in_position" class="gt_row gt_left">2018-05-08</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Croatia">Croatia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Croatia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Croatia  pm_name" class="gt_row gt_left">Andrej Plenkovic</td>
#> <td headers="Croatia  pm_start_in_position" class="gt_row gt_left">2016-10-19</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Cyprus">Cyprus</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Cyprus  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Cyprus  pm_name" class="gt_row gt_left">Nicos Anastasiades</td>
#> <td headers="Cyprus  pm_start_in_position" class="gt_row gt_left">2013-02-18</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Czechia">Czechia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Czechia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Czechia  pm_name" class="gt_row gt_left">Andrej Babis</td>
#> <td headers="Czechia  pm_start_in_position" class="gt_row gt_left">2017-12-06</td></tr>
#>     <tr><td headers="Czechia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Czechia  pm_name" class="gt_row gt_left">Petr Fiala</td>
#> <td headers="Czechia  pm_start_in_position" class="gt_row gt_left">2021-11-28</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Denmark">Denmark</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Denmark  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Denmark  pm_name" class="gt_row gt_left">Mette Frederiksen</td>
#> <td headers="Denmark  pm_start_in_position" class="gt_row gt_left">2019-06-27</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Estonia">Estonia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Estonia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Estonia  pm_name" class="gt_row gt_left">Juri Ratas</td>
#> <td headers="Estonia  pm_start_in_position" class="gt_row gt_left">2019-04-29</td></tr>
#>     <tr><td headers="Estonia  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Estonia  pm_name" class="gt_row gt_left">Kaja Kallas</td>
#> <td headers="Estonia  pm_start_in_position" class="gt_row gt_left">2021-01-26</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Finland">Finland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Finland  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Finland  pm_name" class="gt_row gt_left">Sanna Marin</td>
#> <td headers="Finland  pm_start_in_position" class="gt_row gt_left">2019-12-10</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="France">France</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="France  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="France  pm_name" class="gt_row gt_left">Emmanuel Macron</td>
#> <td headers="France  pm_start_in_position" class="gt_row gt_left">2017-05-14</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Germany">Germany</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Germany  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Germany  pm_name" class="gt_row gt_left">Angela Merkel</td>
#> <td headers="Germany  pm_start_in_position" class="gt_row gt_left">2005-11-22</td></tr>
#>     <tr><td headers="Germany  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Germany  pm_name" class="gt_row gt_left">Olaf Scholz</td>
#> <td headers="Germany  pm_start_in_position" class="gt_row gt_left">2021-12-08</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Ghana">Ghana</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Ghana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Ghana  pm_name" class="gt_row gt_left">Nana Akufo-Addo</td>
#> <td headers="Ghana  pm_start_in_position" class="gt_row gt_left">2017-01-07</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Greece">Greece</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Greece  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Greece  pm_name" class="gt_row gt_left">Kyriakos Mitsotakis</td>
#> <td headers="Greece  pm_start_in_position" class="gt_row gt_left">2019-07-08</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Guyana">Guyana</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Guyana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Guyana  pm_name" class="gt_row gt_left">David Granger</td>
#> <td headers="Guyana  pm_start_in_position" class="gt_row gt_left">2015-05-16</td></tr>
#>     <tr><td headers="Guyana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Guyana  pm_name" class="gt_row gt_left">Irfaan Ali</td>
#> <td headers="Guyana  pm_start_in_position" class="gt_row gt_left">2020-08-02</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Hungary">Hungary</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Hungary  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Hungary  pm_name" class="gt_row gt_left">Viktor Orban</td>
#> <td headers="Hungary  pm_start_in_position" class="gt_row gt_left">2010-05-29</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Iceland">Iceland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Iceland  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Iceland  pm_name" class="gt_row gt_left">Katrin Jakobsdottir</td>
#> <td headers="Iceland  pm_start_in_position" class="gt_row gt_left">2017-11-30</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Ireland">Ireland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Ireland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Ireland  pm_name" class="gt_row gt_left">Leo Varadkar</td>
#> <td headers="Ireland  pm_start_in_position" class="gt_row gt_left">2017-06-14</td></tr>
#>     <tr><td headers="Ireland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Ireland  pm_name" class="gt_row gt_left">Micheal Martin</td>
#> <td headers="Ireland  pm_start_in_position" class="gt_row gt_left">2020-06-27</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Israel">Israel</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Israel  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Israel  pm_name" class="gt_row gt_left">Benjamin Netanyau</td>
#> <td headers="Israel  pm_start_in_position" class="gt_row gt_left">2009-03-31</td></tr>
#>     <tr><td headers="Israel  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Israel  pm_name" class="gt_row gt_left">Naftali Bennett</td>
#> <td headers="Israel  pm_start_in_position" class="gt_row gt_left">2021-06-13</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Italy">Italy</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Italy  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Italy  pm_name" class="gt_row gt_left">Giuseppe Conte</td>
#> <td headers="Italy  pm_start_in_position" class="gt_row gt_left">2018-06-01</td></tr>
#>     <tr><td headers="Italy  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Italy  pm_name" class="gt_row gt_left">Mario Draghi</td>
#> <td headers="Italy  pm_start_in_position" class="gt_row gt_left">2021-02-13</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Jamaica">Jamaica</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Jamaica  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Jamaica  pm_name" class="gt_row gt_left">Andrew Holness</td>
#> <td headers="Jamaica  pm_start_in_position" class="gt_row gt_left">2016-03-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Japan">Japan</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Japan  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Japan  pm_name" class="gt_row gt_left">Shinzo Abe</td>
#> <td headers="Japan  pm_start_in_position" class="gt_row gt_left">2012-12-26</td></tr>
#>     <tr><td headers="Japan  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Japan  pm_name" class="gt_row gt_left">Yoshihide Suga</td>
#> <td headers="Japan  pm_start_in_position" class="gt_row gt_left">2020-09-16</td></tr>
#>     <tr><td headers="Japan  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Japan  pm_name" class="gt_row gt_left">Fumio Kishida</td>
#> <td headers="Japan  pm_start_in_position" class="gt_row gt_left">2021-10-04</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Latvia">Latvia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Latvia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Latvia  pm_name" class="gt_row gt_left">Krisjanis Kariņs</td>
#> <td headers="Latvia  pm_start_in_position" class="gt_row gt_left">2019-01-23</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Lithuania">Lithuania</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Lithuania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Lithuania  pm_name" class="gt_row gt_left">Saulius Skvernelis</td>
#> <td headers="Lithuania  pm_start_in_position" class="gt_row gt_left">2016-11-22</td></tr>
#>     <tr><td headers="Lithuania  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Lithuania  pm_name" class="gt_row gt_left">Ingrida Simonyte</td>
#> <td headers="Lithuania  pm_start_in_position" class="gt_row gt_left">2020-12-11</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Luxembourg">Luxembourg</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Luxembourg  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Luxembourg  pm_name" class="gt_row gt_left">Xavier Bettel</td>
#> <td headers="Luxembourg  pm_start_in_position" class="gt_row gt_left">2013-12-04</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Malta">Malta</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Malta  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Malta  pm_name" class="gt_row gt_left">Robert Abela</td>
#> <td headers="Malta  pm_start_in_position" class="gt_row gt_left">2020-01-13</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Mauritius">Mauritius</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Mauritius  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Mauritius  pm_name" class="gt_row gt_left">Pravind Jugnauth</td>
#> <td headers="Mauritius  pm_start_in_position" class="gt_row gt_left">2017-01-23</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Mexico">Mexico</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Mexico  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Mexico  pm_name" class="gt_row gt_left">Andres Manuel Lopez Obrador</td>
#> <td headers="Mexico  pm_start_in_position" class="gt_row gt_left">2018-12-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Netherlands">Netherlands</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Netherlands  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Netherlands  pm_name" class="gt_row gt_left">Mark Rutte</td>
#> <td headers="Netherlands  pm_start_in_position" class="gt_row gt_left">2010-10-14</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="New Zealand">New Zealand</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="New Zealand  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="New Zealand  pm_name" class="gt_row gt_left">Jacinda Ardern</td>
#> <td headers="New Zealand  pm_start_in_position" class="gt_row gt_left">2017-10-26</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Norway">Norway</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Norway  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Norway  pm_name" class="gt_row gt_left">Erna Solberg</td>
#> <td headers="Norway  pm_start_in_position" class="gt_row gt_left">2013-10-16</td></tr>
#>     <tr><td headers="Norway  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Norway  pm_name" class="gt_row gt_left">Jonas Gahr Store</td>
#> <td headers="Norway  pm_start_in_position" class="gt_row gt_left">2021-10-14</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Panama">Panama</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Panama  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Panama  pm_name" class="gt_row gt_left">Laurentino Cortizo</td>
#> <td headers="Panama  pm_start_in_position" class="gt_row gt_left">2019-07-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Poland">Poland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Poland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Poland  pm_name" class="gt_row gt_left">Mateusz Morawiecki</td>
#> <td headers="Poland  pm_start_in_position" class="gt_row gt_left">2017-12-11</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Portugal">Portugal</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Portugal  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Portugal  pm_name" class="gt_row gt_left">Antonio Costa</td>
#> <td headers="Portugal  pm_start_in_position" class="gt_row gt_left">2015-11-26</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Romania">Romania</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Romania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Romania  pm_name" class="gt_row gt_left">Ludovic Orban</td>
#> <td headers="Romania  pm_start_in_position" class="gt_row gt_left">2019-11-04</td></tr>
#>     <tr><td headers="Romania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Romania  pm_name" class="gt_row gt_left">Florin Citu</td>
#> <td headers="Romania  pm_start_in_position" class="gt_row gt_left">2020-12-07</td></tr>
#>     <tr><td headers="Romania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Romania  pm_name" class="gt_row gt_left">Nicolae Ciuca</td>
#> <td headers="Romania  pm_start_in_position" class="gt_row gt_left">2021-11-25</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Sao Tome and Principe">Sao Tome and Principe</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Sao Tome and Principe  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Sao Tome and Principe  pm_name" class="gt_row gt_left">Jorge Bom Jesus</td>
#> <td headers="Sao Tome and Principe  pm_start_in_position" class="gt_row gt_left">2018-12-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Slovakia">Slovakia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Slovakia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovakia  pm_name" class="gt_row gt_left">Peter Pellegrini</td>
#> <td headers="Slovakia  pm_start_in_position" class="gt_row gt_left">2018-03-22</td></tr>
#>     <tr><td headers="Slovakia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovakia  pm_name" class="gt_row gt_left">Igor Matovic</td>
#> <td headers="Slovakia  pm_start_in_position" class="gt_row gt_left">2020-03-21</td></tr>
#>     <tr><td headers="Slovakia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovakia  pm_name" class="gt_row gt_left">Eduard Heger</td>
#> <td headers="Slovakia  pm_start_in_position" class="gt_row gt_left">2021-04-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Slovenia">Slovenia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Slovenia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovenia  pm_name" class="gt_row gt_left">Marjan Sarec</td>
#> <td headers="Slovenia  pm_start_in_position" class="gt_row gt_left">2018-09-13</td></tr>
#>     <tr><td headers="Slovenia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovenia  pm_name" class="gt_row gt_left">Janez Jansa</td>
#> <td headers="Slovenia  pm_start_in_position" class="gt_row gt_left">2020-03-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="South Africa">South Africa</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="South Africa  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="South Africa  pm_name" class="gt_row gt_left">Cyril Ramaphosa</td>
#> <td headers="South Africa  pm_start_in_position" class="gt_row gt_left">2018-02-15</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="South Korea">South Korea</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="South Korea  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="South Korea  pm_name" class="gt_row gt_left">Moon Jae-in</td>
#> <td headers="South Korea  pm_start_in_position" class="gt_row gt_left">2017-05-10</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Spain">Spain</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Spain  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Spain  pm_name" class="gt_row gt_left">Pedro Sanchez</td>
#> <td headers="Spain  pm_start_in_position" class="gt_row gt_left">2018-06-02</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Suriname">Suriname</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Suriname  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Suriname  pm_name" class="gt_row gt_left">Desi Bouterse</td>
#> <td headers="Suriname  pm_start_in_position" class="gt_row gt_left">2010-08-12</td></tr>
#>     <tr><td headers="Suriname  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Suriname  pm_name" class="gt_row gt_left">Chan Santokhi</td>
#> <td headers="Suriname  pm_start_in_position" class="gt_row gt_left">2020-07-16</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Sweden">Sweden</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Sweden  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Sweden  pm_name" class="gt_row gt_left">Stefan Lofven</td>
#> <td headers="Sweden  pm_start_in_position" class="gt_row gt_left">2014-10-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Switzerland">Switzerland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Switzerland  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Switzerland  pm_name" class="gt_row gt_left">Simonetta Sommaruga</td>
#> <td headers="Switzerland  pm_start_in_position" class="gt_row gt_left">2020-01-01</td></tr>
#>     <tr><td headers="Switzerland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Switzerland  pm_name" class="gt_row gt_left">Guy Parmelin</td>
#> <td headers="Switzerland  pm_start_in_position" class="gt_row gt_left">2021-01-01</td></tr>
#>     <tr><td headers="Switzerland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Switzerland  pm_name" class="gt_row gt_left">Ignazio Cassis</td>
#> <td headers="Switzerland  pm_start_in_position" class="gt_row gt_left">2022-01-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Taiwan">Taiwan</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Taiwan  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Taiwan  pm_name" class="gt_row gt_left">Tsai Ing-wen</td>
#> <td headers="Taiwan  pm_start_in_position" class="gt_row gt_left">2016-05-20</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Trinidad and Tobago">Trinidad and Tobago</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Trinidad and Tobago  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Trinidad and Tobago  pm_name" class="gt_row gt_left">Keith Rowley</td>
#> <td headers="Trinidad and Tobago  pm_start_in_position" class="gt_row gt_left">2015-09-09</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Turkey">Turkey</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Turkey  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Turkey  pm_name" class="gt_row gt_left">Recep Tayyip Erdogan</td>
#> <td headers="Turkey  pm_start_in_position" class="gt_row gt_left">2003-02-09</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="United Kingdom">United Kingdom</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="United Kingdom  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="United Kingdom  pm_name" class="gt_row gt_left">Boris Johnson</td>
#> <td headers="United Kingdom  pm_start_in_position" class="gt_row gt_left">2019-07-24</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="United States">United States</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="United States  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="United States  pm_name" class="gt_row gt_left">Donald Trump</td>
#> <td headers="United States  pm_start_in_position" class="gt_row gt_left">2017-01-20</td></tr>
#>     <tr><td headers="United States  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="United States  pm_name" class="gt_row gt_left">Joe Biden</td>
#> <td headers="United States  pm_start_in_position" class="gt_row gt_left">2021-01-20</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Uruguay">Uruguay</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Uruguay  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Uruguay  pm_name" class="gt_row gt_left">Tabare Vazquez</td>
#> <td headers="Uruguay  pm_start_in_position" class="gt_row gt_left">2015-03-01</td></tr>
#>     <tr><td headers="Uruguay  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Uruguay  pm_name" class="gt_row gt_left">Luis Lacalle Pou</td>
#> <td headers="Uruguay  pm_start_in_position" class="gt_row gt_left">2020-03-01</td></tr>
#>   </tbody>
#>   
#>   
#> </table>
#> </div>
wl_leaders_si_table("tables/country_leaders.rtf")
#> <div id="kpsenyqemm" style="padding-left:0px;padding-right:0px;padding-top:10px;padding-bottom:10px;overflow-x:auto;overflow-y:auto;width:auto;height:auto;">
#>   <style>#kpsenyqemm table {
#>   font-family: system-ui, 'Segoe UI', Roboto, Helvetica, Arial, sans-serif, 'Apple Color Emoji', 'Segoe UI Emoji', 'Segoe UI Symbol', 'Noto Color Emoji';
#>   -webkit-font-smoothing: antialiased;
#>   -moz-osx-font-smoothing: grayscale;
#> }
#> 
#> #kpsenyqemm thead, #kpsenyqemm tbody, #kpsenyqemm tfoot, #kpsenyqemm tr, #kpsenyqemm td, #kpsenyqemm th {
#>   border-style: none;
#> }
#> 
#> #kpsenyqemm p {
#>   margin: 0;
#>   padding: 0;
#> }
#> 
#> #kpsenyqemm .gt_table {
#>   display: table;
#>   border-collapse: collapse;
#>   line-height: normal;
#>   margin-left: auto;
#>   margin-right: auto;
#>   color: #333333;
#>   font-size: 14px;
#>   font-weight: normal;
#>   font-style: normal;
#>   background-color: #FFFFFF;
#>   width: auto;
#>   border-top-style: solid;
#>   border-top-width: 0px;
#>   border-top-color: #A8A8A8;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 0px;
#>   border-bottom-color: #A8A8A8;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_caption {
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#> }
#> 
#> #kpsenyqemm .gt_title {
#>   color: #333333;
#>   font-size: 125%;
#>   font-weight: initial;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-color: #FFFFFF;
#>   border-bottom-width: 0;
#> }
#> 
#> #kpsenyqemm .gt_subtitle {
#>   color: #333333;
#>   font-size: 85%;
#>   font-weight: initial;
#>   padding-top: 3px;
#>   padding-bottom: 5px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-color: #FFFFFF;
#>   border-top-width: 0;
#> }
#> 
#> #kpsenyqemm .gt_heading {
#>   background-color: #FFFFFF;
#>   text-align: center;
#>   border-bottom-color: #FFFFFF;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_bottom_border {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_col_headings {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_col_heading {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 6px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   overflow-x: hidden;
#> }
#> 
#> #kpsenyqemm .gt_column_spanner_outer {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   padding-top: 0;
#>   padding-bottom: 0;
#>   padding-left: 4px;
#>   padding-right: 4px;
#> }
#> 
#> #kpsenyqemm .gt_column_spanner_outer:first-child {
#>   padding-left: 0;
#> }
#> 
#> #kpsenyqemm .gt_column_spanner_outer:last-child {
#>   padding-right: 0;
#> }
#> 
#> #kpsenyqemm .gt_column_spanner {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 5px;
#>   overflow-x: hidden;
#>   display: inline-block;
#>   width: 100%;
#> }
#> 
#> #kpsenyqemm .gt_spanner_row {
#>   border-bottom-style: hidden;
#> }
#> 
#> #kpsenyqemm .gt_group_heading {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   text-align: left;
#> }
#> 
#> #kpsenyqemm .gt_empty_group_heading {
#>   padding: 0.5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: middle;
#> }
#> 
#> #kpsenyqemm .gt_from_md > :first-child {
#>   margin-top: 0;
#> }
#> 
#> #kpsenyqemm .gt_from_md > :last-child {
#>   margin-bottom: 0;
#> }
#> 
#> #kpsenyqemm .gt_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   margin: 10px;
#>   border-top-style: solid;
#>   border-top-width: 1px;
#>   border-top-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   overflow-x: hidden;
#> }
#> 
#> #kpsenyqemm .gt_stub {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #kpsenyqemm .gt_stub_row_group {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   vertical-align: top;
#> }
#> 
#> #kpsenyqemm .gt_row_group_first td {
#>   border-top-width: 2px;
#> }
#> 
#> #kpsenyqemm .gt_row_group_first th {
#>   border-top-width: 2px;
#> }
#> 
#> #kpsenyqemm .gt_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #kpsenyqemm .gt_first_summary_row {
#>   border-top-style: solid;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_first_summary_row.thick {
#>   border-top-width: 2px;
#> }
#> 
#> #kpsenyqemm .gt_last_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_grand_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #kpsenyqemm .gt_first_grand_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-style: double;
#>   border-top-width: 6px;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_last_grand_summary_row_top {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: double;
#>   border-bottom-width: 6px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_striped {
#>   background-color: rgba(128, 128, 128, 0.05);
#> }
#> 
#> #kpsenyqemm .gt_table_body {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_footnotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_footnote {
#>   margin: 0px;
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #kpsenyqemm .gt_sourcenotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #kpsenyqemm .gt_sourcenote {
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #kpsenyqemm .gt_left {
#>   text-align: left;
#> }
#> 
#> #kpsenyqemm .gt_center {
#>   text-align: center;
#> }
#> 
#> #kpsenyqemm .gt_right {
#>   text-align: right;
#>   font-variant-numeric: tabular-nums;
#> }
#> 
#> #kpsenyqemm .gt_font_normal {
#>   font-weight: normal;
#> }
#> 
#> #kpsenyqemm .gt_font_bold {
#>   font-weight: bold;
#> }
#> 
#> #kpsenyqemm .gt_font_italic {
#>   font-style: italic;
#> }
#> 
#> #kpsenyqemm .gt_super {
#>   font-size: 65%;
#> }
#> 
#> #kpsenyqemm .gt_footnote_marks {
#>   font-size: 75%;
#>   vertical-align: 0.4em;
#>   position: initial;
#> }
#> 
#> #kpsenyqemm .gt_asterisk {
#>   font-size: 100%;
#>   vertical-align: 0;
#> }
#> 
#> #kpsenyqemm .gt_indent_1 {
#>   text-indent: 5px;
#> }
#> 
#> #kpsenyqemm .gt_indent_2 {
#>   text-indent: 10px;
#> }
#> 
#> #kpsenyqemm .gt_indent_3 {
#>   text-indent: 15px;
#> }
#> 
#> #kpsenyqemm .gt_indent_4 {
#>   text-indent: 20px;
#> }
#> 
#> #kpsenyqemm .gt_indent_5 {
#>   text-indent: 25px;
#> }
#> </style>
#>   <table class="gt_table" data-quarto-disable-processing="false" data-quarto-bootstrap="false">
#>   <thead>
#>     <tr class="gt_col_headings">
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Leader Gender">Leader Gender</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Leader Name">Leader Name</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Start in Position">Start in Position</th>
#>     </tr>
#>   </thead>
#>   <tbody class="gt_table_body">
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Argentina">Argentina</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Argentina  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Argentina  pm_name" class="gt_row gt_left">Alberto Fernandez</td>
#> <td headers="Argentina  pm_start_in_position" class="gt_row gt_left">2019-12-10</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Australia">Australia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Australia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Australia  pm_name" class="gt_row gt_left">Scott Morrison</td>
#> <td headers="Australia  pm_start_in_position" class="gt_row gt_left">2018-08-24</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Austria">Austria</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Austria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Austria  pm_name" class="gt_row gt_left">Sebastian Kurz</td>
#> <td headers="Austria  pm_start_in_position" class="gt_row gt_left">2020-01-07</td></tr>
#>     <tr><td headers="Austria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Austria  pm_name" class="gt_row gt_left">Karl Nehammer</td>
#> <td headers="Austria  pm_start_in_position" class="gt_row gt_left">2021-12-06</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Bahamas">Bahamas</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Bahamas  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bahamas  pm_name" class="gt_row gt_left">Hubert Minnis</td>
#> <td headers="Bahamas  pm_start_in_position" class="gt_row gt_left">2017-05-11</td></tr>
#>     <tr><td headers="Bahamas  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bahamas  pm_name" class="gt_row gt_left">Philip Davis</td>
#> <td headers="Bahamas  pm_start_in_position" class="gt_row gt_left">2021-09-17</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Barbados">Barbados</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Barbados  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Barbados  pm_name" class="gt_row gt_left">Mia Mottley</td>
#> <td headers="Barbados  pm_start_in_position" class="gt_row gt_left">2018-05-25</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Belgium">Belgium</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Belgium  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Belgium  pm_name" class="gt_row gt_left">Sophie Wilmes</td>
#> <td headers="Belgium  pm_start_in_position" class="gt_row gt_left">2019-10-27</td></tr>
#>     <tr><td headers="Belgium  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Belgium  pm_name" class="gt_row gt_left">Alexander De Croo</td>
#> <td headers="Belgium  pm_start_in_position" class="gt_row gt_left">2020-10-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Belize">Belize</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Belize  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Belize  pm_name" class="gt_row gt_left">Dean Barrow</td>
#> <td headers="Belize  pm_start_in_position" class="gt_row gt_left">2008-02-08</td></tr>
#>     <tr><td headers="Belize  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Belize  pm_name" class="gt_row gt_left">Johnny Briceno</td>
#> <td headers="Belize  pm_start_in_position" class="gt_row gt_left">2020-11-12</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Botswana">Botswana</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Botswana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Botswana  pm_name" class="gt_row gt_left">Mokgweetsi Masisi</td>
#> <td headers="Botswana  pm_start_in_position" class="gt_row gt_left">2018-04-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Brazil">Brazil</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Brazil  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Brazil  pm_name" class="gt_row gt_left">Jair Bolsonaro</td>
#> <td headers="Brazil  pm_start_in_position" class="gt_row gt_left">2019-01-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Bulgaria">Bulgaria</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Boyko Borisov</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2017-05-04</td></tr>
#>     <tr><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Stefan Yanev</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2021-05-12</td></tr>
#>     <tr><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Kiril Petkov</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2021-12-13</td></tr>
#>     <tr><td headers="Bulgaria  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Bulgaria  pm_name" class="gt_row gt_left">Galab Donev</td>
#> <td headers="Bulgaria  pm_start_in_position" class="gt_row gt_left">2022-08-02</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Canada">Canada</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Canada  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Canada  pm_name" class="gt_row gt_left">Justin Trudeau</td>
#> <td headers="Canada  pm_start_in_position" class="gt_row gt_left">2015-11-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Cape Verde">Cape Verde</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Cape Verde  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Cape Verde  pm_name" class="gt_row gt_left">Ulisses Correia e Silva</td>
#> <td headers="Cape Verde  pm_start_in_position" class="gt_row gt_left">2016-04-22</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Chile">Chile</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Chile  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Chile  pm_name" class="gt_row gt_left">Sebastian Pinera</td>
#> <td headers="Chile  pm_start_in_position" class="gt_row gt_left">2018-03-11</td></tr>
#>     <tr><td headers="Chile  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Chile  pm_name" class="gt_row gt_left">Gabriel Boric</td>
#> <td headers="Chile  pm_start_in_position" class="gt_row gt_left">2022-03-11</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Costa Rica">Costa Rica</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Costa Rica  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Costa Rica  pm_name" class="gt_row gt_left">Carlos Alvarado Quesada</td>
#> <td headers="Costa Rica  pm_start_in_position" class="gt_row gt_left">2018-05-08</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Croatia">Croatia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Croatia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Croatia  pm_name" class="gt_row gt_left">Andrej Plenkovic</td>
#> <td headers="Croatia  pm_start_in_position" class="gt_row gt_left">2016-10-19</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Cyprus">Cyprus</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Cyprus  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Cyprus  pm_name" class="gt_row gt_left">Nicos Anastasiades</td>
#> <td headers="Cyprus  pm_start_in_position" class="gt_row gt_left">2013-02-18</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Czechia">Czechia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Czechia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Czechia  pm_name" class="gt_row gt_left">Andrej Babis</td>
#> <td headers="Czechia  pm_start_in_position" class="gt_row gt_left">2017-12-06</td></tr>
#>     <tr><td headers="Czechia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Czechia  pm_name" class="gt_row gt_left">Petr Fiala</td>
#> <td headers="Czechia  pm_start_in_position" class="gt_row gt_left">2021-11-28</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Denmark">Denmark</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Denmark  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Denmark  pm_name" class="gt_row gt_left">Mette Frederiksen</td>
#> <td headers="Denmark  pm_start_in_position" class="gt_row gt_left">2019-06-27</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Estonia">Estonia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Estonia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Estonia  pm_name" class="gt_row gt_left">Juri Ratas</td>
#> <td headers="Estonia  pm_start_in_position" class="gt_row gt_left">2019-04-29</td></tr>
#>     <tr><td headers="Estonia  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Estonia  pm_name" class="gt_row gt_left">Kaja Kallas</td>
#> <td headers="Estonia  pm_start_in_position" class="gt_row gt_left">2021-01-26</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Finland">Finland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Finland  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Finland  pm_name" class="gt_row gt_left">Sanna Marin</td>
#> <td headers="Finland  pm_start_in_position" class="gt_row gt_left">2019-12-10</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="France">France</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="France  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="France  pm_name" class="gt_row gt_left">Emmanuel Macron</td>
#> <td headers="France  pm_start_in_position" class="gt_row gt_left">2017-05-14</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Germany">Germany</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Germany  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Germany  pm_name" class="gt_row gt_left">Angela Merkel</td>
#> <td headers="Germany  pm_start_in_position" class="gt_row gt_left">2005-11-22</td></tr>
#>     <tr><td headers="Germany  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Germany  pm_name" class="gt_row gt_left">Olaf Scholz</td>
#> <td headers="Germany  pm_start_in_position" class="gt_row gt_left">2021-12-08</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Ghana">Ghana</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Ghana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Ghana  pm_name" class="gt_row gt_left">Nana Akufo-Addo</td>
#> <td headers="Ghana  pm_start_in_position" class="gt_row gt_left">2017-01-07</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Greece">Greece</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Greece  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Greece  pm_name" class="gt_row gt_left">Kyriakos Mitsotakis</td>
#> <td headers="Greece  pm_start_in_position" class="gt_row gt_left">2019-07-08</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Guyana">Guyana</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Guyana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Guyana  pm_name" class="gt_row gt_left">David Granger</td>
#> <td headers="Guyana  pm_start_in_position" class="gt_row gt_left">2015-05-16</td></tr>
#>     <tr><td headers="Guyana  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Guyana  pm_name" class="gt_row gt_left">Irfaan Ali</td>
#> <td headers="Guyana  pm_start_in_position" class="gt_row gt_left">2020-08-02</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Hungary">Hungary</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Hungary  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Hungary  pm_name" class="gt_row gt_left">Viktor Orban</td>
#> <td headers="Hungary  pm_start_in_position" class="gt_row gt_left">2010-05-29</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Iceland">Iceland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Iceland  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Iceland  pm_name" class="gt_row gt_left">Katrin Jakobsdottir</td>
#> <td headers="Iceland  pm_start_in_position" class="gt_row gt_left">2017-11-30</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Ireland">Ireland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Ireland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Ireland  pm_name" class="gt_row gt_left">Leo Varadkar</td>
#> <td headers="Ireland  pm_start_in_position" class="gt_row gt_left">2017-06-14</td></tr>
#>     <tr><td headers="Ireland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Ireland  pm_name" class="gt_row gt_left">Micheal Martin</td>
#> <td headers="Ireland  pm_start_in_position" class="gt_row gt_left">2020-06-27</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Israel">Israel</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Israel  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Israel  pm_name" class="gt_row gt_left">Benjamin Netanyau</td>
#> <td headers="Israel  pm_start_in_position" class="gt_row gt_left">2009-03-31</td></tr>
#>     <tr><td headers="Israel  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Israel  pm_name" class="gt_row gt_left">Naftali Bennett</td>
#> <td headers="Israel  pm_start_in_position" class="gt_row gt_left">2021-06-13</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Italy">Italy</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Italy  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Italy  pm_name" class="gt_row gt_left">Giuseppe Conte</td>
#> <td headers="Italy  pm_start_in_position" class="gt_row gt_left">2018-06-01</td></tr>
#>     <tr><td headers="Italy  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Italy  pm_name" class="gt_row gt_left">Mario Draghi</td>
#> <td headers="Italy  pm_start_in_position" class="gt_row gt_left">2021-02-13</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Jamaica">Jamaica</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Jamaica  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Jamaica  pm_name" class="gt_row gt_left">Andrew Holness</td>
#> <td headers="Jamaica  pm_start_in_position" class="gt_row gt_left">2016-03-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Japan">Japan</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Japan  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Japan  pm_name" class="gt_row gt_left">Shinzo Abe</td>
#> <td headers="Japan  pm_start_in_position" class="gt_row gt_left">2012-12-26</td></tr>
#>     <tr><td headers="Japan  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Japan  pm_name" class="gt_row gt_left">Yoshihide Suga</td>
#> <td headers="Japan  pm_start_in_position" class="gt_row gt_left">2020-09-16</td></tr>
#>     <tr><td headers="Japan  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Japan  pm_name" class="gt_row gt_left">Fumio Kishida</td>
#> <td headers="Japan  pm_start_in_position" class="gt_row gt_left">2021-10-04</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Latvia">Latvia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Latvia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Latvia  pm_name" class="gt_row gt_left">Krisjanis Kariņs</td>
#> <td headers="Latvia  pm_start_in_position" class="gt_row gt_left">2019-01-23</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Lithuania">Lithuania</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Lithuania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Lithuania  pm_name" class="gt_row gt_left">Saulius Skvernelis</td>
#> <td headers="Lithuania  pm_start_in_position" class="gt_row gt_left">2016-11-22</td></tr>
#>     <tr><td headers="Lithuania  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Lithuania  pm_name" class="gt_row gt_left">Ingrida Simonyte</td>
#> <td headers="Lithuania  pm_start_in_position" class="gt_row gt_left">2020-12-11</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Luxembourg">Luxembourg</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Luxembourg  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Luxembourg  pm_name" class="gt_row gt_left">Xavier Bettel</td>
#> <td headers="Luxembourg  pm_start_in_position" class="gt_row gt_left">2013-12-04</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Malta">Malta</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Malta  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Malta  pm_name" class="gt_row gt_left">Robert Abela</td>
#> <td headers="Malta  pm_start_in_position" class="gt_row gt_left">2020-01-13</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Mauritius">Mauritius</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Mauritius  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Mauritius  pm_name" class="gt_row gt_left">Pravind Jugnauth</td>
#> <td headers="Mauritius  pm_start_in_position" class="gt_row gt_left">2017-01-23</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Mexico">Mexico</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Mexico  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Mexico  pm_name" class="gt_row gt_left">Andres Manuel Lopez Obrador</td>
#> <td headers="Mexico  pm_start_in_position" class="gt_row gt_left">2018-12-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Netherlands">Netherlands</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Netherlands  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Netherlands  pm_name" class="gt_row gt_left">Mark Rutte</td>
#> <td headers="Netherlands  pm_start_in_position" class="gt_row gt_left">2010-10-14</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="New Zealand">New Zealand</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="New Zealand  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="New Zealand  pm_name" class="gt_row gt_left">Jacinda Ardern</td>
#> <td headers="New Zealand  pm_start_in_position" class="gt_row gt_left">2017-10-26</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Norway">Norway</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Norway  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Norway  pm_name" class="gt_row gt_left">Erna Solberg</td>
#> <td headers="Norway  pm_start_in_position" class="gt_row gt_left">2013-10-16</td></tr>
#>     <tr><td headers="Norway  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Norway  pm_name" class="gt_row gt_left">Jonas Gahr Store</td>
#> <td headers="Norway  pm_start_in_position" class="gt_row gt_left">2021-10-14</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Panama">Panama</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Panama  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Panama  pm_name" class="gt_row gt_left">Laurentino Cortizo</td>
#> <td headers="Panama  pm_start_in_position" class="gt_row gt_left">2019-07-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Poland">Poland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Poland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Poland  pm_name" class="gt_row gt_left">Mateusz Morawiecki</td>
#> <td headers="Poland  pm_start_in_position" class="gt_row gt_left">2017-12-11</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Portugal">Portugal</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Portugal  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Portugal  pm_name" class="gt_row gt_left">Antonio Costa</td>
#> <td headers="Portugal  pm_start_in_position" class="gt_row gt_left">2015-11-26</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Romania">Romania</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Romania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Romania  pm_name" class="gt_row gt_left">Ludovic Orban</td>
#> <td headers="Romania  pm_start_in_position" class="gt_row gt_left">2019-11-04</td></tr>
#>     <tr><td headers="Romania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Romania  pm_name" class="gt_row gt_left">Florin Citu</td>
#> <td headers="Romania  pm_start_in_position" class="gt_row gt_left">2020-12-07</td></tr>
#>     <tr><td headers="Romania  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Romania  pm_name" class="gt_row gt_left">Nicolae Ciuca</td>
#> <td headers="Romania  pm_start_in_position" class="gt_row gt_left">2021-11-25</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Sao Tome and Principe">Sao Tome and Principe</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Sao Tome and Principe  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Sao Tome and Principe  pm_name" class="gt_row gt_left">Jorge Bom Jesus</td>
#> <td headers="Sao Tome and Principe  pm_start_in_position" class="gt_row gt_left">2018-12-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Slovakia">Slovakia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Slovakia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovakia  pm_name" class="gt_row gt_left">Peter Pellegrini</td>
#> <td headers="Slovakia  pm_start_in_position" class="gt_row gt_left">2018-03-22</td></tr>
#>     <tr><td headers="Slovakia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovakia  pm_name" class="gt_row gt_left">Igor Matovic</td>
#> <td headers="Slovakia  pm_start_in_position" class="gt_row gt_left">2020-03-21</td></tr>
#>     <tr><td headers="Slovakia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovakia  pm_name" class="gt_row gt_left">Eduard Heger</td>
#> <td headers="Slovakia  pm_start_in_position" class="gt_row gt_left">2021-04-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Slovenia">Slovenia</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Slovenia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovenia  pm_name" class="gt_row gt_left">Marjan Sarec</td>
#> <td headers="Slovenia  pm_start_in_position" class="gt_row gt_left">2018-09-13</td></tr>
#>     <tr><td headers="Slovenia  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Slovenia  pm_name" class="gt_row gt_left">Janez Jansa</td>
#> <td headers="Slovenia  pm_start_in_position" class="gt_row gt_left">2020-03-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="South Africa">South Africa</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="South Africa  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="South Africa  pm_name" class="gt_row gt_left">Cyril Ramaphosa</td>
#> <td headers="South Africa  pm_start_in_position" class="gt_row gt_left">2018-02-15</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="South Korea">South Korea</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="South Korea  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="South Korea  pm_name" class="gt_row gt_left">Moon Jae-in</td>
#> <td headers="South Korea  pm_start_in_position" class="gt_row gt_left">2017-05-10</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Spain">Spain</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Spain  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Spain  pm_name" class="gt_row gt_left">Pedro Sanchez</td>
#> <td headers="Spain  pm_start_in_position" class="gt_row gt_left">2018-06-02</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Suriname">Suriname</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Suriname  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Suriname  pm_name" class="gt_row gt_left">Desi Bouterse</td>
#> <td headers="Suriname  pm_start_in_position" class="gt_row gt_left">2010-08-12</td></tr>
#>     <tr><td headers="Suriname  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Suriname  pm_name" class="gt_row gt_left">Chan Santokhi</td>
#> <td headers="Suriname  pm_start_in_position" class="gt_row gt_left">2020-07-16</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Sweden">Sweden</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Sweden  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Sweden  pm_name" class="gt_row gt_left">Stefan Lofven</td>
#> <td headers="Sweden  pm_start_in_position" class="gt_row gt_left">2014-10-03</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Switzerland">Switzerland</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Switzerland  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Switzerland  pm_name" class="gt_row gt_left">Simonetta Sommaruga</td>
#> <td headers="Switzerland  pm_start_in_position" class="gt_row gt_left">2020-01-01</td></tr>
#>     <tr><td headers="Switzerland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Switzerland  pm_name" class="gt_row gt_left">Guy Parmelin</td>
#> <td headers="Switzerland  pm_start_in_position" class="gt_row gt_left">2021-01-01</td></tr>
#>     <tr><td headers="Switzerland  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Switzerland  pm_name" class="gt_row gt_left">Ignazio Cassis</td>
#> <td headers="Switzerland  pm_start_in_position" class="gt_row gt_left">2022-01-01</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Taiwan">Taiwan</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Taiwan  pm_gender" class="gt_row gt_left">Female</td>
#> <td headers="Taiwan  pm_name" class="gt_row gt_left">Tsai Ing-wen</td>
#> <td headers="Taiwan  pm_start_in_position" class="gt_row gt_left">2016-05-20</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Trinidad and Tobago">Trinidad and Tobago</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Trinidad and Tobago  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Trinidad and Tobago  pm_name" class="gt_row gt_left">Keith Rowley</td>
#> <td headers="Trinidad and Tobago  pm_start_in_position" class="gt_row gt_left">2015-09-09</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Turkey">Turkey</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Turkey  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Turkey  pm_name" class="gt_row gt_left">Recep Tayyip Erdogan</td>
#> <td headers="Turkey  pm_start_in_position" class="gt_row gt_left">2003-02-09</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="United Kingdom">United Kingdom</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="United Kingdom  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="United Kingdom  pm_name" class="gt_row gt_left">Boris Johnson</td>
#> <td headers="United Kingdom  pm_start_in_position" class="gt_row gt_left">2019-07-24</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="United States">United States</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="United States  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="United States  pm_name" class="gt_row gt_left">Donald Trump</td>
#> <td headers="United States  pm_start_in_position" class="gt_row gt_left">2017-01-20</td></tr>
#>     <tr><td headers="United States  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="United States  pm_name" class="gt_row gt_left">Joe Biden</td>
#> <td headers="United States  pm_start_in_position" class="gt_row gt_left">2021-01-20</td></tr>
#>     <tr class="gt_group_heading_row">
#>       <th colspan="3" class="gt_group_heading" scope="colgroup" id="Uruguay">Uruguay</th>
#>     </tr>
#>     <tr class="gt_row_group_first"><td headers="Uruguay  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Uruguay  pm_name" class="gt_row gt_left">Tabare Vazquez</td>
#> <td headers="Uruguay  pm_start_in_position" class="gt_row gt_left">2015-03-01</td></tr>
#>     <tr><td headers="Uruguay  pm_gender" class="gt_row gt_left">Male</td>
#> <td headers="Uruguay  pm_name" class="gt_row gt_left">Luis Lacalle Pou</td>
#> <td headers="Uruguay  pm_start_in_position" class="gt_row gt_left">2020-03-01</td></tr>
#>   </tbody>
#>   
#>   
#> </table>
#> </div>

2.2 Data Sources Table

data_sources.tex

wl_si_data_sources("tables/data_sources.tex")
#> <div id="syeiargnxc" style="padding-left:0px;padding-right:0px;padding-top:10px;padding-bottom:10px;overflow-x:auto;overflow-y:auto;width:auto;height:auto;">
#>   <style>#syeiargnxc table {
#>   font-family: system-ui, 'Segoe UI', Roboto, Helvetica, Arial, sans-serif, 'Apple Color Emoji', 'Segoe UI Emoji', 'Segoe UI Symbol', 'Noto Color Emoji';
#>   -webkit-font-smoothing: antialiased;
#>   -moz-osx-font-smoothing: grayscale;
#> }
#> 
#> #syeiargnxc thead, #syeiargnxc tbody, #syeiargnxc tfoot, #syeiargnxc tr, #syeiargnxc td, #syeiargnxc th {
#>   border-style: none;
#> }
#> 
#> #syeiargnxc p {
#>   margin: 0;
#>   padding: 0;
#> }
#> 
#> #syeiargnxc .gt_table {
#>   display: table;
#>   border-collapse: collapse;
#>   line-height: normal;
#>   margin-left: auto;
#>   margin-right: auto;
#>   color: #333333;
#>   font-size: 16px;
#>   font-weight: normal;
#>   font-style: normal;
#>   background-color: #FFFFFF;
#>   width: auto;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #A8A8A8;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #A8A8A8;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_caption {
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#> }
#> 
#> #syeiargnxc .gt_title {
#>   color: #333333;
#>   font-size: 125%;
#>   font-weight: initial;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-color: #FFFFFF;
#>   border-bottom-width: 0;
#> }
#> 
#> #syeiargnxc .gt_subtitle {
#>   color: #333333;
#>   font-size: 85%;
#>   font-weight: initial;
#>   padding-top: 3px;
#>   padding-bottom: 5px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-color: #FFFFFF;
#>   border-top-width: 0;
#> }
#> 
#> #syeiargnxc .gt_heading {
#>   background-color: #FFFFFF;
#>   text-align: center;
#>   border-bottom-color: #FFFFFF;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_bottom_border {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_col_headings {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_col_heading {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 6px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   overflow-x: hidden;
#> }
#> 
#> #syeiargnxc .gt_column_spanner_outer {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   padding-top: 0;
#>   padding-bottom: 0;
#>   padding-left: 4px;
#>   padding-right: 4px;
#> }
#> 
#> #syeiargnxc .gt_column_spanner_outer:first-child {
#>   padding-left: 0;
#> }
#> 
#> #syeiargnxc .gt_column_spanner_outer:last-child {
#>   padding-right: 0;
#> }
#> 
#> #syeiargnxc .gt_column_spanner {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 5px;
#>   overflow-x: hidden;
#>   display: inline-block;
#>   width: 100%;
#> }
#> 
#> #syeiargnxc .gt_spanner_row {
#>   border-bottom-style: hidden;
#> }
#> 
#> #syeiargnxc .gt_group_heading {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   text-align: left;
#> }
#> 
#> #syeiargnxc .gt_empty_group_heading {
#>   padding: 0.5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: middle;
#> }
#> 
#> #syeiargnxc .gt_from_md > :first-child {
#>   margin-top: 0;
#> }
#> 
#> #syeiargnxc .gt_from_md > :last-child {
#>   margin-bottom: 0;
#> }
#> 
#> #syeiargnxc .gt_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   margin: 10px;
#>   border-top-style: solid;
#>   border-top-width: 1px;
#>   border-top-color: #000000;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   overflow-x: hidden;
#> }
#> 
#> #syeiargnxc .gt_stub {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #syeiargnxc .gt_stub_row_group {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   vertical-align: top;
#> }
#> 
#> #syeiargnxc .gt_row_group_first td {
#>   border-top-width: 2px;
#> }
#> 
#> #syeiargnxc .gt_row_group_first th {
#>   border-top-width: 2px;
#> }
#> 
#> #syeiargnxc .gt_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #syeiargnxc .gt_first_summary_row {
#>   border-top-style: solid;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_first_summary_row.thick {
#>   border-top-width: 2px;
#> }
#> 
#> #syeiargnxc .gt_last_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_grand_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #syeiargnxc .gt_first_grand_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-style: double;
#>   border-top-width: 6px;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_last_grand_summary_row_top {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: double;
#>   border-bottom-width: 6px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_striped {
#>   background-color: rgba(128, 128, 128, 0.05);
#> }
#> 
#> #syeiargnxc .gt_table_body {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_footnotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_footnote {
#>   margin: 0px;
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #syeiargnxc .gt_sourcenotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #syeiargnxc .gt_sourcenote {
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #syeiargnxc .gt_left {
#>   text-align: left;
#> }
#> 
#> #syeiargnxc .gt_center {
#>   text-align: center;
#> }
#> 
#> #syeiargnxc .gt_right {
#>   text-align: right;
#>   font-variant-numeric: tabular-nums;
#> }
#> 
#> #syeiargnxc .gt_font_normal {
#>   font-weight: normal;
#> }
#> 
#> #syeiargnxc .gt_font_bold {
#>   font-weight: bold;
#> }
#> 
#> #syeiargnxc .gt_font_italic {
#>   font-style: italic;
#> }
#> 
#> #syeiargnxc .gt_super {
#>   font-size: 65%;
#> }
#> 
#> #syeiargnxc .gt_footnote_marks {
#>   font-size: 75%;
#>   vertical-align: 0.4em;
#>   position: initial;
#> }
#> 
#> #syeiargnxc .gt_asterisk {
#>   font-size: 100%;
#>   vertical-align: 0;
#> }
#> 
#> #syeiargnxc .gt_indent_1 {
#>   text-indent: 5px;
#> }
#> 
#> #syeiargnxc .gt_indent_2 {
#>   text-indent: 10px;
#> }
#> 
#> #syeiargnxc .gt_indent_3 {
#>   text-indent: 15px;
#> }
#> 
#> #syeiargnxc .gt_indent_4 {
#>   text-indent: 20px;
#> }
#> 
#> #syeiargnxc .gt_indent_5 {
#>   text-indent: 25px;
#> }
#> </style>
#>   <table class="gt_table" style="table-layout: fixed;; width: 0px" data-quarto-disable-processing="false" data-quarto-bootstrap="false">
#>   <colgroup>
#>     <col style="width:200px;"/>
#>     <col style="width:400px;"/>
#>     <col style="width:200px;"/>
#>   </colgroup>
#>   <thead>
#>     <tr class="gt_col_headings">
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Name">Name</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Credit">Credit</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Variables">Variables</th>
#>     </tr>
#>   </thead>
#>   <tbody class="gt_table_body">
#>     <tr><td headers="Name" class="gt_row gt_left">Freedom House</td>
#> <td headers="Credit" class="gt_row gt_left">Freedom in the World report, FIW 2013-2021</td>
#> <td headers="Variables" class="gt_row gt_left">Democracy Score</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Governmental online resources</td>
#> <td headers="Credit" class="gt_row gt_left">Governmental online resources</td>
#> <td headers="Variables" class="gt_row gt_left">Leader Age, Leader Gender[female]</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Hanson-Sigman State Capacity</td>
#> <td headers="Credit" class="gt_row gt_left">Hanson, Jonathan; Sigman, Rachel, 2020, "Leviathan's Latent Dimensions: Measuring State Capacity for Comparative Political Research", https://doi.org/10.7910/DVN/IFZXQX, Harvard Dataverse, V1</td>
#> <td headers="Variables" class="gt_row gt_left">State Capacity</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Legatum Prosperity Index (LPI)</td>
#> <td headers="Credit" class="gt_row gt_left">Legatum Institute Foundation. 2023 Legatum Prosperity Index (www.prosperity.com)</td>
#> <td headers="Variables" class="gt_row gt_left">Health-system Score</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Our World in Data (OWID)</td>
#> <td headers="Credit" class="gt_row gt_left">Hannah Ritchie, Edouard Mathieu, Lucas Rodes-Guirao, Cameron Appel, Charlie Giattino, Esteban Ortiz-Ospina, Joe Hasell, Bobbie Macdonald, Diana Beltekian and Max Roser (2020) - "Coronavirus Pandemic (COVID-19)". Published online at OurWorldInData.org.  </td>
#> <td headers="Variables" class="gt_row gt_left">Population Aged 70 or Older, Excess Mortality, Immigration Percentage, New Infection Cases, New Death Cases, New Vaccination (delayed), Stringency Index (delayed), GDP per Capita, Population Size</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">The World Bank (WDI)</td>
#> <td headers="Credit" class="gt_row gt_left">The World Bank, Urban Population, United Nations Population Division</td>
#> <td headers="Variables" class="gt_row gt_left">Urbanization Percentage</td></tr>
#>   </tbody>
#>   
#>   
#> </table>
#> </div>
wl_si_data_sources("tables/data_sources.rtf")
#> <div id="ixjoabcoer" style="padding-left:0px;padding-right:0px;padding-top:10px;padding-bottom:10px;overflow-x:auto;overflow-y:auto;width:auto;height:auto;">
#>   <style>#ixjoabcoer table {
#>   font-family: system-ui, 'Segoe UI', Roboto, Helvetica, Arial, sans-serif, 'Apple Color Emoji', 'Segoe UI Emoji', 'Segoe UI Symbol', 'Noto Color Emoji';
#>   -webkit-font-smoothing: antialiased;
#>   -moz-osx-font-smoothing: grayscale;
#> }
#> 
#> #ixjoabcoer thead, #ixjoabcoer tbody, #ixjoabcoer tfoot, #ixjoabcoer tr, #ixjoabcoer td, #ixjoabcoer th {
#>   border-style: none;
#> }
#> 
#> #ixjoabcoer p {
#>   margin: 0;
#>   padding: 0;
#> }
#> 
#> #ixjoabcoer .gt_table {
#>   display: table;
#>   border-collapse: collapse;
#>   line-height: normal;
#>   margin-left: auto;
#>   margin-right: auto;
#>   color: #333333;
#>   font-size: 16px;
#>   font-weight: normal;
#>   font-style: normal;
#>   background-color: #FFFFFF;
#>   width: auto;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #A8A8A8;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #A8A8A8;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_caption {
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#> }
#> 
#> #ixjoabcoer .gt_title {
#>   color: #333333;
#>   font-size: 125%;
#>   font-weight: initial;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-color: #FFFFFF;
#>   border-bottom-width: 0;
#> }
#> 
#> #ixjoabcoer .gt_subtitle {
#>   color: #333333;
#>   font-size: 85%;
#>   font-weight: initial;
#>   padding-top: 3px;
#>   padding-bottom: 5px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-color: #FFFFFF;
#>   border-top-width: 0;
#> }
#> 
#> #ixjoabcoer .gt_heading {
#>   background-color: #FFFFFF;
#>   text-align: center;
#>   border-bottom-color: #FFFFFF;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_bottom_border {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_col_headings {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_col_heading {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 6px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   overflow-x: hidden;
#> }
#> 
#> #ixjoabcoer .gt_column_spanner_outer {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: normal;
#>   text-transform: inherit;
#>   padding-top: 0;
#>   padding-bottom: 0;
#>   padding-left: 4px;
#>   padding-right: 4px;
#> }
#> 
#> #ixjoabcoer .gt_column_spanner_outer:first-child {
#>   padding-left: 0;
#> }
#> 
#> #ixjoabcoer .gt_column_spanner_outer:last-child {
#>   padding-right: 0;
#> }
#> 
#> #ixjoabcoer .gt_column_spanner {
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: bottom;
#>   padding-top: 5px;
#>   padding-bottom: 5px;
#>   overflow-x: hidden;
#>   display: inline-block;
#>   width: 100%;
#> }
#> 
#> #ixjoabcoer .gt_spanner_row {
#>   border-bottom-style: hidden;
#> }
#> 
#> #ixjoabcoer .gt_group_heading {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   text-align: left;
#> }
#> 
#> #ixjoabcoer .gt_empty_group_heading {
#>   padding: 0.5px;
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   vertical-align: middle;
#> }
#> 
#> #ixjoabcoer .gt_from_md > :first-child {
#>   margin-top: 0;
#> }
#> 
#> #ixjoabcoer .gt_from_md > :last-child {
#>   margin-bottom: 0;
#> }
#> 
#> #ixjoabcoer .gt_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   margin: 10px;
#>   border-top-style: solid;
#>   border-top-width: 1px;
#>   border-top-color: #000000;
#>   border-left-style: none;
#>   border-left-width: 1px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 1px;
#>   border-right-color: #D3D3D3;
#>   vertical-align: middle;
#>   overflow-x: hidden;
#> }
#> 
#> #ixjoabcoer .gt_stub {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #ixjoabcoer .gt_stub_row_group {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   font-size: 100%;
#>   font-weight: initial;
#>   text-transform: inherit;
#>   border-right-style: solid;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   vertical-align: top;
#> }
#> 
#> #ixjoabcoer .gt_row_group_first td {
#>   border-top-width: 2px;
#> }
#> 
#> #ixjoabcoer .gt_row_group_first th {
#>   border-top-width: 2px;
#> }
#> 
#> #ixjoabcoer .gt_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #ixjoabcoer .gt_first_summary_row {
#>   border-top-style: solid;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_first_summary_row.thick {
#>   border-top-width: 2px;
#> }
#> 
#> #ixjoabcoer .gt_last_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_grand_summary_row {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   text-transform: inherit;
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #ixjoabcoer .gt_first_grand_summary_row {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-top-style: double;
#>   border-top-width: 6px;
#>   border-top-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_last_grand_summary_row_top {
#>   padding-top: 8px;
#>   padding-bottom: 8px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#>   border-bottom-style: double;
#>   border-bottom-width: 6px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_striped {
#>   background-color: rgba(128, 128, 128, 0.05);
#> }
#> 
#> #ixjoabcoer .gt_table_body {
#>   border-top-style: solid;
#>   border-top-width: 2px;
#>   border-top-color: #D3D3D3;
#>   border-bottom-style: solid;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_footnotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_footnote {
#>   margin: 0px;
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #ixjoabcoer .gt_sourcenotes {
#>   color: #333333;
#>   background-color: #FFFFFF;
#>   border-bottom-style: none;
#>   border-bottom-width: 2px;
#>   border-bottom-color: #D3D3D3;
#>   border-left-style: none;
#>   border-left-width: 2px;
#>   border-left-color: #D3D3D3;
#>   border-right-style: none;
#>   border-right-width: 2px;
#>   border-right-color: #D3D3D3;
#> }
#> 
#> #ixjoabcoer .gt_sourcenote {
#>   font-size: 90%;
#>   padding-top: 4px;
#>   padding-bottom: 4px;
#>   padding-left: 5px;
#>   padding-right: 5px;
#> }
#> 
#> #ixjoabcoer .gt_left {
#>   text-align: left;
#> }
#> 
#> #ixjoabcoer .gt_center {
#>   text-align: center;
#> }
#> 
#> #ixjoabcoer .gt_right {
#>   text-align: right;
#>   font-variant-numeric: tabular-nums;
#> }
#> 
#> #ixjoabcoer .gt_font_normal {
#>   font-weight: normal;
#> }
#> 
#> #ixjoabcoer .gt_font_bold {
#>   font-weight: bold;
#> }
#> 
#> #ixjoabcoer .gt_font_italic {
#>   font-style: italic;
#> }
#> 
#> #ixjoabcoer .gt_super {
#>   font-size: 65%;
#> }
#> 
#> #ixjoabcoer .gt_footnote_marks {
#>   font-size: 75%;
#>   vertical-align: 0.4em;
#>   position: initial;
#> }
#> 
#> #ixjoabcoer .gt_asterisk {
#>   font-size: 100%;
#>   vertical-align: 0;
#> }
#> 
#> #ixjoabcoer .gt_indent_1 {
#>   text-indent: 5px;
#> }
#> 
#> #ixjoabcoer .gt_indent_2 {
#>   text-indent: 10px;
#> }
#> 
#> #ixjoabcoer .gt_indent_3 {
#>   text-indent: 15px;
#> }
#> 
#> #ixjoabcoer .gt_indent_4 {
#>   text-indent: 20px;
#> }
#> 
#> #ixjoabcoer .gt_indent_5 {
#>   text-indent: 25px;
#> }
#> </style>
#>   <table class="gt_table" style="table-layout: fixed;; width: 0px" data-quarto-disable-processing="false" data-quarto-bootstrap="false">
#>   <colgroup>
#>     <col style="width:200px;"/>
#>     <col style="width:400px;"/>
#>     <col style="width:200px;"/>
#>   </colgroup>
#>   <thead>
#>     <tr class="gt_col_headings">
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Name">Name</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Credit">Credit</th>
#>       <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" scope="col" id="Variables">Variables</th>
#>     </tr>
#>   </thead>
#>   <tbody class="gt_table_body">
#>     <tr><td headers="Name" class="gt_row gt_left">Freedom House</td>
#> <td headers="Credit" class="gt_row gt_left">Freedom in the World report, FIW 2013-2021</td>
#> <td headers="Variables" class="gt_row gt_left">Democracy Score</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Governmental online resources</td>
#> <td headers="Credit" class="gt_row gt_left">Governmental online resources</td>
#> <td headers="Variables" class="gt_row gt_left">Leader Age, Leader Gender[female]</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Hanson-Sigman State Capacity</td>
#> <td headers="Credit" class="gt_row gt_left">Hanson, Jonathan; Sigman, Rachel, 2020, "Leviathan's Latent Dimensions: Measuring State Capacity for Comparative Political Research", https://doi.org/10.7910/DVN/IFZXQX, Harvard Dataverse, V1</td>
#> <td headers="Variables" class="gt_row gt_left">State Capacity</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Legatum Prosperity Index (LPI)</td>
#> <td headers="Credit" class="gt_row gt_left">Legatum Institute Foundation. 2023 Legatum Prosperity Index (www.prosperity.com)</td>
#> <td headers="Variables" class="gt_row gt_left">Health-system Score</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">Our World in Data (OWID)</td>
#> <td headers="Credit" class="gt_row gt_left">Hannah Ritchie, Edouard Mathieu, Lucas Rodes-Guirao, Cameron Appel, Charlie Giattino, Esteban Ortiz-Ospina, Joe Hasell, Bobbie Macdonald, Diana Beltekian and Max Roser (2020) - "Coronavirus Pandemic (COVID-19)". Published online at OurWorldInData.org.  </td>
#> <td headers="Variables" class="gt_row gt_left">Population Aged 70 or Older, Excess Mortality, Immigration Percentage, New Infection Cases, New Death Cases, New Vaccination (delayed), Stringency Index (delayed), GDP per Capita, Population Size</td></tr>
#>     <tr><td headers="Name" class="gt_row gt_left">The World Bank (WDI)</td>
#> <td headers="Credit" class="gt_row gt_left">The World Bank, Urban Population, United Nations Population Division</td>
#> <td headers="Variables" class="gt_row gt_left">Urbanization Percentage</td></tr>
#>   </tbody>
#>   
#>   
#> </table>
#> </div>

2.3 Variables’ Descriptive Summary

descriptive.tex

Wave 1 Wave 2 Wave 3 Wave 4 Full
Male Female Male Female Male Female Male Female Male Female
New Infection Cases 18.7 27.0 201.8 118.4 195.3 131.8 162.2 134.3 141.9 94.4
New Death Cases 1.6 2.0 3.5 1.7 3.3 1.4 2.1 0.8 2.4 1.2
Excess Mortality 8.3 6.2 18.6 7.5 9.7 1.8 11.6 8.8 11.8 5.0
Stringency Index (delayed) 41.5 35.7 58.9 43.4 66.9 55.4 55.6 46.8 60.1 48.2
Leader Age 57.4 50.6 56.7 50.6 57.0 48.9 56.6 48.8 56.9 49.9
New Vaccination (delayed) 0.0 0.0 17.8 9.7 2,005.5 2,071.1 5,219.6 6,093.7 1,677.4 1,889.3
Democracy Score 1.5 1.0 1.5 1.0 1.5 1.0 1.5 1.1 1.5 1.0
GDP per Capita 29.8 40.8 29.4 39.3 29.9 36.4 30.0 36.0 29.6 38.3
Health-system Score 75.7 81.4 75.7 81.1 76.1 80.4 76.1 80.4 75.8 80.9
Population Size 30.9 15.3 29.3 14.8 29.7 13.4 29.7 13.5 29.7 14.3
Urbanization Percentage 72.6 80.6 72.7 77.7 73.1 76.4 73.2 76.3 72.8 77.8
Immigration Percentage 11.1 16.4 11.1 15.4 11.4 13.8 11.4 13.7 11.2 15.0
Population Aged 70 or Older 9.6 11.5 9.3 11.4 9.3 11.6 9.3 11.7 9.3 11.5
State Capacity 1.4 2.3 1.4 2.2 1.4 2.1 1.4 2.1 1.4 2.2

2.4 Dependent Variables by Leader Gender

dep_by_gender.png

x_name <- c("Week Number (starting 1-Jan-2020)" = "week_num")
x_lim <- range(wl_wave_range("All_C"))

df <- wl_plotdata("Weekly", x_name, wl_dependents()[1], "ALL", "Gender", x_lim, "All_C")
p1a <- wl_plot_xy(df, x_name, wl_dependents()[1], "Gender", "Color") + xlab("")+
    theme(legend.direction="horizontal", 
        legend.position = c(0.15, 0.8))

df <- wl_plotdata("Weekly", x_name, wl_dependents()[2], "ALL", "Gender", x_lim, "All_C")
p1b <- wl_plot_xy(df, x_name, wl_dependents()[2], "Gender", "Color") + xlab("")+
    theme(legend.direction="horizontal", 
        legend.position = c(0.15, 0.8))

df <- wl_plotdata("Weekly", x_name, wl_dependents()[3], "ALL", "Gender", x_lim, "All_C")
p1c <- wl_plot_xy(df, x_name, wl_dependents()[3], "Gender", "Color")+
  theme(legend.direction="horizontal", 
        legend.position = c(0.15, 0.8))

p <- ggarrange(p1a, p1b, p1c,  
               align = "h", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 1, nrow = 3) 

wl_ggsave("figures/dep_by_gender.png", plot = p, height=150)

3 Analysis of Health Performance

3.1 Infections, Deaths and Excess Mortality (Full)

cov_f <- wl_covariates(cov_group = "Main coviariates")

m_fc <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[1], x_var = cov_f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

m_fd <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[2], x_var = cov_f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

m_fe <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[3], x_var = cov_f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

3.1.1 Full Regression Table

performance_full_reg.tex

#> 
#> --------------------------------------------------------------------------------
#>                             New Infection Cases New Death Cases Excess Mortality
#>                                     (1)               (2)             (3)       
#> --------------------------------------------------------------------------------
#> Leader Gender[female]           -58.253***         -0.980***       -4.486***    
#>                                   (8.836)           (0.152)         (0.918)     
#> Leader Age                        -0.385           -0.016**          0.032      
#>                                   (0.309)           (0.005)         (0.035)     
#> Stringency Index (delayed)       0.629***          0.014***        -0.176***    
#>                                   (0.159)           (0.003)         (0.018)     
#> New Vaccination (delayed)         0.002*          -0.0001***        -0.0001     
#>                                   (0.001)          (0.00002)        (0.0001)    
#> Democracy Score                    2.664           0.644***         3.255***    
#>                                   (8.090)           (0.139)         (0.932)     
#> GDP per Capita                     0.481            -0.002           0.023      
#>                                   (0.305)           (0.005)         (0.037)     
#> Health-system Score               -1.209           -0.063***       -0.761***    
#>                                   (0.790)           (0.014)         (0.097)     
#> Population Size                   -0.014             0.002           -0.001     
#>                                   (0.063)           (0.001)         (0.007)     
#> Urbanization Percentage            0.145           -0.012**        -0.126***    
#>                                   (0.248)           (0.004)         (0.033)     
#> Immigration Percentage             0.222            -0.022*        -0.286***    
#>                                   (0.505)           (0.009)         (0.054)     
#> Population Aged 70 or Older      4.991***          0.233***         0.508***    
#>                                   (1.049)           (0.018)         (0.130)     
#> N                                  4,337             4,864           3,571      
#> R2                                 0.027             0.080           0.138      
#> Adjusted R2                        0.024             0.078           0.135      
#> --------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

New Infections interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 58.25 units assuming all other covariates stay constant. This represents a change of more than 0.307 standard deviations (SDNew Infection Cases= 189.7).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 0.6287 units, assuming all other covariates stay constant. This represents a change of more than 0.003314 standard deviations (SDNew Infection Cases= 189.7).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases New Infection Cases by 0.002106 units, assuming all other covariates stay constant. This represents a change of more than 1.11e-05 standard deviations (SDNew Infection Cases= 189.7).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Infection Cases by 4.991 units, assuming all other covariates stay constant. This represents a change of more than 0.02631 standard deviations (SDNew Infection Cases= 189.7).

New Deaths interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 0.9802 units assuming all other covariates stay constant. This represents a change of more than 0.2788 standard deviations (SDNew Death Cases= 3.516).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.01612 units, assuming all other covariates stay constant. This represents a change of more than 0.004585 standard deviations (SDNew Death Cases= 3.516).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.01442 units, assuming all other covariates stay constant. This represents a change of more than 0.004102 standard deviations (SDNew Death Cases= 3.516).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 8.236e-05 units, assuming all other covariates stay constant. This represents a change of more than 2.342e-05 standard deviations (SDNew Death Cases= 3.516).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases New Death Cases by 0.6441 units, assuming all other covariates stay constant. This represents a change of more than 0.1832 standard deviations (SDNew Death Cases= 3.516).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Death Cases by 0.06253 units, assuming all other covariates stay constant. This represents a change of more than 0.01778 standard deviations (SDNew Death Cases= 3.516).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.01168 units, assuming all other covariates stay constant. This represents a change of more than 0.003321 standard deviations (SDNew Death Cases= 3.516).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.02191 units, assuming all other covariates stay constant. This represents a change of more than 0.006231 standard deviations (SDNew Death Cases= 3.516).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.2327 units, assuming all other covariates stay constant. This represents a change of more than 0.06618 standard deviations (SDNew Death Cases= 3.516).

Excess Mortality interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 4.486 units assuming all other covariates stay constant. This represents a change of more than 0.2222 standard deviations (SDExcess Mortality= 20.19).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1757 units, assuming all other covariates stay constant. This represents a change of more than 0.008706 standard deviations (SDExcess Mortality= 20.19).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Excess Mortality by 3.255 units, assuming all other covariates stay constant. This represents a change of more than 0.1613 standard deviations (SDExcess Mortality= 20.19).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 0.7615 units, assuming all other covariates stay constant. This represents a change of more than 0.03772 standard deviations (SDExcess Mortality= 20.19).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.1264 units, assuming all other covariates stay constant. This represents a change of more than 0.006259 standard deviations (SDExcess Mortality= 20.19).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.2855 units, assuming all other covariates stay constant. This represents a change of more than 0.01414 standard deviations (SDExcess Mortality= 20.19).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Excess Mortality by 0.508 units, assuming all other covariates stay constant. This represents a change of more than 0.02516 standard deviations (SDExcess Mortality= 20.19).

3.1.2 Full Regression Coefficients plot

performance_full.png

title1 <- textGrob(names(wl_dependents())[1], gp = gpar(fontsize = 14))
title2 <- textGrob(names(wl_dependents())[2], gp = gpar(fontsize = 14))
title3 <- textGrob(names(wl_dependents())[3], gp = gpar(fontsize = 14))

p_fc <- wl_plot_coef (model = m_fc[[1]], cov_f, palette = "Color")
p_fd <- wl_plot_coef (model = m_fd[[1]], cov_f, palette = "Color")
p_fe <- wl_plot_coef (model = m_fe[[1]], cov_f, palette = "Color")


p <- ggarrange(title1, title2, title3, 
               p_fc, p_fd, p_fe, 
               font.label = list(size = 10, face = "bold"),
               heights = c(0.1, 1),
               ncol = 3, nrow = 2) 

wl_ggsave("figures/performance_full.png", plot = p, width=250)

3.2 New Infection Cases

dep <- wl_dependents()[1]
cov_34f <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("Wave_1C", "Wave_2C"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("Wave_3C", "Wave_4C"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

3.2.1 Infection Cases Regression Table

cases_reg.tex

#> 
#> --------------------------------------------------------------------------------
#>                              Wave 1     Wave 2     Wave 3    Wave 4      Full   
#>                                (1)       (2)        (3)        (4)       (5)    
#> --------------------------------------------------------------------------------
#> Leader Gender[female]       -12.086** -84.868*** -95.760*** -45.331** -58.253***
#>                              (3.931)   (18.609)   (21.159)  (16.591)   (8.836)  
#> Leader Age                   -0.152     0.836     -1.684*    -0.366     -0.385  
#>                              (0.153)   (0.633)    (0.697)    (0.600)   (0.309)  
#> Stringency Index (delayed)  -0.160***  1.788***   1.858***   -1.084*   0.629*** 
#>                              (0.047)   (0.377)    (0.525)    (0.425)   (0.159)  
#> New Vaccination (delayed)                         0.008**   -0.007***   0.002*  
#>                                                   (0.003)    (0.002)   (0.001)  
#> Democracy Score              -9.481*    0.272    -61.975***  22.067     2.664   
#>                              (4.131)   (16.347)   (18.481)  (15.477)   (8.090)  
#> GDP per Capita              0.614***   1.801**     -0.389     0.054     0.481   
#>                              (0.147)   (0.616)    (0.714)    (0.587)   (0.305)  
#> Health-system Score          -0.274     1.520     -4.061*   -4.895**    -1.209  
#>                              (0.378)   (1.588)    (1.832)    (1.514)   (0.790)  
#> Population Size              0.091**    0.024     -0.316*    -0.108     -0.014  
#>                              (0.028)   (0.127)    (0.146)    (0.122)   (0.063)  
#> Urbanization Percentage      -0.057   -2.008***   2.372***   0.990*     0.145   
#>                              (0.130)   (0.497)    (0.564)    (0.466)   (0.248)  
#> Immigration Percentage       -0.128     1.304     -3.934**    1.294     0.222   
#>                              (0.239)   (1.025)    (1.197)    (0.971)   (0.505)  
#> Population Aged 70 or Older  -0.288   12.801***   8.646***    2.685    4.991*** 
#>                              (0.522)   (2.134)    (2.403)    (1.999)   (1.049)  
#> N                              362      1,325       848       1,060     4,337   
#> R2                            0.221     0.125      0.136      0.065     0.027   
#> Adjusted R2                   0.199     0.118      0.124      0.055     0.024   
#> --------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 12.09 units assuming all other covariates stay constant. This represents a change of more than 0.3828 standard deviations (SDNew Infection Cases= 31.57).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Infection Cases by 0.1596 units, assuming all other covariates stay constant. This represents a change of more than 0.005055 standard deviations (SDNew Infection Cases= 31.57).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 9.481 units, assuming all other covariates stay constant. This represents a change of more than 0.3003 standard deviations (SDNew Infection Cases= 31.57).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases New Infection Cases by 0.6141 units, assuming all other covariates stay constant. This represents a change of more than 0.01945 standard deviations (SDNew Infection Cases= 31.57).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Infection Cases by 0.09115 units, assuming all other covariates stay constant. This represents a change of more than 0.002887 standard deviations (SDNew Infection Cases= 31.57).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 84.87 units assuming all other covariates stay constant. This represents a change of more than 0.3778 standard deviations (SDNew Infection Cases= 224.6).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 1.788 units, assuming all other covariates stay constant. This represents a change of more than 0.007961 standard deviations (SDNew Infection Cases= 224.6).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases New Infection Cases by 1.801 units, assuming all other covariates stay constant. This represents a change of more than 0.008016 standard deviations (SDNew Infection Cases= 224.6).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Infection Cases by 2.008 units, assuming all other covariates stay constant. This represents a change of more than 0.008938 standard deviations (SDNew Infection Cases= 224.6).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Infection Cases by 12.8 units, assuming all other covariates stay constant. This represents a change of more than 0.05699 standard deviations (SDNew Infection Cases= 224.6).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 95.76 units assuming all other covariates stay constant. This represents a change of more than 0.4722 standard deviations (SDNew Infection Cases= 202.8).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 1.684 units, assuming all other covariates stay constant. This represents a change of more than 0.008306 standard deviations (SDNew Infection Cases= 202.8).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 1.858 units, assuming all other covariates stay constant. This represents a change of more than 0.009161 standard deviations (SDNew Infection Cases= 202.8).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases New Infection Cases by 0.008333 units, assuming all other covariates stay constant. This represents a change of more than 4.109e-05 standard deviations (SDNew Infection Cases= 202.8).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 61.98 units, assuming all other covariates stay constant. This represents a change of more than 0.3056 standard deviations (SDNew Infection Cases= 202.8).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Infection Cases by 4.061 units, assuming all other covariates stay constant. This represents a change of more than 0.02003 standard deviations (SDNew Infection Cases= 202.8).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size decreases New Infection Cases by 0.3161 units, assuming all other covariates stay constant. This represents a change of more than 0.001559 standard deviations (SDNew Infection Cases= 202.8).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 2.372 units, assuming all other covariates stay constant. This represents a change of more than 0.0117 standard deviations (SDNew Infection Cases= 202.8).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Infection Cases by 3.934 units, assuming all other covariates stay constant. This represents a change of more than 0.0194 standard deviations (SDNew Infection Cases= 202.8).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Infection Cases by 8.646 units, assuming all other covariates stay constant. This represents a change of more than 0.04264 standard deviations (SDNew Infection Cases= 202.8).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 45.33 units assuming all other covariates stay constant. This represents a change of more than 0.2414 standard deviations (SDNew Infection Cases= 187.8).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Infection Cases by 1.084 units, assuming all other covariates stay constant. This represents a change of more than 0.005772 standard deviations (SDNew Infection Cases= 187.8).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Infection Cases by 0.006707 units, assuming all other covariates stay constant. This represents a change of more than 3.572e-05 standard deviations (SDNew Infection Cases= 187.8).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Infection Cases by 4.895 units, assuming all other covariates stay constant. This represents a change of more than 0.02607 standard deviations (SDNew Infection Cases= 187.8).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 0.9895 units, assuming all other covariates stay constant. This represents a change of more than 0.00527 standard deviations (SDNew Infection Cases= 187.8).

Full period interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 58.25 units assuming all other covariates stay constant. This represents a change of more than 0.307 standard deviations (SDNew Infection Cases= 189.7).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 0.6287 units, assuming all other covariates stay constant. This represents a change of more than 0.003314 standard deviations (SDNew Infection Cases= 189.7).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases New Infection Cases by 0.002106 units, assuming all other covariates stay constant. This represents a change of more than 1.11e-05 standard deviations (SDNew Infection Cases= 189.7).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Infection Cases by 4.991 units, assuming all other covariates stay constant. This represents a change of more than 0.02631 standard deviations (SDNew Infection Cases= 189.7).

3.2.2 Infection Cases (Waves 1-4) Coefficients Plot

cases_waves_reg.png

# 
p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/cases_waves_reg.png", plot = p, height=130)

3.3 New Death Cases

dep <- wl_dependents()[2]
cov_34f <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_1D", "Wave_2D"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_3D", "Wave_4D"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

3.3.1 Deaths Cases Regression Table

deaths_reg.tex

#> 
#> -------------------------------------------------------------------------------
#>                              Wave 1    Wave 2    Wave 3     Wave 4      Full   
#>                                (1)       (2)       (3)       (4)        (5)    
#> -------------------------------------------------------------------------------
#> Leader Gender[female]       -1.618*** -1.236*** -1.440***   -0.245   -0.980*** 
#>                              (0.296)   (0.360)   (0.327)   (0.257)    (0.152)  
#> Leader Age                   -0.011     0.022   -0.046***   -0.010    -0.016** 
#>                              (0.011)   (0.012)   (0.011)   (0.009)    (0.005)  
#> Stringency Index (delayed)  -0.016*** 0.043***  0.052***   0.021**    0.014*** 
#>                              (0.003)   (0.007)   (0.008)   (0.006)    (0.003)  
#> New Vaccination (delayed)                        -0.0001  -0.0002*** -0.0001***
#>                                                 (0.00004) (0.00003)  (0.00002) 
#> Democracy Score             -0.992***  -0.073    0.697*    1.575***   0.644*** 
#>                              (0.293)   (0.317)   (0.284)   (0.234)    (0.139)  
#> GDP per Capita               0.026*    -0.003    -0.019     0.003      -0.002  
#>                              (0.011)   (0.012)   (0.011)   (0.009)    (0.005)  
#> Health-system Score          -0.033   -0.084**  -0.113*** -0.088***  -0.063*** 
#>                              (0.027)   (0.031)   (0.028)   (0.023)    (0.014)  
#> Population Size             0.009***    0.002    0.0001    -0.005**    0.002   
#>                              (0.002)   (0.002)   (0.002)   (0.002)    (0.001)  
#> Urbanization Percentage       0.007   -0.059*** 0.055***  -0.033***   -0.012** 
#>                              (0.009)   (0.010)   (0.009)   (0.007)    (0.004)  
#> Immigration Percentage       -0.008     0.027   -0.091***  -0.031*    -0.022*  
#>                              (0.017)   (0.020)   (0.019)   (0.015)    (0.009)  
#> Population Aged 70 or Older   0.034   0.546***  0.187***   0.281***   0.233*** 
#>                              (0.037)   (0.041)   (0.037)   (0.030)    (0.018)  
#> N                              624      1,166     1,007     1,378      4,864   
#> R2                            0.180     0.221     0.216     0.235      0.080   
#> Adjusted R2                   0.167     0.214     0.207     0.228      0.078   
#> -------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.618 units assuming all other covariates stay constant. This represents a change of more than 0.6446 standard deviations (SDNew Death Cases= 2.51).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Death Cases by 0.01588 units, assuming all other covariates stay constant. This represents a change of more than 0.006326 standard deviations (SDNew Death Cases= 2.51).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Death Cases by 0.9922 units, assuming all other covariates stay constant. This represents a change of more than 0.3954 standard deviations (SDNew Death Cases= 2.51).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases New Death Cases by 0.02647 units, assuming all other covariates stay constant. This represents a change of more than 0.01055 standard deviations (SDNew Death Cases= 2.51).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.008759 units, assuming all other covariates stay constant. This represents a change of more than 0.00349 standard deviations (SDNew Death Cases= 2.51).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.236 units assuming all other covariates stay constant. This represents a change of more than 0.2848 standard deviations (SDNew Death Cases= 4.338).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.04301 units, assuming all other covariates stay constant. This represents a change of more than 0.009913 standard deviations (SDNew Death Cases= 4.338).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Death Cases by 0.0837 units, assuming all other covariates stay constant. This represents a change of more than 0.01929 standard deviations (SDNew Death Cases= 4.338).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.05909 units, assuming all other covariates stay constant. This represents a change of more than 0.01362 standard deviations (SDNew Death Cases= 4.338).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.5455 units, assuming all other covariates stay constant. This represents a change of more than 0.1257 standard deviations (SDNew Death Cases= 4.338).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.44 units assuming all other covariates stay constant. This represents a change of more than 0.4017 standard deviations (SDNew Death Cases= 3.586).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.04617 units, assuming all other covariates stay constant. This represents a change of more than 0.01288 standard deviations (SDNew Death Cases= 3.586).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.05154 units, assuming all other covariates stay constant. This represents a change of more than 0.01438 standard deviations (SDNew Death Cases= 3.586).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases New Death Cases by 0.6971 units, assuming all other covariates stay constant. This represents a change of more than 0.1944 standard deviations (SDNew Death Cases= 3.586).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Death Cases by 0.1125 units, assuming all other covariates stay constant. This represents a change of more than 0.03138 standard deviations (SDNew Death Cases= 3.586).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Death Cases by 0.05464 units, assuming all other covariates stay constant. This represents a change of more than 0.01524 standard deviations (SDNew Death Cases= 3.586).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.09114 units, assuming all other covariates stay constant. This represents a change of more than 0.02542 standard deviations (SDNew Death Cases= 3.586).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.1872 units, assuming all other covariates stay constant. This represents a change of more than 0.05222 standard deviations (SDNew Death Cases= 3.586).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.02057 units, assuming all other covariates stay constant. This represents a change of more than 0.00598 standard deviations (SDNew Death Cases= 3.441).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 0.0002407 units, assuming all other covariates stay constant. This represents a change of more than 6.996e-05 standard deviations (SDNew Death Cases= 3.441).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases New Death Cases by 1.575 units, assuming all other covariates stay constant. This represents a change of more than 0.4577 standard deviations (SDNew Death Cases= 3.441).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Death Cases by 0.08806 units, assuming all other covariates stay constant. This represents a change of more than 0.0256 standard deviations (SDNew Death Cases= 3.441).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size decreases New Death Cases by 0.005279 units, assuming all other covariates stay constant. This represents a change of more than 0.001534 standard deviations (SDNew Death Cases= 3.441).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.03325 units, assuming all other covariates stay constant. This represents a change of more than 0.009663 standard deviations (SDNew Death Cases= 3.441).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.03141 units, assuming all other covariates stay constant. This represents a change of more than 0.00913 standard deviations (SDNew Death Cases= 3.441).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.2814 units, assuming all other covariates stay constant. This represents a change of more than 0.08179 standard deviations (SDNew Death Cases= 3.441).

Full period interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 0.9802 units assuming all other covariates stay constant. This represents a change of more than 0.2788 standard deviations (SDNew Death Cases= 3.516).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.01612 units, assuming all other covariates stay constant. This represents a change of more than 0.004585 standard deviations (SDNew Death Cases= 3.516).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.01442 units, assuming all other covariates stay constant. This represents a change of more than 0.004102 standard deviations (SDNew Death Cases= 3.516).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 8.236e-05 units, assuming all other covariates stay constant. This represents a change of more than 2.342e-05 standard deviations (SDNew Death Cases= 3.516).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases New Death Cases by 0.6441 units, assuming all other covariates stay constant. This represents a change of more than 0.1832 standard deviations (SDNew Death Cases= 3.516).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Death Cases by 0.06253 units, assuming all other covariates stay constant. This represents a change of more than 0.01778 standard deviations (SDNew Death Cases= 3.516).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.01168 units, assuming all other covariates stay constant. This represents a change of more than 0.003321 standard deviations (SDNew Death Cases= 3.516).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.02191 units, assuming all other covariates stay constant. This represents a change of more than 0.006231 standard deviations (SDNew Death Cases= 3.516).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.2327 units, assuming all other covariates stay constant. This represents a change of more than 0.06618 standard deviations (SDNew Death Cases= 3.516).

3.3.2 Deaths Cases (Waves 1-4) Coefficients Plot

deaths_waves_reg.png

# 
p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/deaths_waves_reg.png", plot = p, height=130)

3.4 Excess Mortality

dep <- wl_dependents()[3]
cov_34f <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_1D", "Wave_2D"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_3D", "Wave_4D"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

3.4.1 Excess Mortality Regression Table

excess_reg.tex

#> 
#> ------------------------------------------------------------------------------
#>                               Wave 1    Wave 2    Wave 3    Wave 4     Full   
#>                                (1)        (2)       (3)       (4)       (5)   
#> ------------------------------------------------------------------------------
#> Leader Gender[female]       -13.893*** -7.942***  -3.588*    0.839   -4.486***
#>                              (2.464)    (2.263)   (1.605)   (1.651)   (0.918) 
#> Leader Age                    0.068     0.198*     0.005     0.011     0.032  
#>                              (0.095)    (0.085)   (0.057)   (0.063)   (0.035) 
#> Stringency Index (delayed)  -0.127***  -0.301***  -0.028    -0.071   -0.176***
#>                              (0.028)    (0.052)   (0.047)   (0.050)   (0.018) 
#> New Vaccination (delayed)                         0.001**  -0.001***  -0.0001 
#>                                                  (0.0002)  (0.0002)  (0.0001) 
#> Democracy Score             -10.506***   0.621     2.697   7.634***  3.255*** 
#>                              (2.568)    (2.215)   (1.586)   (1.685)   (0.932) 
#> GDP per Capita                0.057     0.202*    -0.134*    0.029     0.023  
#>                              (0.104)    (0.089)   (0.062)   (0.067)   (0.037) 
#> Health-system Score           -0.023   -1.174*** -0.991*** -0.876*** -0.761***
#>                              (0.250)    (0.229)   (0.166)   (0.175)   (0.097) 
#> Population Size              0.051**    -0.001    -0.031*   -0.033*   -0.001  
#>                              (0.018)    (0.017)   (0.012)   (0.013)   (0.007) 
#> Urbanization Percentage      0.278**   -0.402***  0.109*   -0.214*** -0.126***
#>                              (0.091)    (0.078)   (0.055)   (0.057)   (0.033) 
#> Immigration Percentage        -0.222   -0.363**  -0.424*** -0.308**  -0.286***
#>                              (0.147)    (0.130)   (0.096)   (0.098)   (0.054) 
#> Population Aged 70 or Older   0.832*    0.612*     0.166    0.525*   0.508*** 
#>                              (0.364)    (0.309)   (0.220)   (0.229)   (0.130) 
#> N                              483        851       724       993      3,571  
#> R2                            0.186      0.208     0.275     0.296     0.138  
#> Adjusted R2                   0.168      0.198     0.264     0.288     0.135  
#> ------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 13.89 units assuming all other covariates stay constant. This represents a change of more than 0.7261 standard deviations (SDExcess Mortality= 19.13).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1266 units, assuming all other covariates stay constant. This represents a change of more than 0.006618 standard deviations (SDExcess Mortality= 19.13).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Excess Mortality by 10.51 units, assuming all other covariates stay constant. This represents a change of more than 0.5491 standard deviations (SDExcess Mortality= 19.13).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.05129 units, assuming all other covariates stay constant. This represents a change of more than 0.002681 standard deviations (SDExcess Mortality= 19.13).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Excess Mortality by 0.2776 units, assuming all other covariates stay constant. This represents a change of more than 0.01451 standard deviations (SDExcess Mortality= 19.13).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Excess Mortality by 0.8317 units, assuming all other covariates stay constant. This represents a change of more than 0.04347 standard deviations (SDExcess Mortality= 19.13).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 7.942 units assuming all other covariates stay constant. This represents a change of more than 0.3179 standard deviations (SDExcess Mortality= 24.99).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age increases Excess Mortality by 0.1976 units, assuming all other covariates stay constant. This represents a change of more than 0.00791 standard deviations (SDExcess Mortality= 24.99).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.301 units, assuming all other covariates stay constant. This represents a change of more than 0.01205 standard deviations (SDExcess Mortality= 24.99).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Excess Mortality by 0.2018 units, assuming all other covariates stay constant. This represents a change of more than 0.008078 standard deviations (SDExcess Mortality= 24.99).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 1.174 units, assuming all other covariates stay constant. This represents a change of more than 0.047 standard deviations (SDExcess Mortality= 24.99).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.4022 units, assuming all other covariates stay constant. This represents a change of more than 0.0161 standard deviations (SDExcess Mortality= 24.99).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.3627 units, assuming all other covariates stay constant. This represents a change of more than 0.01452 standard deviations (SDExcess Mortality= 24.99).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Excess Mortality by 0.6122 units, assuming all other covariates stay constant. This represents a change of more than 0.0245 standard deviations (SDExcess Mortality= 24.99).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 3.588 units assuming all other covariates stay constant. This represents a change of more than 0.215 standard deviations (SDExcess Mortality= 16.69).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases Excess Mortality by 0.0005047 units, assuming all other covariates stay constant. This represents a change of more than 3.024e-05 standard deviations (SDExcess Mortality= 16.69).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita decreases Excess Mortality by 0.134 units, assuming all other covariates stay constant. This represents a change of more than 0.008032 standard deviations (SDExcess Mortality= 16.69).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 0.991 units, assuming all other covariates stay constant. This represents a change of more than 0.05938 standard deviations (SDExcess Mortality= 16.69).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size decreases Excess Mortality by 0.03057 units, assuming all other covariates stay constant. This represents a change of more than 0.001832 standard deviations (SDExcess Mortality= 16.69).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Excess Mortality by 0.1089 units, assuming all other covariates stay constant. This represents a change of more than 0.006528 standard deviations (SDExcess Mortality= 16.69).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.4242 units, assuming all other covariates stay constant. This represents a change of more than 0.02542 standard deviations (SDExcess Mortality= 16.69).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Excess Mortality by 0.001495 units, assuming all other covariates stay constant. This represents a change of more than 7.471e-05 standard deviations (SDExcess Mortality= 20.02).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Excess Mortality by 7.634 units, assuming all other covariates stay constant. This represents a change of more than 0.3814 standard deviations (SDExcess Mortality= 20.02).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 0.8764 units, assuming all other covariates stay constant. This represents a change of more than 0.04378 standard deviations (SDExcess Mortality= 20.02).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size decreases Excess Mortality by 0.0329 units, assuming all other covariates stay constant. This represents a change of more than 0.001644 standard deviations (SDExcess Mortality= 20.02).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.2143 units, assuming all other covariates stay constant. This represents a change of more than 0.0107 standard deviations (SDExcess Mortality= 20.02).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.3077 units, assuming all other covariates stay constant. This represents a change of more than 0.01537 standard deviations (SDExcess Mortality= 20.02).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Excess Mortality by 0.5249 units, assuming all other covariates stay constant. This represents a change of more than 0.02622 standard deviations (SDExcess Mortality= 20.02).

Full period interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 4.486 units assuming all other covariates stay constant. This represents a change of more than 0.2222 standard deviations (SDExcess Mortality= 20.19).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1757 units, assuming all other covariates stay constant. This represents a change of more than 0.008706 standard deviations (SDExcess Mortality= 20.19).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Excess Mortality by 3.255 units, assuming all other covariates stay constant. This represents a change of more than 0.1613 standard deviations (SDExcess Mortality= 20.19).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 0.7615 units, assuming all other covariates stay constant. This represents a change of more than 0.03772 standard deviations (SDExcess Mortality= 20.19).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.1264 units, assuming all other covariates stay constant. This represents a change of more than 0.006259 standard deviations (SDExcess Mortality= 20.19).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.2855 units, assuming all other covariates stay constant. This represents a change of more than 0.01414 standard deviations (SDExcess Mortality= 20.19).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Excess Mortality by 0.508 units, assuming all other covariates stay constant. This represents a change of more than 0.02516 standard deviations (SDExcess Mortality= 20.19).

3.4.2 Excess Mortality (Waves 1-4) Coefficients Plot

excess_waves_reg.png

# 
p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/excess_waves_reg.png", plot = p, height=130)

3.5 Leader Gender Coefficient Summary

gender_coef_summary.png

title1 <- textGrob(names(wl_dependents())[1], gp = gpar(fontsize = 14))
title2 <- textGrob(names(wl_dependents())[2], gp = gpar(fontsize = 14))
title3 <- textGrob(names(wl_dependents())[3], gp = gpar(fontsize = 14))
title4 <- textGrob("Leader Gender Coefficient",gp = gpar(fontsize = 14))
tblank <- textGrob(" ",gp = gpar(fontsize = 14))

p <- ggarrange(title1, title2, title3, 
               p_cases, p_deaths, p_excess, 
               tblank,title4, tblank,
               font.label = list(size = 10, face = "bold"),
               heights = c(0.1, 1, 0.1),
               ncol = 3, nrow = 3) 

wl_ggsave("figures/gender_coef_summary.png", plot = p, width=250)

4 Analysis of Relative Performance

4.1 Relative New Infection Cases

order_vars <- c("Relative Infection cases" = "ord_c", 
                "Relative Death cases" = "ord_d", 
                "Relative Excess Mortality" = "ord_e")

dep <- order_vars[1]
cov_34f <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_1 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("Wave_1C"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("Wave_1C", "Wave_2C"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("Wave_3C", "Wave_4C"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

4.1.1 Relative Infection Cases Regression Table

relative_cases_reg.tex

#> 
#> -----------------------------------------------------------------------------
#>                              Wave 1    Wave 2    Wave 3    Wave 4     Full   
#>                               (1)       (2)        (3)       (4)       (5)   
#> -----------------------------------------------------------------------------
#> Leader Gender[female]       -4.290*  -10.077*** -8.317*** -5.268*** -8.284***
#>                             (2.091)   (1.262)    (1.602)   (1.474)   (0.715) 
#> Leader Age                   -0.121    -0.066   -0.165**  -0.144**  -0.090***
#>                             (0.081)   (0.043)    (0.053)   (0.053)   (0.025) 
#> Stringency Index (delayed)   0.013    0.219***  0.264***   0.079*   0.127*** 
#>                             (0.025)   (0.026)    (0.040)   (0.038)   (0.013) 
#> New Vaccination (delayed)                       0.001***  -0.001*** -0.0002* 
#>                                                 (0.0002)  (0.0002)  (0.0001) 
#> Democracy Score              -2.496    1.119     -1.832     0.959     0.591  
#>                             (2.197)   (1.109)    (1.399)   (1.375)   (0.654) 
#> GDP per Capita              0.443***  0.151***    0.062   0.229***  0.160*** 
#>                             (0.078)   (0.042)    (0.054)   (0.052)   (0.025) 
#> Health-system Score         -0.629**   0.014    -0.638*** -0.699*** -0.441***
#>                             (0.201)   (0.108)    (0.139)   (0.135)   (0.064) 
#> Population Size              0.026     0.014     -0.021    -0.001    0.015** 
#>                             (0.015)   (0.009)    (0.011)   (0.011)   (0.005) 
#> Urbanization Percentage      0.046    -0.074*   0.165***    0.034   0.079*** 
#>                             (0.069)   (0.034)    (0.043)   (0.041)   (0.020) 
#> Immigration Percentage       -0.088    0.117    -0.260**   -0.009     0.024  
#>                             (0.127)   (0.070)    (0.091)   (0.086)   (0.041) 
#> Population Aged 70 or Older 0.758**   1.024***  1.218***    0.188   0.613*** 
#>                             (0.278)   (0.145)    (0.182)   (0.178)   (0.085) 
#> N                             362      1,325       848      1,060     4,337  
#> R2                           0.261     0.213      0.235     0.089     0.109  
#> Adjusted R2                  0.240     0.207      0.225     0.079     0.107  
#> -----------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Relative Infection cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Infection cases decreases by 4.29 units assuming all other covariates stay constant. This represents a change of more than 0.2707 standard deviations (SD[Relative Infection cases]= 15.85).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Relative Infection cases by 0.4431 units, assuming all other covariates stay constant. This represents a change of more than 0.02796 standard deviations (SD[Relative Infection cases]= 15.85).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Infection cases by 0.6289 units, assuming all other covariates stay constant. This represents a change of more than 0.03968 standard deviations (SD[Relative Infection cases]= 15.85).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Infection cases by 0.7585 units, assuming all other covariates stay constant. This represents a change of more than 0.04786 standard deviations (SD[Relative Infection cases]= 15.85).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Relative Infection cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Infection cases decreases by 10.08 units assuming all other covariates stay constant. This represents a change of more than 0.6114 standard deviations (SD[Relative Infection cases]= 16.48).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Infection cases by 0.2185 units, assuming all other covariates stay constant. This represents a change of more than 0.01326 standard deviations (SD[Relative Infection cases]= 16.48).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Relative Infection cases by 0.151 units, assuming all other covariates stay constant. This represents a change of more than 0.009164 standard deviations (SD[Relative Infection cases]= 16.48).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Relative Infection cases by 0.07396 units, assuming all other covariates stay constant. This represents a change of more than 0.004487 standard deviations (SD[Relative Infection cases]= 16.48).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Infection cases by 1.024 units, assuming all other covariates stay constant. This represents a change of more than 0.06213 standard deviations (SD[Relative Infection cases]= 16.48).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Relative Infection cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Infection cases decreases by 8.317 units assuming all other covariates stay constant. This represents a change of more than 0.5037 standard deviations (SD[Relative Infection cases]= 16.51).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Relative Infection cases by 0.165 units, assuming all other covariates stay constant. This represents a change of more than 0.009993 standard deviations (SD[Relative Infection cases]= 16.51).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Infection cases by 0.2637 units, assuming all other covariates stay constant. This represents a change of more than 0.01597 standard deviations (SD[Relative Infection cases]= 16.51).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases Relative Infection cases by 0.00102 units, assuming all other covariates stay constant. This represents a change of more than 6.179e-05 standard deviations (SD[Relative Infection cases]= 16.51).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Infection cases by 0.6377 units, assuming all other covariates stay constant. This represents a change of more than 0.03863 standard deviations (SD[Relative Infection cases]= 16.51).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Relative Infection cases by 0.1653 units, assuming all other covariates stay constant. This represents a change of more than 0.01001 standard deviations (SD[Relative Infection cases]= 16.51).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Relative Infection cases by 0.2597 units, assuming all other covariates stay constant. This represents a change of more than 0.01573 standard deviations (SD[Relative Infection cases]= 16.51).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Infection cases by 1.218 units, assuming all other covariates stay constant. This represents a change of more than 0.07379 standard deviations (SD[Relative Infection cases]= 16.51).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Relative Infection cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Infection cases decreases by 5.268 units assuming all other covariates stay constant. This represents a change of more than 0.3179 standard deviations (SD[Relative Infection cases]= 16.57).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Relative Infection cases by 0.144 units, assuming all other covariates stay constant. This represents a change of more than 0.008692 standard deviations (SD[Relative Infection cases]= 16.57).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Infection cases by 0.07899 units, assuming all other covariates stay constant. This represents a change of more than 0.004767 standard deviations (SD[Relative Infection cases]= 16.57).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Relative Infection cases by 0.0006246 units, assuming all other covariates stay constant. This represents a change of more than 3.77e-05 standard deviations (SD[Relative Infection cases]= 16.57).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Relative Infection cases by 0.2289 units, assuming all other covariates stay constant. This represents a change of more than 0.01381 standard deviations (SD[Relative Infection cases]= 16.57).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Infection cases by 0.6992 units, assuming all other covariates stay constant. This represents a change of more than 0.04219 standard deviations (SD[Relative Infection cases]= 16.57).

Full period interpretation

OLS Linear regression was used to analyze the effects on Relative Infection cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Infection cases decreases by 8.284 units assuming all other covariates stay constant. This represents a change of more than 0.5039 standard deviations (SD[Relative Infection cases]= 16.44).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Relative Infection cases by 0.08961 units, assuming all other covariates stay constant. This represents a change of more than 0.005451 standard deviations (SD[Relative Infection cases]= 16.44).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Infection cases by 0.1271 units, assuming all other covariates stay constant. This represents a change of more than 0.007731 standard deviations (SD[Relative Infection cases]= 16.44).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Relative Infection cases by 0.0001608 units, assuming all other covariates stay constant. This represents a change of more than 9.784e-06 standard deviations (SD[Relative Infection cases]= 16.44).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Relative Infection cases by 0.16 units, assuming all other covariates stay constant. This represents a change of more than 0.009731 standard deviations (SD[Relative Infection cases]= 16.44).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Infection cases by 0.4414 units, assuming all other covariates stay constant. This represents a change of more than 0.02685 standard deviations (SD[Relative Infection cases]= 16.44).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Infection cases by 0.01501 units, assuming all other covariates stay constant. This represents a change of more than 0.0009132 standard deviations (SD[Relative Infection cases]= 16.44).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Relative Infection cases by 0.07867 units, assuming all other covariates stay constant. This represents a change of more than 0.004786 standard deviations (SD[Relative Infection cases]= 16.44).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Infection cases by 0.6126 units, assuming all other covariates stay constant. This represents a change of more than 0.03727 standard deviations (SD[Relative Infection cases]= 16.44).

4.1.2 Relative Infection Cases (Full) Regression Coefficients

relative_cases_full_reg.png

p <- wl_plot_coef (model = m_f[[1]], cov_34f, palette = "Color")

wl_ggsave("figures/relative_cases_full_reg.png", plot = p)

4.1.3 Relative Infection Cases (Waves 1-4) Regression Coefficients

relative_cases_waves_reg.png

# 
p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/relative_cases_waves_reg.png", plot = p, height=130)

4.2 Relative New Death Cases

dep <- order_vars[2]
cov_34f <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_1D", "Wave_2D"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_3D", "Wave_4D"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

4.2.1 Relative Deaths Cases Regression Table

relative_deaths_reg.tex

#> 
#> ------------------------------------------------------------------------------
#>                              Wave 1    Wave 2    Wave 3    Wave 4      Full   
#>                                (1)       (2)       (3)       (4)       (5)    
#> ------------------------------------------------------------------------------
#> Leader Gender[female]       -6.323*** -7.426*** -7.544*** -5.940*** -7.426*** 
#>                              (1.629)   (1.227)   (1.334)   (1.128)   (0.630)  
#> Leader Age                   -0.004    -0.015   -0.216*** -0.151*** -0.120*** 
#>                              (0.061)   (0.042)   (0.044)   (0.040)   (0.022)  
#> Stringency Index (delayed)  0.092***  0.247***  0.295***  0.220***   0.188*** 
#>                              (0.019)   (0.025)   (0.033)   (0.028)   (0.011)  
#> New Vaccination (delayed)                        0.0003*  -0.001*** -0.0003***
#>                                                 (0.0002)  (0.0001)   (0.0001) 
#> Democracy Score              -2.033    2.332*   4.304***  5.330***   3.884*** 
#>                              (1.615)   (1.080)   (1.160)   (1.027)   (0.575)  
#> GDP per Capita              0.369***    0.005   -0.143**    0.037     0.010   
#>                              (0.059)   (0.041)   (0.045)   (0.040)   (0.022)  
#> Health-system Score         -0.653*** -0.434*** -0.961*** -0.581*** -0.537*** 
#>                              (0.151)   (0.105)   (0.115)   (0.102)   (0.056)  
#> Population Size             0.041***   0.026**    0.018    0.016*    0.032*** 
#>                              (0.011)   (0.008)   (0.009)   (0.008)   (0.004)  
#> Urbanization Percentage       0.088   -0.154*** 0.122***  -0.234***  -0.055** 
#>                              (0.050)   (0.033)   (0.036)   (0.031)   (0.018)  
#> Immigration Percentage       -0.225*   0.153*   -0.250***  -0.066     -0.030  
#>                              (0.096)   (0.068)   (0.076)   (0.065)   (0.036)  
#> Population Aged 70 or Older 1.203***  2.265***  1.676***  1.355***   1.396*** 
#>                              (0.203)   (0.140)   (0.152)   (0.134)   (0.075)  
#> N                              624      1,166     1,007     1,378     4,864   
#> R2                            0.264     0.330     0.355     0.307     0.207   
#> Adjusted R2                   0.252     0.324     0.348     0.301     0.205   
#> ------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Relative Death cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Death cases decreases by 6.323 units assuming all other covariates stay constant. This represents a change of more than 0.3986 standard deviations (SD[Relative Death cases]= 15.86).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Death cases by 0.09172 units, assuming all other covariates stay constant. This represents a change of more than 0.005782 standard deviations (SD[Relative Death cases]= 15.86).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Relative Death cases by 0.3693 units, assuming all other covariates stay constant. This represents a change of more than 0.02328 standard deviations (SD[Relative Death cases]= 15.86).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Death cases by 0.6534 units, assuming all other covariates stay constant. This represents a change of more than 0.04119 standard deviations (SD[Relative Death cases]= 15.86).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Death cases by 0.04086 units, assuming all other covariates stay constant. This represents a change of more than 0.002576 standard deviations (SD[Relative Death cases]= 15.86).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Relative Death cases by 0.2247 units, assuming all other covariates stay constant. This represents a change of more than 0.01417 standard deviations (SD[Relative Death cases]= 15.86).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Death cases by 1.203 units, assuming all other covariates stay constant. This represents a change of more than 0.07581 standard deviations (SD[Relative Death cases]= 15.86).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Relative Death cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Death cases decreases by 7.426 units assuming all other covariates stay constant. This represents a change of more than 0.4515 standard deviations (SD[Relative Death cases]= 16.45).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Death cases by 0.2468 units, assuming all other covariates stay constant. This represents a change of more than 0.01501 standard deviations (SD[Relative Death cases]= 16.45).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Relative Death cases by 2.332 units, assuming all other covariates stay constant. This represents a change of more than 0.1418 standard deviations (SD[Relative Death cases]= 16.45).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Death cases by 0.4335 units, assuming all other covariates stay constant. This represents a change of more than 0.02636 standard deviations (SD[Relative Death cases]= 16.45).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Death cases by 0.02555 units, assuming all other covariates stay constant. This represents a change of more than 0.001553 standard deviations (SD[Relative Death cases]= 16.45).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Relative Death cases by 0.1544 units, assuming all other covariates stay constant. This represents a change of more than 0.00939 standard deviations (SD[Relative Death cases]= 16.45).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage increases Relative Death cases by 0.153 units, assuming all other covariates stay constant. This represents a change of more than 0.009305 standard deviations (SD[Relative Death cases]= 16.45).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Death cases by 2.265 units, assuming all other covariates stay constant. This represents a change of more than 0.1377 standard deviations (SD[Relative Death cases]= 16.45).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Relative Death cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Death cases decreases by 7.544 units assuming all other covariates stay constant. This represents a change of more than 0.4556 standard deviations (SD[Relative Death cases]= 16.56).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Relative Death cases by 0.216 units, assuming all other covariates stay constant. This represents a change of more than 0.01305 standard deviations (SD[Relative Death cases]= 16.56).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Death cases by 0.2951 units, assuming all other covariates stay constant. This represents a change of more than 0.01782 standard deviations (SD[Relative Death cases]= 16.56).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases Relative Death cases by 0.0003071 units, assuming all other covariates stay constant. This represents a change of more than 1.855e-05 standard deviations (SD[Relative Death cases]= 16.56).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Relative Death cases by 4.304 units, assuming all other covariates stay constant. This represents a change of more than 0.2599 standard deviations (SD[Relative Death cases]= 16.56).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita decreases Relative Death cases by 0.1428 units, assuming all other covariates stay constant. This represents a change of more than 0.008624 standard deviations (SD[Relative Death cases]= 16.56).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Death cases by 0.9606 units, assuming all other covariates stay constant. This represents a change of more than 0.05802 standard deviations (SD[Relative Death cases]= 16.56).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Relative Death cases by 0.1216 units, assuming all other covariates stay constant. This represents a change of more than 0.007342 standard deviations (SD[Relative Death cases]= 16.56).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Relative Death cases by 0.2502 units, assuming all other covariates stay constant. This represents a change of more than 0.01511 standard deviations (SD[Relative Death cases]= 16.56).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Death cases by 1.676 units, assuming all other covariates stay constant. This represents a change of more than 0.1012 standard deviations (SD[Relative Death cases]= 16.56).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Relative Death cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Death cases decreases by 5.94 units assuming all other covariates stay constant. This represents a change of more than 0.3586 standard deviations (SD[Relative Death cases]= 16.57).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Relative Death cases by 0.1514 units, assuming all other covariates stay constant. This represents a change of more than 0.009138 standard deviations (SD[Relative Death cases]= 16.57).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Death cases by 0.2201 units, assuming all other covariates stay constant. This represents a change of more than 0.01329 standard deviations (SD[Relative Death cases]= 16.57).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Relative Death cases by 0.0009421 units, assuming all other covariates stay constant. This represents a change of more than 5.687e-05 standard deviations (SD[Relative Death cases]= 16.57).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Relative Death cases by 5.33 units, assuming all other covariates stay constant. This represents a change of more than 0.3218 standard deviations (SD[Relative Death cases]= 16.57).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Death cases by 0.5814 units, assuming all other covariates stay constant. This represents a change of more than 0.0351 standard deviations (SD[Relative Death cases]= 16.57).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Death cases by 0.01609 units, assuming all other covariates stay constant. This represents a change of more than 0.000971 standard deviations (SD[Relative Death cases]= 16.57).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Relative Death cases by 0.2339 units, assuming all other covariates stay constant. This represents a change of more than 0.01412 standard deviations (SD[Relative Death cases]= 16.57).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Death cases by 1.355 units, assuming all other covariates stay constant. This represents a change of more than 0.08179 standard deviations (SD[Relative Death cases]= 16.57).

Full period interpretation

OLS Linear regression was used to analyze the effects on Relative Death cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Death cases decreases by 7.426 units assuming all other covariates stay constant. This represents a change of more than 0.4519 standard deviations (SD[Relative Death cases]= 16.43).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Relative Death cases by 0.12 units, assuming all other covariates stay constant. This represents a change of more than 0.007302 standard deviations (SD[Relative Death cases]= 16.43).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Relative Death cases by 0.1876 units, assuming all other covariates stay constant. This represents a change of more than 0.01141 standard deviations (SD[Relative Death cases]= 16.43).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Relative Death cases by 0.0002917 units, assuming all other covariates stay constant. This represents a change of more than 1.775e-05 standard deviations (SD[Relative Death cases]= 16.43).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Relative Death cases by 3.884 units, assuming all other covariates stay constant. This represents a change of more than 0.2363 standard deviations (SD[Relative Death cases]= 16.43).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Death cases by 0.5366 units, assuming all other covariates stay constant. This represents a change of more than 0.03265 standard deviations (SD[Relative Death cases]= 16.43).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Death cases by 0.03177 units, assuming all other covariates stay constant. This represents a change of more than 0.001933 standard deviations (SD[Relative Death cases]= 16.43).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Relative Death cases by 0.05534 units, assuming all other covariates stay constant. This represents a change of more than 0.003367 standard deviations (SD[Relative Death cases]= 16.43).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Death cases by 1.396 units, assuming all other covariates stay constant. This represents a change of more than 0.08496 standard deviations (SD[Relative Death cases]= 16.43).

4.2.2 Relative Deaths Cases (Full) Regression Coefficients

relative_deaths_full_reg.png

p <- wl_plot_coef (model = m_f[[1]], cov_34f, palette = "Color")

wl_ggsave("figures/relative_deaths_full_reg.png", plot = p)

4.2.3 Relative Deaths Cases (Waves 1-4) Regression Coefficients

relative_deaths_waves_reg.png

# 
p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/relative_deaths_waves_reg.png", plot = p, height=130)

4.3 Relative Excess Mortality

dep <- order_vars[3]
cov_34f <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_1D", "Wave_2D"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_3D", "Wave_4D"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

4.3.1 Relative Excess Mortality Regression Table

relative_excess_reg.tex

#> 
#> -----------------------------------------------------------------------------
#>                              Wave 1    Wave 2    Wave 3    Wave 4     Full   
#>                                (1)       (2)       (3)       (4)       (5)   
#> -----------------------------------------------------------------------------
#> Leader Gender[female]       -8.515*** -3.708**   -1.951     1.032   -2.473***
#>                              (1.503)   (1.131)   (1.132)   (0.994)   (0.543) 
#> Leader Age                    0.024     0.022    -0.039     0.026    -0.004  
#>                              (0.058)   (0.042)   (0.040)   (0.038)   (0.021) 
#> Stringency Index (delayed)    0.015    -0.016     0.004    -0.049    -0.009  
#>                              (0.017)   (0.026)   (0.033)   (0.030)   (0.010) 
#> New Vaccination (delayed)                       0.001***  -0.0003**  -0.0001 
#>                                                 (0.0001)  (0.0001)  (0.0001) 
#> Democracy Score             -4.264**    0.683    2.520*   4.155***  2.777*** 
#>                              (1.567)   (1.107)   (1.119)   (1.015)   (0.551) 
#> GDP per Capita               0.127*    -0.010   -0.168***   0.022    -0.010  
#>                              (0.063)   (0.044)   (0.044)   (0.041)   (0.022) 
#> Health-system Score           0.125   -0.360**  -0.702*** -0.560*** -0.400***
#>                              (0.152)   (0.115)   (0.117)   (0.105)   (0.057) 
#> Population Size             0.041***   0.020*    -0.005    -0.002    0.013** 
#>                              (0.011)   (0.008)   (0.008)   (0.008)   (0.004) 
#> Urbanization Percentage      0.135*   -0.225***  -0.010   -0.116*** -0.081***
#>                              (0.056)   (0.039)   (0.039)   (0.035)   (0.019) 
#> Immigration Percentage       -0.224*   -0.088    -0.140*  -0.193**  -0.162***
#>                              (0.089)   (0.065)   (0.068)   (0.059)   (0.032) 
#> Population Aged 70 or Older   0.381     0.178     0.179    0.383**  0.315*** 
#>                              (0.222)   (0.154)   (0.155)   (0.138)   (0.077) 
#> N                              483       851       724       993      3,571  
#> R2                            0.157     0.189     0.304     0.222     0.146  
#> Adjusted R2                   0.139     0.180     0.293     0.213     0.144  
#> -----------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Relative Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Excess Mortality decreases by 8.515 units assuming all other covariates stay constant. This represents a change of more than 0.697 standard deviations (SDRelative Excess Mortality= 12.22).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Relative Excess Mortality by 4.264 units, assuming all other covariates stay constant. This represents a change of more than 0.349 standard deviations (SDRelative Excess Mortality= 12.22).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Relative Excess Mortality by 0.1271 units, assuming all other covariates stay constant. This represents a change of more than 0.0104 standard deviations (SDRelative Excess Mortality= 12.22).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Excess Mortality by 0.0414 units, assuming all other covariates stay constant. This represents a change of more than 0.003389 standard deviations (SDRelative Excess Mortality= 12.22).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Relative Excess Mortality by 0.1348 units, assuming all other covariates stay constant. This represents a change of more than 0.01103 standard deviations (SDRelative Excess Mortality= 12.22).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Relative Excess Mortality by 0.2235 units, assuming all other covariates stay constant. This represents a change of more than 0.0183 standard deviations (SDRelative Excess Mortality= 12.22).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Relative Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Excess Mortality decreases by 3.708 units assuming all other covariates stay constant. This represents a change of more than 0.2998 standard deviations (SDRelative Excess Mortality= 12.37).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Excess Mortality by 0.3599 units, assuming all other covariates stay constant. This represents a change of more than 0.0291 standard deviations (SDRelative Excess Mortality= 12.37).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Excess Mortality by 0.0197 units, assuming all other covariates stay constant. This represents a change of more than 0.001593 standard deviations (SDRelative Excess Mortality= 12.37).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Relative Excess Mortality by 0.2254 units, assuming all other covariates stay constant. This represents a change of more than 0.01822 standard deviations (SDRelative Excess Mortality= 12.37).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Relative Excess Mortality.
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases Relative Excess Mortality by 0.0005779 units, assuming all other covariates stay constant. This represents a change of more than 4.807e-05 standard deviations (SDRelative Excess Mortality= 12.02).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Relative Excess Mortality by 2.52 units, assuming all other covariates stay constant. This represents a change of more than 0.2096 standard deviations (SDRelative Excess Mortality= 12.02).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita decreases Relative Excess Mortality by 0.1679 units, assuming all other covariates stay constant. This represents a change of more than 0.01396 standard deviations (SDRelative Excess Mortality= 12.02).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Excess Mortality by 0.7016 units, assuming all other covariates stay constant. This represents a change of more than 0.05836 standard deviations (SDRelative Excess Mortality= 12.02).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Relative Excess Mortality by 0.1402 units, assuming all other covariates stay constant. This represents a change of more than 0.01166 standard deviations (SDRelative Excess Mortality= 12.02).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Relative Excess Mortality.
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Relative Excess Mortality by 0.0003041 units, assuming all other covariates stay constant. This represents a change of more than 2.564e-05 standard deviations (SDRelative Excess Mortality= 11.86).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Relative Excess Mortality by 4.155 units, assuming all other covariates stay constant. This represents a change of more than 0.3504 standard deviations (SDRelative Excess Mortality= 11.86).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Excess Mortality by 0.5599 units, assuming all other covariates stay constant. This represents a change of more than 0.04722 standard deviations (SDRelative Excess Mortality= 11.86).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Relative Excess Mortality by 0.1156 units, assuming all other covariates stay constant. This represents a change of more than 0.009745 standard deviations (SDRelative Excess Mortality= 11.86).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Relative Excess Mortality by 0.1926 units, assuming all other covariates stay constant. This represents a change of more than 0.01624 standard deviations (SDRelative Excess Mortality= 11.86).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Excess Mortality by 0.3827 units, assuming all other covariates stay constant. This represents a change of more than 0.03228 standard deviations (SDRelative Excess Mortality= 11.86).

Full period interpretation

OLS Linear regression was used to analyze the effects on Relative Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Relative Excess Mortality decreases by 2.473 units assuming all other covariates stay constant. This represents a change of more than 0.2024 standard deviations (SDRelative Excess Mortality= 12.22).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Relative Excess Mortality by 2.777 units, assuming all other covariates stay constant. This represents a change of more than 0.2272 standard deviations (SDRelative Excess Mortality= 12.22).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Relative Excess Mortality by 0.4004 units, assuming all other covariates stay constant. This represents a change of more than 0.03276 standard deviations (SDRelative Excess Mortality= 12.22).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Relative Excess Mortality by 0.01312 units, assuming all other covariates stay constant. This represents a change of more than 0.001074 standard deviations (SDRelative Excess Mortality= 12.22).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Relative Excess Mortality by 0.08089 units, assuming all other covariates stay constant. This represents a change of more than 0.006619 standard deviations (SDRelative Excess Mortality= 12.22).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Relative Excess Mortality by 0.1621 units, assuming all other covariates stay constant. This represents a change of more than 0.01326 standard deviations (SDRelative Excess Mortality= 12.22).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Relative Excess Mortality by 0.3147 units, assuming all other covariates stay constant. This represents a change of more than 0.02575 standard deviations (SDRelative Excess Mortality= 12.22).

4.3.2 Relative Excess Mortality (Full) Regression Coefficients

relative_excess_full_reg.png

p <- wl_plot_coef (model = m_f[[1]], cov_34f, palette = "Color")

wl_ggsave("figures/relative_excess_full_reg.png", plot = p)

4.3.3 Relative Excess Mortality (Waves 1-4) Regression Coefficients

relative_excess_waves_reg.png

# 
p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/relative_excess_waves_reg.png", plot = p, height=130)

4.4 Relative Performance - Leader Gender Coefficient Summary

relative_gender_coef_summary.png

title1 <- textGrob(names(wl_dependents())[1], gp = gpar(fontsize = 14))
title2 <- textGrob(names(wl_dependents())[2], gp = gpar(fontsize = 14))
title3 <- textGrob(names(wl_dependents())[3], gp = gpar(fontsize = 14))
title4 <- textGrob("Leader Gender Coefficient",gp = gpar(fontsize = 14))
tblank <- textGrob(" ",gp = gpar(fontsize = 14))

p <- ggarrange(title1, title2, title3, 
               p_cases, p_deaths, p_excess, 
               tblank,title4, tblank,
               font.label = list(size = 10, face = "bold"),
               heights = c(0.1, 1, 0.1),
               ncol = 3, nrow = 3) 

wl_ggsave("figures/relative_gender_coef_summary.png", plot = p, width=250)

4.5 Relative Perforamnce Regressions Results - Aligned timeline

4.5.1 Aligned - Relative New Infection cases Regression models

order_aligned_vars <- c("Infection Cases Order - Wave 1" = "ord_w1c", 
                        "Infection Cases Order - Wave 2" = "ord_w2c",
                        "Infection Cases Order - Wave 3" = "ord_w3c",
                        "Infection Cases Order - Wave 4" = "ord_w4c")

cov_34 <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

dep <- order_aligned_vars[1]
m_1 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_1AC"))

dep <- order_aligned_vars[2]
m_2 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_2AC"))

dep <- order_aligned_vars[3]
m_3 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_3AC"))

dep <- order_aligned_vars[4]
m_4 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_4AC"))

4.5.2 Aligned - Relative Infection Cases Regression Table

relative_aligned_cases_reg.tex

#> 
#> --------------------------------------------------------------------
#>                              Wave 1    Wave 2     Wave 3    Wave 4  
#>                                (1)       (2)       (3)        (4)   
#> --------------------------------------------------------------------
#> Leader Gender[female]       -7.708*** -9.041*** -12.642*** -7.017***
#>                              (1.728)   (1.338)   (1.378)    (1.426) 
#> Leader Age                    0.046    -0.001   -0.159***  -0.177***
#>                              (0.063)   (0.047)   (0.046)    (0.050) 
#> Stringency Index (delayed)   0.054**  0.203***   0.325***   -0.013  
#>                              (0.019)   (0.028)   (0.033)    (0.034) 
#> New Vaccination (delayed)                         0.0003   -0.001***
#>                                                  (0.0002)  (0.0002) 
#> Democracy Score             -4.818**   -0.349    -2.718*     0.080  
#>                              (1.594)   (1.212)   (1.193)    (1.295) 
#> GDP per Capita              0.312***   0.124**    0.036     0.149** 
#>                              (0.062)   (0.046)   (0.046)    (0.049) 
#> Health-system Score         -0.770***  -0.253*  -0.657***  -0.622***
#>                              (0.148)   (0.118)   (0.119)    (0.126) 
#> Population Size             0.038***    0.004    -0.024*    -0.009  
#>                              (0.011)   (0.009)   (0.009)    (0.010) 
#> Urbanization Percentage     0.289***   -0.046    0.180***    0.004  
#>                              (0.049)   (0.037)   (0.037)    (0.040) 
#> Immigration Percentage       -0.146    0.150*    -0.198**   -0.100  
#>                              (0.101)   (0.076)   (0.077)    (0.081) 
#> Population Aged 70 or Older   0.145   1.560***   1.026***  0.929*** 
#>                              (0.208)   (0.158)   (0.156)    (0.167) 
#> N                              630      1,113     1,113      1,113  
#> R2                            0.210     0.238     0.258      0.148  
#> Adjusted R2                   0.197     0.231     0.251      0.139  
#> --------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Infection Cases Order - Wave 1.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Infection Cases Order - Wave 1 decreases by 7.708 units assuming all other covariates stay constant. This represents a change of more than 0.478 standard deviations (SD[Infection Cases Order - Wave 1]= 16.13).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Infection Cases Order - Wave 1 by 0.05409 units, assuming all other covariates stay constant. This represents a change of more than 0.003354 standard deviations (SD[Infection Cases Order - Wave 1]= 16.13).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Infection Cases Order - Wave 1 by 4.818 units, assuming all other covariates stay constant. This represents a change of more than 0.2987 standard deviations (SD[Infection Cases Order - Wave 1]= 16.13).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Infection Cases Order - Wave 1 by 0.3124 units, assuming all other covariates stay constant. This represents a change of more than 0.01937 standard deviations (SD[Infection Cases Order - Wave 1]= 16.13).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Infection Cases Order - Wave 1 by 0.7705 units, assuming all other covariates stay constant. This represents a change of more than 0.04778 standard deviations (SD[Infection Cases Order - Wave 1]= 16.13).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Infection Cases Order - Wave 1 by 0.03821 units, assuming all other covariates stay constant. This represents a change of more than 0.002369 standard deviations (SD[Infection Cases Order - Wave 1]= 16.13).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Infection Cases Order - Wave 1 by 0.2888 units, assuming all other covariates stay constant. This represents a change of more than 0.01791 standard deviations (SD[Infection Cases Order - Wave 1]= 16.13).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Infection Cases Order - Wave 2.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Infection Cases Order - Wave 2 decreases by 9.041 units assuming all other covariates stay constant. This represents a change of more than 0.5475 standard deviations (SD[Infection Cases Order - Wave 2]= 16.51).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Infection Cases Order - Wave 2 by 0.2034 units, assuming all other covariates stay constant. This represents a change of more than 0.01232 standard deviations (SD[Infection Cases Order - Wave 2]= 16.51).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Infection Cases Order - Wave 2 by 0.1238 units, assuming all other covariates stay constant. This represents a change of more than 0.007495 standard deviations (SD[Infection Cases Order - Wave 2]= 16.51).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Infection Cases Order - Wave 2 by 0.2534 units, assuming all other covariates stay constant. This represents a change of more than 0.01535 standard deviations (SD[Infection Cases Order - Wave 2]= 16.51).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage increases Infection Cases Order - Wave 2 by 0.15 units, assuming all other covariates stay constant. This represents a change of more than 0.009082 standard deviations (SD[Infection Cases Order - Wave 2]= 16.51).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Infection Cases Order - Wave 2 by 1.56 units, assuming all other covariates stay constant. This represents a change of more than 0.09448 standard deviations (SD[Infection Cases Order - Wave 2]= 16.51).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Infection Cases Order - Wave 3.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Infection Cases Order - Wave 3 decreases by 12.64 units assuming all other covariates stay constant. This represents a change of more than 0.7652 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Infection Cases Order - Wave 3 by 0.1594 units, assuming all other covariates stay constant. This represents a change of more than 0.00965 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Infection Cases Order - Wave 3 by 0.3254 units, assuming all other covariates stay constant. This represents a change of more than 0.01969 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Infection Cases Order - Wave 3 by 2.718 units, assuming all other covariates stay constant. This represents a change of more than 0.1645 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Infection Cases Order - Wave 3 by 0.6574 units, assuming all other covariates stay constant. This represents a change of more than 0.03979 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size decreases Infection Cases Order - Wave 3 by 0.02392 units, assuming all other covariates stay constant. This represents a change of more than 0.001448 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Infection Cases Order - Wave 3 by 0.1804 units, assuming all other covariates stay constant. This represents a change of more than 0.01092 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Infection Cases Order - Wave 3 by 0.1984 units, assuming all other covariates stay constant. This represents a change of more than 0.01201 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Infection Cases Order - Wave 3 by 1.026 units, assuming all other covariates stay constant. This represents a change of more than 0.0621 standard deviations (SD[Infection Cases Order - Wave 3]= 16.52).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Infection Cases Order - Wave 4.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Infection Cases Order - Wave 4 decreases by 7.017 units assuming all other covariates stay constant. This represents a change of more than 0.4267 standard deviations (SD[Infection Cases Order - Wave 4]= 16.45).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Infection Cases Order - Wave 4 by 0.1769 units, assuming all other covariates stay constant. This represents a change of more than 0.01076 standard deviations (SD[Infection Cases Order - Wave 4]= 16.45).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Infection Cases Order - Wave 4 by 0.001258 units, assuming all other covariates stay constant. This represents a change of more than 7.649e-05 standard deviations (SD[Infection Cases Order - Wave 4]= 16.45).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Infection Cases Order - Wave 4 by 0.1488 units, assuming all other covariates stay constant. This represents a change of more than 0.009047 standard deviations (SD[Infection Cases Order - Wave 4]= 16.45).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Infection Cases Order - Wave 4 by 0.6224 units, assuming all other covariates stay constant. This represents a change of more than 0.03784 standard deviations (SD[Infection Cases Order - Wave 4]= 16.45).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Infection Cases Order - Wave 4 by 0.9291 units, assuming all other covariates stay constant. This represents a change of more than 0.05649 standard deviations (SD[Infection Cases Order - Wave 4]= 16.45).

4.5.3 Aligned - Relative Infection Cases (Waves 1-4) Regression Coefficients

relative_aligned_cases_waves_reg.png

p1 <- wl_plot_coef (model = m_1[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_2[[1]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_3[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_4[[1]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/relative_aligned_cases_waves_reg.png", plot = p, height=130)

4.6 Aligned - Relative New Death Cases Regression models

order_aligned_vars <- c("Death Cases Order - Wave 1" = "ord_w1d", 
                        "Death Cases Order - Wave 2" = "ord_w2d",
                        "Death Cases Order - Wave 3" = "ord_w3d",
                        "Death Cases Order - Wave 4" = "ord_w4d")

cov_34 <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

dep <- order_aligned_vars[1]
m_1 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_1AD"))

dep <- order_aligned_vars[2]
m_2 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_2AD"))

dep <- order_aligned_vars[3]
m_3 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_3AD"))

dep <- order_aligned_vars[4]
m_4 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_4AD"))

4.6.1 Aligned - Relative Deaths Cases Regression Table

relative_aligned_deaths_reg.tex

#> 
#> --------------------------------------------------------------------
#>                              Wave 1    Wave 2     Wave 3    Wave 4  
#>                                (1)       (2)       (3)        (4)   
#> --------------------------------------------------------------------
#> Leader Gender[female]       -6.798*** -7.548*** -10.395*** -4.991** 
#>                              (1.670)   (1.233)   (1.261)    (1.706) 
#> Leader Age                    0.031    -0.019    -0.138**  -0.269***
#>                              (0.060)   (0.043)   (0.042)    (0.057) 
#> Stringency Index (delayed)  0.135***  0.241***   0.315***  0.173*** 
#>                              (0.020)   (0.025)   (0.031)    (0.043) 
#> New Vaccination (delayed)                        0.00003   0.001*** 
#>                                                  (0.0002)  (0.0003) 
#> Democracy Score              -3.465*    1.130    3.725***  9.325*** 
#>                              (1.489)   (1.114)   (1.108)    (1.623) 
#> GDP per Capita              0.347***   -0.042     0.018      0.076  
#>                              (0.059)   (0.042)   (0.042)    (0.060) 
#> Health-system Score         -0.608*** -0.708*** -1.108***  -0.704***
#>                              (0.139)   (0.110)   (0.110)    (0.129) 
#> Population Size             0.036***  0.034***    0.009      0.022  
#>                              (0.011)   (0.009)   (0.009)    (0.012) 
#> Urbanization Percentage      0.159**  -0.127***  0.154***  -0.245***
#>                              (0.049)   (0.034)   (0.034)    (0.040) 
#> Immigration Percentage       -0.233*  0.252***  -0.251***  -0.464***
#>                              (0.097)   (0.069)   (0.071)    (0.103) 
#> Population Aged 70 or Older 0.689***  2.421***   1.802***  2.503*** 
#>                              (0.191)   (0.145)   (0.144)    (0.187) 
#> N                              714      1,113     1,113       628   
#> R2                            0.212     0.339     0.356      0.438  
#> Adjusted R2                   0.201     0.333     0.349      0.428  
#> --------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 1.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Death Cases Order - Wave 1 decreases by 6.798 units assuming all other covariates stay constant. This represents a change of more than 0.4253 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Death Cases Order - Wave 1 by 0.1348 units, assuming all other covariates stay constant. This represents a change of more than 0.008433 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Death Cases Order - Wave 1 by 3.465 units, assuming all other covariates stay constant. This represents a change of more than 0.2168 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Death Cases Order - Wave 1 by 0.3467 units, assuming all other covariates stay constant. This represents a change of more than 0.02169 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 1 by 0.6083 units, assuming all other covariates stay constant. This represents a change of more than 0.03806 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Death Cases Order - Wave 1 by 0.03643 units, assuming all other covariates stay constant. This represents a change of more than 0.002279 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Death Cases Order - Wave 1 by 0.1592 units, assuming all other covariates stay constant. This represents a change of more than 0.009959 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Death Cases Order - Wave 1 by 0.233 units, assuming all other covariates stay constant. This represents a change of more than 0.01457 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Death Cases Order - Wave 1 by 0.6894 units, assuming all other covariates stay constant. This represents a change of more than 0.04313 standard deviations (SD[Death Cases Order - Wave 1]= 15.98).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 2.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Death Cases Order - Wave 2 decreases by 7.548 units assuming all other covariates stay constant. This represents a change of more than 0.4587 standard deviations (SD[Death Cases Order - Wave 2]= 16.46).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Death Cases Order - Wave 2 by 0.2408 units, assuming all other covariates stay constant. This represents a change of more than 0.01463 standard deviations (SD[Death Cases Order - Wave 2]= 16.46).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 2 by 0.7076 units, assuming all other covariates stay constant. This represents a change of more than 0.043 standard deviations (SD[Death Cases Order - Wave 2]= 16.46).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Death Cases Order - Wave 2 by 0.03439 units, assuming all other covariates stay constant. This represents a change of more than 0.002089 standard deviations (SD[Death Cases Order - Wave 2]= 16.46).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Death Cases Order - Wave 2 by 0.1267 units, assuming all other covariates stay constant. This represents a change of more than 0.007698 standard deviations (SD[Death Cases Order - Wave 2]= 16.46).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage increases Death Cases Order - Wave 2 by 0.2521 units, assuming all other covariates stay constant. This represents a change of more than 0.01532 standard deviations (SD[Death Cases Order - Wave 2]= 16.46).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Death Cases Order - Wave 2 by 2.421 units, assuming all other covariates stay constant. This represents a change of more than 0.1471 standard deviations (SD[Death Cases Order - Wave 2]= 16.46).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 3.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Death Cases Order - Wave 3 decreases by 10.4 units assuming all other covariates stay constant. This represents a change of more than 0.6267 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Death Cases Order - Wave 3 by 0.1378 units, assuming all other covariates stay constant. This represents a change of more than 0.008305 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Death Cases Order - Wave 3 by 0.315 units, assuming all other covariates stay constant. This represents a change of more than 0.01899 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Death Cases Order - Wave 3 by 3.725 units, assuming all other covariates stay constant. This represents a change of more than 0.2246 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 3 by 1.108 units, assuming all other covariates stay constant. This represents a change of more than 0.06678 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Death Cases Order - Wave 3 by 0.1539 units, assuming all other covariates stay constant. This represents a change of more than 0.009277 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Death Cases Order - Wave 3 by 0.2511 units, assuming all other covariates stay constant. This represents a change of more than 0.01514 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Death Cases Order - Wave 3 by 1.802 units, assuming all other covariates stay constant. This represents a change of more than 0.1086 standard deviations (SD[Death Cases Order - Wave 3]= 16.59).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 4.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Death Cases Order - Wave 4 decreases by 4.991 units assuming all other covariates stay constant. This represents a change of more than 0.3023 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Death Cases Order - Wave 4 by 0.2693 units, assuming all other covariates stay constant. This represents a change of more than 0.01631 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases Death Cases Order - Wave 4 by 0.1726 units, assuming all other covariates stay constant. This represents a change of more than 0.01045 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases Death Cases Order - Wave 4 by 0.001246 units, assuming all other covariates stay constant. This represents a change of more than 7.544e-05 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Death Cases Order - Wave 4 by 9.325 units, assuming all other covariates stay constant. This represents a change of more than 0.5647 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 4 by 0.7042 units, assuming all other covariates stay constant. This represents a change of more than 0.04265 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Death Cases Order - Wave 4 by 0.245 units, assuming all other covariates stay constant. This represents a change of more than 0.01484 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Death Cases Order - Wave 4 by 0.464 units, assuming all other covariates stay constant. This represents a change of more than 0.0281 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Death Cases Order - Wave 4 by 2.503 units, assuming all other covariates stay constant. This represents a change of more than 0.1516 standard deviations (SD[Death Cases Order - Wave 4]= 16.51).

4.6.2 Aligned - Relative Deaths Cases (Waves 1-4) Regression Coefficients

relative_aligned_deaths_waves_reg.png

p1 <- wl_plot_coef (model = m_1[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_2[[1]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_3[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_4[[1]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/relative_aligned_deaths_waves_reg.png", plot = p, height=130)

4.7 Aligned - Relative Excess Mortality Regression models

order_aligned_vars <- c("Death Cases Order - Wave 1" = "ord_w1e", 
                        "Death Cases Order - Wave 2" = "ord_w2e",
                        "Death Cases Order - Wave 3" = "ord_w3e",
                        "Death Cases Order - Wave 4" = "ord_w4e")

cov_34 <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

dep <- order_aligned_vars[1]
m_1 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_1AD"))

dep <- order_aligned_vars[2]
m_2 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_2AD"))

dep <- order_aligned_vars[3]
m_3 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_3AD"))

dep <- order_aligned_vars[4]
m_4 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_4AD"))

4.7.1 Aligned - Relative Excess Mortality Regression Table

relative_aligned_excess_reg.tex

#> 
#> -------------------------------------------------------------------
#>                              Wave 1    Wave 2    Wave 3    Wave 4  
#>                                (1)       (2)       (3)       (4)   
#> -------------------------------------------------------------------
#> Leader Gender[female]       -7.725*** -4.186*** -4.716***   0.148  
#>                              (1.402)   (1.016)   (1.055)   (1.336) 
#> Leader Age                   0.108*    -0.004     0.034    -0.099  
#>                              (0.053)   (0.040)   (0.037)   (0.052) 
#> Stringency Index (delayed)    0.001    -0.036    -0.061*   -0.093* 
#>                              (0.016)   (0.024)   (0.030)   (0.042) 
#> New Vaccination (delayed)                       0.0004***  0.0001  
#>                                                 (0.0001)  (0.0002) 
#> Democracy Score             -4.435***   1.271    -1.115    3.952** 
#>                              (1.333)   (1.032)   (1.023)   (1.353) 
#> GDP per Capita               0.131*     0.052    -0.092*    0.045  
#>                              (0.058)   (0.041)   (0.040)   (0.054) 
#> Health-system Score         -0.424**  -0.707*** -0.716*** -0.682***
#>                              (0.129)   (0.107)   (0.108)   (0.116) 
#> Population Size              0.026**    0.013     0.004    -0.005  
#>                              (0.010)   (0.008)   (0.008)   (0.011) 
#> Urbanization Percentage     0.190***  -0.220***   0.016   -0.121** 
#>                              (0.049)   (0.036)   (0.036)   (0.041) 
#> Immigration Percentage       -0.201*   -0.086   -0.252*** -0.289***
#>                              (0.083)   (0.060)   (0.061)   (0.084) 
#> Population Aged 70 or Older  0.495**   0.412**    0.059    0.483** 
#>                              (0.191)   (0.142)   (0.143)   (0.169) 
#> N                              532       816       804       396   
#> R2                            0.153     0.265     0.288     0.393  
#> Adjusted R2                   0.137     0.256     0.278     0.375  
#> -------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 1.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Death Cases Order - Wave 1 decreases by 7.725 units assuming all other covariates stay constant. This represents a change of more than 0.6776 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age increases Death Cases Order - Wave 1 by 0.1083 units, assuming all other covariates stay constant. This represents a change of more than 0.009504 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Death Cases Order - Wave 1 by 4.435 units, assuming all other covariates stay constant. This represents a change of more than 0.389 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases Death Cases Order - Wave 1 by 0.1309 units, assuming all other covariates stay constant. This represents a change of more than 0.01149 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 1 by 0.4243 units, assuming all other covariates stay constant. This represents a change of more than 0.03722 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Death Cases Order - Wave 1 by 0.02645 units, assuming all other covariates stay constant. This represents a change of more than 0.00232 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Death Cases Order - Wave 1 by 0.1895 units, assuming all other covariates stay constant. This represents a change of more than 0.01662 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Death Cases Order - Wave 1 by 0.2006 units, assuming all other covariates stay constant. This represents a change of more than 0.0176 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Death Cases Order - Wave 1 by 0.4953 units, assuming all other covariates stay constant. This represents a change of more than 0.04344 standard deviations (SD[Death Cases Order - Wave 1]= 11.4).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 2.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Death Cases Order - Wave 2 decreases by 4.186 units assuming all other covariates stay constant. This represents a change of more than 0.3564 standard deviations (SD[Death Cases Order - Wave 2]= 11.75).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 2 by 0.7068 units, assuming all other covariates stay constant. This represents a change of more than 0.06017 standard deviations (SD[Death Cases Order - Wave 2]= 11.75).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Death Cases Order - Wave 2 by 0.2196 units, assuming all other covariates stay constant. This represents a change of more than 0.01869 standard deviations (SD[Death Cases Order - Wave 2]= 11.75).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Death Cases Order - Wave 2 by 0.4122 units, assuming all other covariates stay constant. This represents a change of more than 0.03509 standard deviations (SD[Death Cases Order - Wave 2]= 11.75).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 3.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Death Cases Order - Wave 3 decreases by 4.716 units assuming all other covariates stay constant. This represents a change of more than 0.4089 standard deviations (SD[Death Cases Order - Wave 3]= 11.53).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Death Cases Order - Wave 3 by 0.0612 units, assuming all other covariates stay constant. This represents a change of more than 0.005307 standard deviations (SD[Death Cases Order - Wave 3]= 11.53).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases Death Cases Order - Wave 3 by 0.0004461 units, assuming all other covariates stay constant. This represents a change of more than 3.868e-05 standard deviations (SD[Death Cases Order - Wave 3]= 11.53).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita decreases Death Cases Order - Wave 3 by 0.09225 units, assuming all other covariates stay constant. This represents a change of more than 0.008 standard deviations (SD[Death Cases Order - Wave 3]= 11.53).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 3 by 0.7155 units, assuming all other covariates stay constant. This represents a change of more than 0.06205 standard deviations (SD[Death Cases Order - Wave 3]= 11.53).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Death Cases Order - Wave 3 by 0.2515 units, assuming all other covariates stay constant. This represents a change of more than 0.02181 standard deviations (SD[Death Cases Order - Wave 3]= 11.53).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Death Cases Order - Wave 4.
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Death Cases Order - Wave 4 by 0.09292 units, assuming all other covariates stay constant. This represents a change of more than 0.008138 standard deviations (SD[Death Cases Order - Wave 4]= 11.42).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Death Cases Order - Wave 4 by 3.952 units, assuming all other covariates stay constant. This represents a change of more than 0.3461 standard deviations (SD[Death Cases Order - Wave 4]= 11.42).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Death Cases Order - Wave 4 by 0.6817 units, assuming all other covariates stay constant. This represents a change of more than 0.0597 standard deviations (SD[Death Cases Order - Wave 4]= 11.42).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Death Cases Order - Wave 4 by 0.1212 units, assuming all other covariates stay constant. This represents a change of more than 0.01061 standard deviations (SD[Death Cases Order - Wave 4]= 11.42).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Death Cases Order - Wave 4 by 0.2889 units, assuming all other covariates stay constant. This represents a change of more than 0.0253 standard deviations (SD[Death Cases Order - Wave 4]= 11.42).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases Death Cases Order - Wave 4 by 0.4833 units, assuming all other covariates stay constant. This represents a change of more than 0.04233 standard deviations (SD[Death Cases Order - Wave 4]= 11.42).

4.7.2 Aligned - Relative Excess Mortality (Waves 1-4) Regression Coefficients

relative_aligned_excess_waves_reg.png

p1 <- wl_plot_coef (model = m_1[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_2[[1]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_3[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_4[[1]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/relative_aligned_excess_waves_reg.png", plot = p, height=130)

4.8 Aligned Relative Performance - Leader Gender Coefficient Summary

relative_aligned_gender_coef_summary.png

title1 <- textGrob(names(wl_dependents())[1], gp = gpar(fontsize = 14))
title2 <- textGrob(names(wl_dependents())[2], gp = gpar(fontsize = 14))
title3 <- textGrob(names(wl_dependents())[3], gp = gpar(fontsize = 14))
title4 <- textGrob("Leader Gender Coefficient",gp = gpar(fontsize = 14))
tblank <- textGrob(" ",gp = gpar(fontsize = 14))

p <- ggarrange(title1, title2, title3, 
               p_cases, p_deaths, p_excess, 
               tblank,title4, tblank,
               font.label = list(size = 10, face = "bold"),
               heights = c(0.1, 1, 0.1),
               ncol = 3, nrow = 3) 

wl_ggsave("figures/relative_aligned_gender_coef_summary.png", plot = p, width=250)

5 Spurious Claims

5.1 Scandinavia / Northern EU Correlations

dep <- wl_dependents()[1]
cov_fs <- c(wl_covariates(cov_group = "Main coviariates"), "Scandinavia" = "scandinavia")
cov_fn <- c(wl_covariates(cov_group = "Main coviariates"), "Northern EU" = "northern_eu")

m_fcs <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fs, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

m_fcn <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fn, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

dep <- wl_dependents()[2]

m_fds <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fs, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

m_fdn <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fn, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

dep <- wl_dependents()[3]

m_fes <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fs, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

m_fen <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fn, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

5.1.1 Scandinavia Regression Table

scandinavia_reg.tex

#> 
#> --------------------------------------------------------------------------------
#>                             New Infection Cases New Death Cases Excess Mortality
#>                                     (1)               (2)             (3)       
#> --------------------------------------------------------------------------------
#> Leader Gender[female]           -40.058***         -0.631***        -2.513*     
#>                                  (10.047)           (0.171)         (1.018)     
#> Leader Age                        -0.486           -0.018***         0.026      
#>                                   (0.309)           (0.005)         (0.035)     
#> Stringency Index (delayed)       0.628***          0.014***        -0.180***    
#>                                   (0.158)           (0.003)         (0.018)     
#> New Vaccination (delayed)         0.002*          -0.0001***        -0.0001     
#>                                   (0.001)          (0.00002)        (0.0001)    
#> Democracy Score                    2.606           0.642***         3.033**     
#>                                   (8.078)           (0.138)         (0.931)     
#> GDP per Capita                    0.988**            0.009           0.097*     
#>                                   (0.332)           (0.006)         (0.041)     
#> Health-system Score               -1.407           -0.066***       -0.782***    
#>                                   (0.791)           (0.014)         (0.096)     
#> Population Size                   -0.072             0.001           -0.010     
#>                                   (0.064)           (0.001)         (0.007)     
#> Urbanization Percentage            0.383            -0.007          -0.076*     
#>                                   (0.255)           (0.004)         (0.035)     
#> Immigration Percentage            -0.511           -0.036***       -0.410***    
#>                                   (0.540)           (0.009)         (0.061)     
#> Population Aged 70 or Older      4.494***          0.222***         0.459***    
#>                                   (1.056)           (0.018)         (0.130)     
#> Scandinavia                     -50.944***         -1.009***       -6.102***    
#>                                  (13.458)           (0.229)         (1.376)     
#> N                                  4,337             4,864           3,571      
#> R2                                 0.030             0.083           0.143      
#> Adjusted R2                        0.027             0.081           0.140      
#> --------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001

5.1.2 Northern EU Regression Table

northern_eu_reg.tex

#> 
#> --------------------------------------------------------------------------------
#>                             New Infection Cases New Death Cases Excess Mortality
#>                                     (1)               (2)             (3)       
#> --------------------------------------------------------------------------------
#> Leader Gender[female]           -69.630***         -0.939***       -3.571***    
#>                                   (9.473)           (0.163)         (0.986)     
#> Leader Age                        -0.308           -0.016**          0.029      
#>                                   (0.309)           (0.005)         (0.035)     
#> Stringency Index (delayed)       0.641***          0.014***        -0.180***    
#>                                   (0.158)           (0.003)         (0.018)     
#> New Vaccination (delayed)         0.002*          -0.0001***        -0.0001     
#>                                   (0.001)          (0.00002)        (0.0001)    
#> Democracy Score                    4.623           0.637***         3.117***    
#>                                   (8.102)           (0.139)         (0.933)     
#> GDP per Capita                     0.044            -0.0001          0.055      
#>                                   (0.332)           (0.006)         (0.039)     
#> Health-system Score               -0.532           -0.065***       -0.832***    
#>                                   (0.816)           (0.014)         (0.101)     
#> Population Size                    0.038             0.002           -0.006     
#>                                   (0.064)           (0.001)         (0.007)     
#> Urbanization Percentage            0.031           -0.011**         -0.104**    
#>                                   (0.250)           (0.004)         (0.034)     
#> Immigration Percentage             0.775           -0.024**        -0.337***    
#>                                   (0.532)           (0.009)         (0.058)     
#> Population Aged 70 or Older      4.566***          0.234***         0.577***    
#>                                   (1.056)           (0.018)         (0.133)     
#> Northern EU                      31.301***          -0.111          -2.475*     
#>                                   (9.468)           (0.163)         (0.979)     
#> N                                  4,337             4,864           3,571      
#> R2                                 0.029             0.080           0.140      
#> Adjusted R2                        0.027             0.078           0.137      
#> --------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001

5.2 State Capacity Analysis

5.2.1 Capacity - Infections, Deaths and Excess Mortality (Full)

cov_f <- wl_covariates(cov_group = "Capacity covariates")

m_fc <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[1], x_var = cov_f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

m_fd <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[2], x_var = cov_f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

m_fe <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[3], x_var = cov_f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

5.2.1.1 Capacity - Full Regression Table

capacity_reg.tex

#> 
#> -------------------------------------------------------------------------------
#>                            New Infection Cases New Death Cases Excess Mortality
#>                                    (1)               (2)             (3)       
#> -------------------------------------------------------------------------------
#> Leader Gender[female]          -49.960***         -1.033***         -1.857     
#>                                  (9.692)           (0.171)         (1.014)     
#> Leader Age                      -0.903**          -0.030***         -0.027     
#>                                  (0.323)           (0.006)         (0.037)     
#> Stringency Index (delayed)      0.615***          0.013***        -0.176***    
#>                                  (0.165)           (0.003)         (0.018)     
#> New Vaccination (delayed)        0.003*          -0.0001***        -0.0002     
#>                                  (0.001)          (0.00002)        (0.0001)    
#> Democracy Score                 -17.250*            0.162           1.347      
#>                                  (7.981)           (0.141)         (0.930)     
#> Population Size                   0.098           0.005***         0.022***    
#>                                  (0.062)           (0.001)         (0.007)     
#> Urbanization Percentage           0.361           -0.013**        -0.168***    
#>                                  (0.259)           (0.005)         (0.035)     
#> Immigration Percentage            0.990           -0.042***       -0.275***    
#>                                  (0.531)           (0.009)         (0.056)     
#> State capacity                   -11.646            0.200         -6.744***    
#>                                  (8.221)           (0.144)         (1.084)     
#> N                                 4,018             4,505           3,279      
#> R2                                0.019             0.042           0.128      
#> Adjusted R2                       0.016             0.040           0.126      
#> -------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Capacity: New Infections interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 49.96 units assuming all other covariates stay constant. This represents a change of more than 0.2633 standard deviations (SDNew Infection Cases= 189.7).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 0.9027 units, assuming all other covariates stay constant. This represents a change of more than 0.004758 standard deviations (SDNew Infection Cases= 189.7).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 0.6148 units, assuming all other covariates stay constant. This represents a change of more than 0.003241 standard deviations (SDNew Infection Cases= 189.7).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases New Infection Cases by 0.002649 units, assuming all other covariates stay constant. This represents a change of more than 1.396e-05 standard deviations (SDNew Infection Cases= 189.7).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 17.25 units, assuming all other covariates stay constant. This represents a change of more than 0.09092 standard deviations (SDNew Infection Cases= 189.7).

Capacity: New Deaths interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.033 units assuming all other covariates stay constant. This represents a change of more than 0.2939 standard deviations (SDNew Death Cases= 3.516).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.02971 units, assuming all other covariates stay constant. This represents a change of more than 0.00845 standard deviations (SDNew Death Cases= 3.516).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.01304 units, assuming all other covariates stay constant. This represents a change of more than 0.003709 standard deviations (SDNew Death Cases= 3.516).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 8.837e-05 units, assuming all other covariates stay constant. This represents a change of more than 2.513e-05 standard deviations (SDNew Death Cases= 3.516).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.004591 units, assuming all other covariates stay constant. This represents a change of more than 0.001306 standard deviations (SDNew Death Cases= 3.516).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.01297 units, assuming all other covariates stay constant. This represents a change of more than 0.00369 standard deviations (SDNew Death Cases= 3.516).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.04244 units, assuming all other covariates stay constant. This represents a change of more than 0.01207 standard deviations (SDNew Death Cases= 3.516).

Capacity: Excess Mortality interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1756 units, assuming all other covariates stay constant. This represents a change of more than 0.008701 standard deviations (SDExcess Mortality= 20.19).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.02195 units, assuming all other covariates stay constant. This represents a change of more than 0.001087 standard deviations (SDExcess Mortality= 20.19).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.1682 units, assuming all other covariates stay constant. This represents a change of more than 0.008332 standard deviations (SDExcess Mortality= 20.19).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.2749 units, assuming all other covariates stay constant. This represents a change of more than 0.01362 standard deviations (SDExcess Mortality= 20.19).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases Excess Mortality by 6.744 units, assuming all other covariates stay constant. This represents a change of more than 0.3341 standard deviations (SDExcess Mortality= 20.19).

5.2.1.2 Capacity - Full Regression Coefficients plot

capacity_full_reg.png

p_fc <- wl_plot_coef (model = m_fc[[1]], cov_f, palette = "Color")
p_fd <- wl_plot_coef (model = m_fd[[1]], cov_f, palette = "Color")
p_fe <- wl_plot_coef (model = m_fe[[1]], cov_f, palette = "Color")

p <- ggarrange(p_fc, p_fd, p_fe,  
               labels = names(wl_dependents()), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               hjust = -0.15,  ncol = 3, nrow = 1) 

wl_ggsave("figures/capacity_full_reg.png", plot = p, width=250)

5.2.2 Capacity - New Infection cases

dep <- wl_dependents()[1]
cov_34f <- wl_covariates(cov_group = "Capacity covariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("Wave_1C", "Wave_2C"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("Wave_3C", "Wave_4C"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

5.2.2.1 Capacity - Infection Cases Regression Table

capacity_cases_reg.tex

#> 
#> ----------------------------------------------------------------------------------
#>                             Wave 1     Wave 2     Wave 3      Wave 4       Full   
#>                               (1)       (2)         (3)         (4)        (5)    
#> ----------------------------------------------------------------------------------
#> Leader Gender[female]      -13.932** -98.362***  -46.504*     -29.980   -49.960***
#>                             (4.368)   (20.705)   (23.110)    (17.574)    (9.692)  
#> Leader Age                  -0.166     1.162     -2.792***   -1.574**    -0.903** 
#>                             (0.162)   (0.671)     (0.722)     (0.605)    (0.323)  
#> Stringency Index (delayed)  -0.121*   2.030***   2.469***    -1.913***   0.615*** 
#>                             (0.050)   (0.402)     (0.525)     (0.435)    (0.165)  
#> New Vaccination (delayed)                        0.014***    -0.006**     0.003*  
#>                                                   (0.003)     (0.002)    (0.001)  
#> Democracy Score             -8.009*   -32.595*  -100.984***    1.146     -17.250* 
#>                             (3.985)   (16.225)   (18.234)    (15.263)    (7.981)  
#> Population Size            0.120***    0.129      -0.106       0.089      0.098   
#>                             (0.026)   (0.128)     (0.143)     (0.120)    (0.062)  
#> Urbanization Percentage     -0.105   -2.476***   2.417***    1.846***     0.361   
#>                             (0.129)   (0.534)     (0.587)     (0.478)    (0.259)  
#> Immigration Percentage       0.459     -1.077    -3.768**    5.960***     0.990   
#>                             (0.240)   (1.116)     (1.217)     (0.991)    (0.531)  
#> State capacity               6.325   104.774*** -65.859***  -103.750***  -11.646  
#>                             (4.164)   (16.871)   (18.640)    (15.883)    (8.221)  
#> N                             343      1,225        784         980       4,018   
#> R2                           0.182     0.088       0.148       0.114      0.019   
#> Adjusted R2                  0.162     0.082       0.138       0.106      0.016   
#> ----------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 13.93 units assuming all other covariates stay constant. This represents a change of more than 0.4413 standard deviations (SDNew Infection Cases= 31.57).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Infection Cases by 0.1209 units, assuming all other covariates stay constant. This represents a change of more than 0.00383 standard deviations (SDNew Infection Cases= 31.57).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 8.009 units, assuming all other covariates stay constant. This represents a change of more than 0.2537 standard deviations (SDNew Infection Cases= 31.57).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Infection Cases by 0.1199 units, assuming all other covariates stay constant. This represents a change of more than 0.003799 standard deviations (SDNew Infection Cases= 31.57).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 98.36 units assuming all other covariates stay constant. This represents a change of more than 0.4379 standard deviations (SDNew Infection Cases= 224.6).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 2.03 units, assuming all other covariates stay constant. This represents a change of more than 0.009036 standard deviations (SDNew Infection Cases= 224.6).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 32.59 units, assuming all other covariates stay constant. This represents a change of more than 0.1451 standard deviations (SDNew Infection Cases= 224.6).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Infection Cases by 2.476 units, assuming all other covariates stay constant. This represents a change of more than 0.01102 standard deviations (SDNew Infection Cases= 224.6).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity increases New Infection Cases by 104.8 units, assuming all other covariates stay constant. This represents a change of more than 0.4664 standard deviations (SDNew Infection Cases= 224.6).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 46.5 units assuming all other covariates stay constant. This represents a change of more than 0.2293 standard deviations (SDNew Infection Cases= 202.8).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 2.792 units, assuming all other covariates stay constant. This represents a change of more than 0.01377 standard deviations (SDNew Infection Cases= 202.8).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 2.469 units, assuming all other covariates stay constant. This represents a change of more than 0.01218 standard deviations (SDNew Infection Cases= 202.8).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases New Infection Cases by 0.01377 units, assuming all other covariates stay constant. This represents a change of more than 6.791e-05 standard deviations (SDNew Infection Cases= 202.8).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 101 units, assuming all other covariates stay constant. This represents a change of more than 0.498 standard deviations (SDNew Infection Cases= 202.8).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 2.417 units, assuming all other covariates stay constant. This represents a change of more than 0.01192 standard deviations (SDNew Infection Cases= 202.8).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Infection Cases by 3.768 units, assuming all other covariates stay constant. This represents a change of more than 0.01858 standard deviations (SDNew Infection Cases= 202.8).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases New Infection Cases by 65.86 units, assuming all other covariates stay constant. This represents a change of more than 0.3248 standard deviations (SDNew Infection Cases= 202.8).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 1.574 units, assuming all other covariates stay constant. This represents a change of more than 0.008383 standard deviations (SDNew Infection Cases= 187.8).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Infection Cases by 1.913 units, assuming all other covariates stay constant. This represents a change of more than 0.01019 standard deviations (SDNew Infection Cases= 187.8).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Infection Cases by 0.006237 units, assuming all other covariates stay constant. This represents a change of more than 3.322e-05 standard deviations (SDNew Infection Cases= 187.8).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 1.846 units, assuming all other covariates stay constant. This represents a change of more than 0.009832 standard deviations (SDNew Infection Cases= 187.8).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage increases New Infection Cases by 5.96 units, assuming all other covariates stay constant. This represents a change of more than 0.03174 standard deviations (SDNew Infection Cases= 187.8).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases New Infection Cases by 103.7 units, assuming all other covariates stay constant. This represents a change of more than 0.5526 standard deviations (SDNew Infection Cases= 187.8).

Full period interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 49.96 units assuming all other covariates stay constant. This represents a change of more than 0.2633 standard deviations (SDNew Infection Cases= 189.7).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 0.9027 units, assuming all other covariates stay constant. This represents a change of more than 0.004758 standard deviations (SDNew Infection Cases= 189.7).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 0.6148 units, assuming all other covariates stay constant. This represents a change of more than 0.003241 standard deviations (SDNew Infection Cases= 189.7).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases New Infection Cases by 0.002649 units, assuming all other covariates stay constant. This represents a change of more than 1.396e-05 standard deviations (SDNew Infection Cases= 189.7).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 17.25 units, assuming all other covariates stay constant. This represents a change of more than 0.09092 standard deviations (SDNew Infection Cases= 189.7).

5.2.2.2 Capacity - Infection Cases (Waves 1-4) Regression Coefficients

capacity_cases_waves_reg.png

# 
p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/capacity_cases_waves_reg.png", plot = p, height=130)

5.2.2.3 Capacity - New Death Cases Regression models

dep <- wl_dependents()[2]
cov_34f <- wl_covariates(cov_group = "Capacity covariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_1D", "Wave_2D"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_3D", "Wave_4D"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

5.2.2.4 Capacity - Deaths Cases Regression Table

capacity_deaths_reg.tex

#> 
#> ------------------------------------------------------------------------------
#>                             Wave 1    Wave 2    Wave 3     Wave 4      Full   
#>                               (1)       (2)       (3)       (4)        (5)    
#> ------------------------------------------------------------------------------
#> Leader Gender[female]      -1.776*** -1.874***  -0.743*    -0.268   -1.033*** 
#>                             (0.332)   (0.424)   (0.360)   (0.289)    (0.171)  
#> Leader Age                  -0.017     0.013   -0.068***  -0.027**  -0.030*** 
#>                             (0.012)   (0.014)   (0.011)   (0.010)    (0.006)  
#> Stringency Index (delayed) -0.015*** 0.048***  0.060***    0.013     0.013*** 
#>                             (0.004)   (0.008)   (0.008)   (0.007)    (0.003)  
#> New Vaccination (delayed)                       0.00001  -0.0003*** -0.0001***
#>                                                (0.00004) (0.00003)  (0.00002) 
#> Democracy Score            -0.935**   -0.771*    0.118    0.952***    0.162   
#>                             (0.288)   (0.333)   (0.284)   (0.244)    (0.141)  
#> Population Size            0.010***   0.005*     0.004     -0.001    0.005*** 
#>                             (0.002)   (0.003)   (0.002)   (0.002)    (0.001)  
#> Urbanization Percentage      0.006   -0.079*** 0.053***  -0.033***   -0.013** 
#>                             (0.009)   (0.011)   (0.009)   (0.008)    (0.005)  
#> Immigration Percentage      0.037*   -0.089*** -0.117***  -0.044**  -0.042*** 
#>                             (0.018)   (0.023)   (0.019)   (0.016)    (0.009)  
#> State capacity               0.164   2.515***  -1.279***   0.136      0.200   
#>                             (0.301)   (0.340)   (0.292)   (0.260)    (0.144)  
#> N                             585      1,078      931      1,274      4,505   
#> R2                           0.173     0.128     0.217     0.181      0.042   
#> Adjusted R2                  0.161     0.121     0.210     0.175      0.040   
#> ------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.776 units assuming all other covariates stay constant. This represents a change of more than 0.7078 standard deviations (SDNew Death Cases= 2.51).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Death Cases by 0.01472 units, assuming all other covariates stay constant. This represents a change of more than 0.005865 standard deviations (SDNew Death Cases= 2.51).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Death Cases by 0.935 units, assuming all other covariates stay constant. This represents a change of more than 0.3726 standard deviations (SDNew Death Cases= 2.51).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.01008 units, assuming all other covariates stay constant. This represents a change of more than 0.004015 standard deviations (SDNew Death Cases= 2.51).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage increases New Death Cases by 0.03709 units, assuming all other covariates stay constant. This represents a change of more than 0.01478 standard deviations (SDNew Death Cases= 2.51).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.874 units assuming all other covariates stay constant. This represents a change of more than 0.432 standard deviations (SDNew Death Cases= 4.338).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.04829 units, assuming all other covariates stay constant. This represents a change of more than 0.01113 standard deviations (SDNew Death Cases= 4.338).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Death Cases by 0.7714 units, assuming all other covariates stay constant. This represents a change of more than 0.1778 standard deviations (SDNew Death Cases= 4.338).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.00519 units, assuming all other covariates stay constant. This represents a change of more than 0.001196 standard deviations (SDNew Death Cases= 4.338).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.0785 units, assuming all other covariates stay constant. This represents a change of more than 0.01809 standard deviations (SDNew Death Cases= 4.338).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.08913 units, assuming all other covariates stay constant. This represents a change of more than 0.02054 standard deviations (SDNew Death Cases= 4.338).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity increases New Death Cases by 2.515 units, assuming all other covariates stay constant. This represents a change of more than 0.5796 standard deviations (SDNew Death Cases= 4.338).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 0.7428 units assuming all other covariates stay constant. This represents a change of more than 0.2072 standard deviations (SDNew Death Cases= 3.586).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.06766 units, assuming all other covariates stay constant. This represents a change of more than 0.01887 standard deviations (SDNew Death Cases= 3.586).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.06 units, assuming all other covariates stay constant. This represents a change of more than 0.01674 standard deviations (SDNew Death Cases= 3.586).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Death Cases by 0.05282 units, assuming all other covariates stay constant. This represents a change of more than 0.01473 standard deviations (SDNew Death Cases= 3.586).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.1166 units, assuming all other covariates stay constant. This represents a change of more than 0.03251 standard deviations (SDNew Death Cases= 3.586).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases New Death Cases by 1.279 units, assuming all other covariates stay constant. This represents a change of more than 0.3567 standard deviations (SDNew Death Cases= 3.586).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.02668 units, assuming all other covariates stay constant. This represents a change of more than 0.007754 standard deviations (SDNew Death Cases= 3.441).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 0.0002866 units, assuming all other covariates stay constant. This represents a change of more than 8.33e-05 standard deviations (SDNew Death Cases= 3.441).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases New Death Cases by 0.9517 units, assuming all other covariates stay constant. This represents a change of more than 0.2766 standard deviations (SDNew Death Cases= 3.441).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.0328 units, assuming all other covariates stay constant. This represents a change of more than 0.009534 standard deviations (SDNew Death Cases= 3.441).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.04399 units, assuming all other covariates stay constant. This represents a change of more than 0.01279 standard deviations (SDNew Death Cases= 3.441).

Full period interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.033 units assuming all other covariates stay constant. This represents a change of more than 0.2939 standard deviations (SDNew Death Cases= 3.516).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.02971 units, assuming all other covariates stay constant. This represents a change of more than 0.00845 standard deviations (SDNew Death Cases= 3.516).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.01304 units, assuming all other covariates stay constant. This represents a change of more than 0.003709 standard deviations (SDNew Death Cases= 3.516).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 8.837e-05 units, assuming all other covariates stay constant. This represents a change of more than 2.513e-05 standard deviations (SDNew Death Cases= 3.516).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.004591 units, assuming all other covariates stay constant. This represents a change of more than 0.001306 standard deviations (SDNew Death Cases= 3.516).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.01297 units, assuming all other covariates stay constant. This represents a change of more than 0.00369 standard deviations (SDNew Death Cases= 3.516).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.04244 units, assuming all other covariates stay constant. This represents a change of more than 0.01207 standard deviations (SDNew Death Cases= 3.516).

5.2.2.5 Capacity - Deaths Cases (Waves 1-4) Regression Coefficients

capacity_deaths_waves_reg.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/capacity_deaths_waves_reg.png", plot = p, height=130)

5.2.2.6 Capacity - Excess Mortality Regression models

dep <- wl_dependents()[3]
cov_34f <- wl_covariates(cov_group = "Capacity covariates")
cov_12 <- wl_setdiff(cov_34f, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_1D", "Wave_2D"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("Wave_3D", "Wave_4D"))

m_f <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34f, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

5.2.2.7 Capacity - Excess Mortality Regression Table

capacity_excess_reg.tex

#> 
#> ------------------------------------------------------------------------------
#>                              Wave 1    Wave 2     Wave 3    Wave 4     Full   
#>                               (1)        (2)       (3)        (4)       (5)   
#> ------------------------------------------------------------------------------
#> Leader Gender[female]      -14.728***  -5.053*    0.950     3.474*    -1.857  
#>                             (2.853)    (2.565)   (1.676)    (1.727)   (1.014) 
#> Leader Age                   -0.011    0.226*    -0.141*    -0.052    -0.027  
#>                             (0.103)    (0.091)   (0.059)    (0.065)   (0.037) 
#> Stringency Index (delayed) -0.114***  -0.284***   -0.013    -0.068   -0.176***
#>                             (0.030)    (0.054)   (0.046)    (0.050)   (0.018) 
#> New Vaccination (delayed)                        0.0005*   -0.002***  -0.0002 
#>                                                  (0.0002)  (0.0002)  (0.0001) 
#> Democracy Score            -12.220***  -1.323     0.900     5.313**    1.347  
#>                             (2.551)    (2.246)   (1.548)    (1.649)   (0.930) 
#> Population Size             0.058***   0.037*     -0.005    -0.009   0.022*** 
#>                             (0.016)    (0.016)   (0.011)    (0.012)   (0.007) 
#> Urbanization Percentage     0.280**   -0.595***  0.149**   -0.222*** -0.168***
#>                             (0.096)    (0.085)   (0.057)    (0.060)   (0.035) 
#> Immigration Percentage       -0.115   -0.438**  -0.387***  -0.264**  -0.275***
#>                             (0.149)    (0.136)   (0.095)    (0.099)   (0.056) 
#> State capacity               1.169     -2.675   -16.807*** -7.100*** -6.744***
#>                             (3.036)    (2.603)   (1.815)    (1.946)   (1.084) 
#> N                             450        780       663        909      3,279  
#> R2                           0.174      0.201     0.311      0.295     0.128  
#> Adjusted R2                  0.159      0.192     0.302      0.288     0.126  
#> ------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 14.73 units assuming all other covariates stay constant. This represents a change of more than 0.7698 standard deviations (SDExcess Mortality= 19.13).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1143 units, assuming all other covariates stay constant. This represents a change of more than 0.005973 standard deviations (SDExcess Mortality= 19.13).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Excess Mortality by 12.22 units, assuming all other covariates stay constant. This represents a change of more than 0.6387 standard deviations (SDExcess Mortality= 19.13).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.05781 units, assuming all other covariates stay constant. This represents a change of more than 0.003021 standard deviations (SDExcess Mortality= 19.13).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Excess Mortality by 0.2805 units, assuming all other covariates stay constant. This represents a change of more than 0.01466 standard deviations (SDExcess Mortality= 19.13).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 5.053 units assuming all other covariates stay constant. This represents a change of more than 0.2022 standard deviations (SDExcess Mortality= 24.99).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age increases Excess Mortality by 0.2264 units, assuming all other covariates stay constant. This represents a change of more than 0.00906 standard deviations (SDExcess Mortality= 24.99).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.2845 units, assuming all other covariates stay constant. This represents a change of more than 0.01139 standard deviations (SDExcess Mortality= 24.99).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.03732 units, assuming all other covariates stay constant. This represents a change of more than 0.001494 standard deviations (SDExcess Mortality= 24.99).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.5955 units, assuming all other covariates stay constant. This represents a change of more than 0.02383 standard deviations (SDExcess Mortality= 24.99).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.4376 units, assuming all other covariates stay constant. This represents a change of more than 0.01751 standard deviations (SDExcess Mortality= 24.99).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases Excess Mortality by 0.1409 units, assuming all other covariates stay constant. This represents a change of more than 0.00844 standard deviations (SDExcess Mortality= 16.69).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) increases Excess Mortality by 0.0004678 units, assuming all other covariates stay constant. This represents a change of more than 2.803e-05 standard deviations (SDExcess Mortality= 16.69).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Excess Mortality by 0.1487 units, assuming all other covariates stay constant. This represents a change of more than 0.008912 standard deviations (SDExcess Mortality= 16.69).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.3872 units, assuming all other covariates stay constant. This represents a change of more than 0.0232 standard deviations (SDExcess Mortality= 16.69).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases Excess Mortality by 16.81 units, assuming all other covariates stay constant. This represents a change of more than 1.007 standard deviations (SDExcess Mortality= 16.69).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality increases by 3.474 units assuming all other covariates stay constant. This represents a change of more than 0.1735 standard deviations (SDExcess Mortality= 20.02).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Excess Mortality by 0.001816 units, assuming all other covariates stay constant. This represents a change of more than 9.073e-05 standard deviations (SDExcess Mortality= 20.02).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Excess Mortality by 5.313 units, assuming all other covariates stay constant. This represents a change of more than 0.2654 standard deviations (SDExcess Mortality= 20.02).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.222 units, assuming all other covariates stay constant. This represents a change of more than 0.01109 standard deviations (SDExcess Mortality= 20.02).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.2641 units, assuming all other covariates stay constant. This represents a change of more than 0.0132 standard deviations (SDExcess Mortality= 20.02).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases Excess Mortality by 7.1 units, assuming all other covariates stay constant. This represents a change of more than 0.3547 standard deviations (SDExcess Mortality= 20.02).

Full period interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1756 units, assuming all other covariates stay constant. This represents a change of more than 0.008701 standard deviations (SDExcess Mortality= 20.19).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.02195 units, assuming all other covariates stay constant. This represents a change of more than 0.001087 standard deviations (SDExcess Mortality= 20.19).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.1682 units, assuming all other covariates stay constant. This represents a change of more than 0.008332 standard deviations (SDExcess Mortality= 20.19).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.2749 units, assuming all other covariates stay constant. This represents a change of more than 0.01362 standard deviations (SDExcess Mortality= 20.19).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases Excess Mortality by 6.744 units, assuming all other covariates stay constant. This represents a change of more than 0.3341 standard deviations (SDExcess Mortality= 20.19).

5.2.2.8 Capacity - Excess Mortality (Waves 1-4) Regression Coefficients

capacity_excess_waves_reg.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34f, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34f, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/capacity_excess_waves_reg.png", plot = p, height=130)

6 Potential Mechanisms

6.1 Public Support

# Important: 
#   change saverds to TRUE in order to re-calculate all regression models
#   change saverds to FALSE in order to save time when re-running 
  models <- wl_polls_ml_olm(out = NULL, saverds = FALSE)

6.1.1 Public support probability graphs

pub_support_reg.tex

s1q1 s2q1 s3q1 s1q2 s2q2 s3q2
Totally oppose|Tend to oppose -19.7*** -10.3*** -5.53***
Tend to oppose|Tend to support -18.3*** -8.78*** -4.02***
Tend to support|Totally support -15.9*** -6.31*** -1.55
pm_genderFemale -0.0773 0.315*** 0.34*** 0.695*** 1.05*** 0.964***
pm_age -0.0165*** -0.00382* -0.00676*** -0.00462** -0.00177 0.00577***
stringency_delayed 0.00504 0.00914*** 0.015*** 0.0275*** 0.0175*** 0.0207***
fh_score -1.24*** -1*** -0.732*** -0.818*** -0.888*** -0.197**
gdp_per_capita 0.0569*** 0.0108*** 0.0129*** 0.0273*** -0.00386 0.0138***
lpi_health_score -0.259*** -0.107*** -0.0612*** -0.145*** -0.0824*** 0.0051
population -0.00248* -7.85e-05 -0.00221* -0.0209*** -0.0123*** -0.0132***
wdi_urbanization 0.0159*** 0.0167*** 0.0108*** 0.00219 0.00293 0.00167
immigrants_pct -0.047*** 0.0102** 0.00372 -0.00761 0.0301*** 0.0122***
aged_70_older 0.161*** -0.0523*** 0.0201* 0.0649*** -0.117*** 0.0515***
Not at all satisfied|Not very satisfied -11.6*** -9.96*** 1.05
Not very satisfied|Fairly satisfied -10.2*** -8.54*** 2.61**
Fairly satisfied|Very satisfied -7.92*** -6.28*** 5***
SD (Intercept age_r) 0.0737 0.529 0.458 0.0748 0.434 0.5
SD (Intercept gender) 0.064 0 0.0422 0.121 0.314 0
SD (Intercept social_grade) 0.0265 0.0422 0.0486 0.0652 0.0458 0.0233
SD (Intercept work_status) 0.101 0.107 0.0955 0.166 0.0298 0.0999
N 17593 20255 20177 17593 20255 20177
* p < 0.05, ** p < 0.01, *** p < 0.001

6.1.2 Public support probabilities plot

pub_support_prob.png

p <- wl_public_support_plot_probs(models)
wl_ggsave("figures/pub_support_prob.png", plot = p, height=50)

6.2 Effectiveness (Interaction)

6.2.1 Effectiveness (Health System Score) Regression models

cov_if <- wl_covariates(cov_group = "Main coviariates")
cov_if <- c(cov_if[2:length(cov_if)], cov_if[1]) # Re-arrange covariates so we have pm_gender last interaction term for plot_model
interaction <- c("Health System Score x Gender" = "pm_gender*lpi_health_score") #"capacity*pm_gender"

m_ic <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[1], x_var = cov_if, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"), interaction=interaction, arrange=FALSE)

m_id <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[2], x_var = cov_if, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"), interaction=interaction, arrange=FALSE)

m_ie <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[3], x_var = cov_if, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"), interaction=interaction, arrange=FALSE)

6.2.2 Effectiveness (Health System Score) Regression Table

effectiveness_health_reg.tex

#> 
#> -------------------------------------------------------------------------
#>                              Infection Cases Death Cases Excess Mortality
#>                                    (1)           (2)           (3)       
#> -------------------------------------------------------------------------
#> Leader Age                        0.176        -0.010         0.065      
#>                                  (0.308)       (0.005)       (0.035)     
#> Stringency Index (delayed)      0.522***      0.014***      -0.176***    
#>                                  (0.156)       (0.003)       (0.017)     
#> New Vaccination (delayed)         0.001      -0.0001***      -0.0002     
#>                                  (0.001)      (0.00002)      (0.0001)    
#> Democracy Score                  -5.267       0.558***       2.563**     
#>                                  (7.995)       (0.138)       (0.930)     
#> GDP per Capita                   0.721*        0.0003         0.056      
#>                                  (0.301)       (0.005)       (0.037)     
#> Health-system Score               1.120       -0.037**      -0.611***    
#>                                  (0.804)       (0.014)       (0.098)     
#> Population Size                  -0.022         0.002         -0.002     
#>                                  (0.062)       (0.001)       (0.007)     
#> Urbanization Percentage           0.205       -0.011**      -0.116***    
#>                                  (0.244)       (0.004)       (0.033)     
#> Immigration Percentage           -0.354       -0.027**      -0.339***    
#>                                  (0.500)       (0.009)       (0.054)     
#> Population Aged 70 or Older       1.394       0.193***        0.277*     
#>                                  (1.078)       (0.019)       (0.133)     
#> Leader Gender[female]         2,893.188***    30.518***     181.131***   
#>                                 (252.894)      (4.295)       (25.250)    
#> Health System Score x Gender   -36.549***     -0.390***     -2.310***    
#>                                  (3.130)       (0.053)       (0.314)     
#> N                                 4,337         4,864         3,571      
#> R2                                0.057         0.090         0.151      
#> Adjusted R2                       0.054         0.088         0.148      
#> -------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001

6.2.3 Effectiveness (Health System Score) Interaction plots

effectiveness_health_system_score.png

x_label <- x_label <- wl_replace_by_name(cov_if, "lpi_health_score")

p_ic <- wl_plot_interaction(m_ic[[1]], interaction, "Color") +
    theme(legend.position = "none") + labs(title = "") + 
  ylab(names(wl_dependents()[1])) + xlab(x_label)

p_id <- wl_plot_interaction(m_id[[1]], interaction, "Color") +
    theme(legend.position = "none") + labs(title = "") + 
  ylab(names(wl_dependents()[2])) + xlab(x_label)

p_ie <- wl_plot_interaction(m_ie[[1]], interaction, "Color") +
    theme(legend.position = c(0.7, 0.8)) + labs(title = "") + 
  ylab(names(wl_dependents()[3])) + xlab(x_label)

p <- ggarrange(p_ic, p_id, p_ie,
               labels = c("New Infection Cases", "New Death Cases", "Excess Mortality"), 
               font.label = list(size = 10, face = "bold"),
               ncol = 3, nrow = 1) 

wl_ggsave("figures/effectiveness_health_system_score.png", plot = p)

6.2.4 Effectiveness (State Capacity) Regression models

cov_if <- wl_covariates(cov_group = "Capacity covariates")
cov_if <- c(cov_if[2:length(cov_if)], cov_if[1]) # Re-arrange covariates so we have pm_gender last interaction term for plot_model
interaction <- c("State Capacity x Gender" = "pm_gender*capacity") 

m_ic <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[1], x_var = cov_if, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"), interaction=interaction, arrange=FALSE)

m_id <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[2], x_var = cov_if, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"), interaction=interaction, arrange=FALSE)

m_ie <- wl_wave_regression(
  db_type = "Weekly", y_name = wl_dependents()[3], x_var = cov_if, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"), interaction=interaction, arrange=FALSE)

6.2.5 Effectiveness (State Capacity) Regression Table

effectiveness_capacity_reg.tex

#> 
#> -----------------------------------------------------------------------
#>                            Infection Cases Death Cases Excess Mortality
#>                                  (1)           (2)           (3)       
#> -----------------------------------------------------------------------
#> Leader Age                    -0.968**      -0.031***       -0.013     
#>                                (0.322)       (0.006)       (0.037)     
#> Stringency Index (delayed)    0.739***      0.015***      -0.169***    
#>                                (0.166)       (0.003)       (0.018)     
#> New Vaccination (delayed)      0.002*      -0.0001***      -0.0002     
#>                                (0.001)      (0.00002)      (0.0001)    
#> Democracy Score                -10.146       0.280*         1.640      
#>                                (8.049)       (0.142)       (0.930)     
#> Population Size                 0.072       0.004***       0.019**     
#>                                (0.062)       (0.001)       (0.007)     
#> Urbanization Percentage         0.388       -0.012**      -0.161***    
#>                                (0.258)       (0.005)       (0.035)     
#> Immigration Percentage         1.041*       -0.042***     -0.306***    
#>                                (0.529)       (0.009)       (0.056)     
#> State capacity                 -0.146        0.395**      -5.005***    
#>                                (8.438)       (0.148)       (1.151)     
#> Leader Gender[female]        181.419***     2.871***      19.472***    
#>                               (42.072)       (0.733)       (4.969)     
#> State Capacity x Gender      -109.554***    -1.861***     -9.891***    
#>                               (19.389)       (0.340)       (2.256)     
#> N                               4,018         4,505         3,279      
#> R2                              0.026         0.048         0.133      
#> Adjusted R2                     0.024         0.046         0.131      
#> -----------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001

6.2.6 Effectiveness (State Capacity) Interaction plots

effectiveness_state_capacity.png

x_label <- wl_replace_by_name(cov_if, "capacity")

p_ic <- wl_plot_interaction(m_ic[[1]], interaction, "Color") +
    theme(legend.position = "none") + labs(title = "") + 
  ylab(names(wl_dependents()[1])) + xlab(x_label)

p_id <- wl_plot_interaction(m_id[[1]], interaction, "Color") +
    theme(legend.position = "none") + labs(title = "") + 
  ylab(names(wl_dependents()[2])) + xlab(x_label)

p_ie <- wl_plot_interaction(m_ie[[1]], interaction, "Color") +
    theme(legend.position = c(0.7, 0.8)) + labs(title = "") + 
  ylab(names(wl_dependents()[3])) + xlab(x_label)

p <- ggarrange(p_ic, p_id, p_ie,
               labels = c("New Infection Cases", "New Death Cases", "Excess Mortality"), 
               font.label = list(size = 10, face = "bold"),
               ncol = 3, nrow = 1) 

wl_ggsave("figures/effectiveness_state_capacity.png", plot = p)

7 SI Robustness

7.1 Wave Alignment Results

7.1.1 Aligned - New Infection cases Regression models

dep <- wl_dependents()[1]
cov_34 <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_1AC", "Wave_2AC"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_3AC", "Wave_4AC"))

7.1.2 Aligned - Infection Cases Regression Table

aligned_cases_reg.tex

#> 
#> -------------------------------------------------------------------------
#>                               Wave 1     Wave 2     Wave 3      Wave 4   
#>                                (1)        (2)         (3)         (4)    
#> -------------------------------------------------------------------------
#> Leader Gender[female]       -17.383*** -94.776*** -138.459*** -116.154***
#>                              (5.007)    (21.622)   (20.628)    (29.780)  
#> Leader Age                   0.487**     0.933      -1.081     -3.578*** 
#>                              (0.183)    (0.762)     (0.683)     (1.051)  
#> Stringency Index (delayed)    0.049     1.315**    2.570***    -2.735*** 
#>                              (0.056)    (0.446)     (0.494)     (0.719)  
#> New Vaccination (delayed)                           -0.004     -0.023*** 
#>                                                     (0.003)     (0.003)  
#> Democracy Score             -20.265***  -35.962   -65.710***   -85.946** 
#>                              (4.618)    (19.581)   (17.856)    (27.058)  
#> GDP per Capita                0.313     1.963**     -0.003      2.662**  
#>                              (0.180)    (0.736)     (0.684)     (1.026)  
#> Health-system Score          -1.120**    -1.360     -3.209      -5.733*  
#>                              (0.429)    (1.914)     (1.775)     (2.632)  
#> Population Size              0.128***    -0.007     -0.363*     -0.265   
#>                              (0.033)    (0.152)     (0.141)     (0.213)  
#> Urbanization Percentage      0.802***  -2.303***   2.101***     -0.182   
#>                              (0.142)    (0.594)     (0.553)     (0.826)  
#> Immigration Percentage        -0.484     0.838     -3.644**    -5.165**  
#>                              (0.294)    (1.221)     (1.146)     (1.682)  
#> Population Aged 70 or Older -2.286***  16.916***   7.950***     8.290*   
#>                              (0.603)    (2.546)     (2.334)     (3.487)  
#> N                              630       1,113       1,113       1,113   
#> R2                            0.171      0.133       0.133       0.127   
#> Adjusted R2                   0.158      0.125       0.124       0.118   
#> -------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 17.38 units assuming all other covariates stay constant. This represents a change of more than 0.4228 standard deviations (SDNew Infection Cases= 41.12).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age increases New Infection Cases by 0.4866 units, assuming all other covariates stay constant. This represents a change of more than 0.01184 standard deviations (SDNew Infection Cases= 41.12).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 20.26 units, assuming all other covariates stay constant. This represents a change of more than 0.4929 standard deviations (SDNew Infection Cases= 41.12).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Infection Cases by 1.12 units, assuming all other covariates stay constant. This represents a change of more than 0.02723 standard deviations (SDNew Infection Cases= 41.12).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Infection Cases by 0.1282 units, assuming all other covariates stay constant. This represents a change of more than 0.003118 standard deviations (SDNew Infection Cases= 41.12).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 0.8017 units, assuming all other covariates stay constant. This represents a change of more than 0.0195 standard deviations (SDNew Infection Cases= 41.12).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older decreases New Infection Cases by 2.286 units, assuming all other covariates stay constant. This represents a change of more than 0.0556 standard deviations (SDNew Infection Cases= 41.12).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 94.78 units assuming all other covariates stay constant. This represents a change of more than 0.3837 standard deviations (SDNew Infection Cases= 247).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 1.315 units, assuming all other covariates stay constant. This represents a change of more than 0.005325 standard deviations (SDNew Infection Cases= 247).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases New Infection Cases by 1.963 units, assuming all other covariates stay constant. This represents a change of more than 0.007948 standard deviations (SDNew Infection Cases= 247).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Infection Cases by 2.303 units, assuming all other covariates stay constant. This represents a change of more than 0.009323 standard deviations (SDNew Infection Cases= 247).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Infection Cases by 16.92 units, assuming all other covariates stay constant. This represents a change of more than 0.06849 standard deviations (SDNew Infection Cases= 247).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 138.5 units assuming all other covariates stay constant. This represents a change of more than 0.6129 standard deviations (SDNew Infection Cases= 225.9).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 2.57 units, assuming all other covariates stay constant. This represents a change of more than 0.01138 standard deviations (SDNew Infection Cases= 225.9).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 65.71 units, assuming all other covariates stay constant. This represents a change of more than 0.2909 standard deviations (SDNew Infection Cases= 225.9).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size decreases New Infection Cases by 0.3629 units, assuming all other covariates stay constant. This represents a change of more than 0.001607 standard deviations (SDNew Infection Cases= 225.9).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 2.101 units, assuming all other covariates stay constant. This represents a change of more than 0.009303 standard deviations (SDNew Infection Cases= 225.9).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Infection Cases by 3.644 units, assuming all other covariates stay constant. This represents a change of more than 0.01613 standard deviations (SDNew Infection Cases= 225.9).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Infection Cases by 7.95 units, assuming all other covariates stay constant. This represents a change of more than 0.03519 standard deviations (SDNew Infection Cases= 225.9).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 116.2 units assuming all other covariates stay constant. This represents a change of more than 0.3444 standard deviations (SDNew Infection Cases= 337.3).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 3.578 units, assuming all other covariates stay constant. This represents a change of more than 0.01061 standard deviations (SDNew Infection Cases= 337.3).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Infection Cases by 2.735 units, assuming all other covariates stay constant. This represents a change of more than 0.008109 standard deviations (SDNew Infection Cases= 337.3).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Infection Cases by 0.02265 units, assuming all other covariates stay constant. This represents a change of more than 6.715e-05 standard deviations (SDNew Infection Cases= 337.3).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 85.95 units, assuming all other covariates stay constant. This represents a change of more than 0.2548 standard deviations (SDNew Infection Cases= 337.3).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases New Infection Cases by 2.662 units, assuming all other covariates stay constant. This represents a change of more than 0.007891 standard deviations (SDNew Infection Cases= 337.3).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Infection Cases by 5.733 units, assuming all other covariates stay constant. This represents a change of more than 0.017 standard deviations (SDNew Infection Cases= 337.3).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Infection Cases by 5.165 units, assuming all other covariates stay constant. This represents a change of more than 0.01531 standard deviations (SDNew Infection Cases= 337.3).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Infection Cases by 8.29 units, assuming all other covariates stay constant. This represents a change of more than 0.02458 standard deviations (SDNew Infection Cases= 337.3).

7.1.3 Aligned - Infection Cases (Waves 1-4) Regression Coefficients

aligned_cases_waves_reg.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/aligned_cases_waves_reg.png", plot = p, height=130)

7.1.4 Aligned - New Death Cases Regression models

dep <- wl_dependents()[2]
cov_34 <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_1AD", "Wave_2AD"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_3AD", "Wave_4AD"))

7.1.5 Aligned - Deaths Cases Regression Table

aligned_deaths_reg.tex

#> 
#> -------------------------------------------------------------------
#>                              Wave 1    Wave 2     Wave 3    Wave 4 
#>                                (1)       (2)       (3)       (4)   
#> -------------------------------------------------------------------
#> Leader Gender[female]       -1.720*** -1.252*** -2.391***  -1.494**
#>                              (0.293)   (0.378)   (0.365)   (0.526) 
#> Leader Age                   -0.006     0.004     -0.022    -0.021 
#>                              (0.011)   (0.013)   (0.012)   (0.018) 
#> Stringency Index (delayed)  -0.012*** 0.037***   0.047***   0.025  
#>                              (0.003)   (0.008)   (0.009)   (0.013) 
#> New Vaccination (delayed)                       -0.0002*** 0.00003 
#>                                                 (0.00004)  (0.0001)
#> Democracy Score             -0.829**   -0.132     0.139    3.038***
#>                              (0.261)   (0.342)   (0.321)   (0.501) 
#> GDP per Capita               0.023*    -0.008     0.012     0.001  
#>                              (0.010)   (0.013)   (0.012)   (0.018) 
#> Health-system Score          -0.030   -0.103**  -0.184***   -0.073 
#>                              (0.024)   (0.034)   (0.032)   (0.040) 
#> Population Size             0.008***   0.008**   -0.0002   -0.008* 
#>                              (0.002)   (0.003)   (0.003)   (0.004) 
#> Urbanization Percentage      0.023**  -0.056***  0.046***   -0.023 
#>                              (0.009)   (0.010)   (0.010)   (0.012) 
#> Immigration Percentage       -0.016     0.028   -0.112***  -0.099**
#>                              (0.017)   (0.021)   (0.020)   (0.032) 
#> Population Aged 70 or Older  -0.025   0.504***   0.306***  0.559***
#>                              (0.034)   (0.044)   (0.042)   (0.058) 
#> N                              714      1,113     1,113      628   
#> R2                            0.143     0.184     0.200     0.267  
#> Adjusted R2                   0.131     0.176     0.192     0.254  
#> -------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.72 units assuming all other covariates stay constant. This represents a change of more than 0.6919 standard deviations (SDNew Death Cases= 2.486).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Death Cases by 0.01157 units, assuming all other covariates stay constant. This represents a change of more than 0.004653 standard deviations (SDNew Death Cases= 2.486).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Death Cases by 0.8292 units, assuming all other covariates stay constant. This represents a change of more than 0.3336 standard deviations (SDNew Death Cases= 2.486).
The GDP per Capita covariate has statistically significant effect. An increase of 1 unit in GDP per Capita increases New Death Cases by 0.0234 units, assuming all other covariates stay constant. This represents a change of more than 0.009413 standard deviations (SDNew Death Cases= 2.486).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.008303 units, assuming all other covariates stay constant. This represents a change of more than 0.00334 standard deviations (SDNew Death Cases= 2.486).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Death Cases by 0.02347 units, assuming all other covariates stay constant. This represents a change of more than 0.009443 standard deviations (SDNew Death Cases= 2.486).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.252 units assuming all other covariates stay constant. This represents a change of more than 0.2819 standard deviations (SDNew Death Cases= 4.442).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.03733 units, assuming all other covariates stay constant. This represents a change of more than 0.008405 standard deviations (SDNew Death Cases= 4.442).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Death Cases by 0.1034 units, assuming all other covariates stay constant. This represents a change of more than 0.02328 standard deviations (SDNew Death Cases= 4.442).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.007619 units, assuming all other covariates stay constant. This represents a change of more than 0.001715 standard deviations (SDNew Death Cases= 4.442).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.05619 units, assuming all other covariates stay constant. This represents a change of more than 0.01265 standard deviations (SDNew Death Cases= 4.442).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.5036 units, assuming all other covariates stay constant. This represents a change of more than 0.1134 standard deviations (SDNew Death Cases= 4.442).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 2.391 units assuming all other covariates stay constant. This represents a change of more than 0.5668 standard deviations (SDNew Death Cases= 4.218).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.04746 units, assuming all other covariates stay constant. This represents a change of more than 0.01125 standard deviations (SDNew Death Cases= 4.218).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 0.0001633 units, assuming all other covariates stay constant. This represents a change of more than 3.871e-05 standard deviations (SDNew Death Cases= 4.218).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases New Death Cases by 0.1841 units, assuming all other covariates stay constant. This represents a change of more than 0.04365 standard deviations (SDNew Death Cases= 4.218).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Death Cases by 0.04599 units, assuming all other covariates stay constant. This represents a change of more than 0.0109 standard deviations (SDNew Death Cases= 4.218).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.1117 units, assuming all other covariates stay constant. This represents a change of more than 0.02648 standard deviations (SDNew Death Cases= 4.218).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.3063 units, assuming all other covariates stay constant. This represents a change of more than 0.0726 standard deviations (SDNew Death Cases= 4.218).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.494 units assuming all other covariates stay constant. This represents a change of more than 0.381 standard deviations (SDNew Death Cases= 3.921).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases New Death Cases by 3.038 units, assuming all other covariates stay constant. This represents a change of more than 0.7749 standard deviations (SDNew Death Cases= 3.921).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size decreases New Death Cases by 0.007848 units, assuming all other covariates stay constant. This represents a change of more than 0.002002 standard deviations (SDNew Death Cases= 3.921).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.09864 units, assuming all other covariates stay constant. This represents a change of more than 0.02516 standard deviations (SDNew Death Cases= 3.921).
The Population Aged 70 or Older covariate has statistically significant effect. An increase of 1 unit in Population Aged 70 or Older increases New Death Cases by 0.5594 units, assuming all other covariates stay constant. This represents a change of more than 0.1427 standard deviations (SDNew Death Cases= 3.921).

7.1.6 Aligned - Deaths Cases (Waves 1-4) Regression Coefficients

aligned_deaths_waves_reg.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/aligned_deaths_waves_reg.png", plot = p, height=130)

7.1.7 Aligned - Excess Mortality Regression models

dep <- wl_dependents()[3]
cov_34 <- wl_covariates(cov_group = "Main coviariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_1AD", "Wave_2AD"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_3AD", "Wave_4AD"))

7.1.8 Aligned - Excess Mortality Regression Table

aligned_excess_reg.tex

#> 
#> --------------------------------------------------------------------
#>                               Wave 1    Wave 2    Wave 3    Wave 4  
#>                                (1)        (2)       (3)       (4)   
#> --------------------------------------------------------------------
#> Leader Gender[female]       -12.677*** -7.894*** -8.598***  -3.367  
#>                              (2.451)    (2.316)   (1.781)   (3.342) 
#> Leader Age                    0.091      0.153     0.112    -0.216  
#>                              (0.092)    (0.090)   (0.063)   (0.130) 
#> Stringency Index (delayed)  -0.109***  -0.298*** -0.138**   -0.102  
#>                              (0.028)    (0.055)   (0.051)   (0.105) 
#> New Vaccination (delayed)                         0.0001   -0.002** 
#>                                                  (0.0002)   (0.001) 
#> Democracy Score             -9.900***    0.690    -4.049*  12.020***
#>                              (2.331)    (2.352)   (1.727)   (3.385) 
#> GDP per Capita                0.077      0.180    -0.027     0.087  
#>                              (0.101)    (0.093)   (0.068)   (0.134) 
#> Health-system Score          -0.491*   -1.340*** -1.332*** -0.772** 
#>                              (0.226)    (0.244)   (0.182)   (0.291) 
#> Population Size               0.042*     0.002    -0.009    -0.044  
#>                              (0.017)    (0.018)   (0.013)   (0.027) 
#> Urbanization Percentage      0.280**   -0.408***   0.082   -0.360***
#>                              (0.086)    (0.081)   (0.060)   (0.103) 
#> Immigration Percentage        -0.251    -0.338*  -0.685*** -0.609** 
#>                              (0.145)    (0.136)   (0.103)   (0.210) 
#> Population Aged 70 or Older   0.487      0.463    -0.320     0.449  
#>                              (0.334)    (0.324)   (0.242)   (0.423) 
#> N                              532        816       804       396   
#> R2                            0.143      0.219     0.307     0.344  
#> Adjusted R2                   0.127      0.209     0.297     0.325  
#> --------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 12.68 units assuming all other covariates stay constant. This represents a change of more than 0.6874 standard deviations (SDExcess Mortality= 18.44).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1087 units, assuming all other covariates stay constant. This represents a change of more than 0.005895 standard deviations (SDExcess Mortality= 18.44).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Excess Mortality by 9.9 units, assuming all other covariates stay constant. This represents a change of more than 0.5368 standard deviations (SDExcess Mortality= 18.44).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 0.4913 units, assuming all other covariates stay constant. This represents a change of more than 0.02664 standard deviations (SDExcess Mortality= 18.44).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.04169 units, assuming all other covariates stay constant. This represents a change of more than 0.00226 standard deviations (SDExcess Mortality= 18.44).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Excess Mortality by 0.2799 units, assuming all other covariates stay constant. This represents a change of more than 0.01518 standard deviations (SDExcess Mortality= 18.44).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 7.894 units assuming all other covariates stay constant. This represents a change of more than 0.3044 standard deviations (SDExcess Mortality= 25.93).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.2983 units, assuming all other covariates stay constant. This represents a change of more than 0.0115 standard deviations (SDExcess Mortality= 25.93).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 1.34 units, assuming all other covariates stay constant. This represents a change of more than 0.05168 standard deviations (SDExcess Mortality= 25.93).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.4084 units, assuming all other covariates stay constant. This represents a change of more than 0.01575 standard deviations (SDExcess Mortality= 25.93).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.3383 units, assuming all other covariates stay constant. This represents a change of more than 0.01305 standard deviations (SDExcess Mortality= 25.93).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 8.598 units assuming all other covariates stay constant. This represents a change of more than 0.4352 standard deviations (SDExcess Mortality= 19.76).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1379 units, assuming all other covariates stay constant. This represents a change of more than 0.006977 standard deviations (SDExcess Mortality= 19.76).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Excess Mortality by 4.049 units, assuming all other covariates stay constant. This represents a change of more than 0.2049 standard deviations (SDExcess Mortality= 19.76).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 1.332 units, assuming all other covariates stay constant. This represents a change of more than 0.06742 standard deviations (SDExcess Mortality= 19.76).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.6845 units, assuming all other covariates stay constant. This represents a change of more than 0.03464 standard deviations (SDExcess Mortality= 19.76).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Excess Mortality by 0.001555 units, assuming all other covariates stay constant. This represents a change of more than 6.634e-05 standard deviations (SDExcess Mortality= 23.44).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Excess Mortality by 12.02 units, assuming all other covariates stay constant. This represents a change of more than 0.5129 standard deviations (SDExcess Mortality= 23.44).
The Health-system Score covariate has statistically significant effect. An increase of 1 unit in Health-system Score decreases Excess Mortality by 0.7721 units, assuming all other covariates stay constant. This represents a change of more than 0.03295 standard deviations (SDExcess Mortality= 23.44).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.3597 units, assuming all other covariates stay constant. This represents a change of more than 0.01535 standard deviations (SDExcess Mortality= 23.44).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.6086 units, assuming all other covariates stay constant. This represents a change of more than 0.02597 standard deviations (SDExcess Mortality= 23.44).

7.1.9 Aligned - Excess Mortality (Waves 1-4) Regression Coefficients

aligned_excess_waves_reg.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/aligned_excess_waves_reg.png", plot = p, height=130)

7.2 Wave Alignment & State Capacity Results

7.2.1 Capacity Aligned - New Infection cases Regression models

dep <- wl_dependents()[1]
cov_34 <- wl_covariates(cov_group = "Capacity covariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_1AC", "Wave_2AC"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "aligned", waves = c("Wave_3AC", "Wave_4AC"))

7.2.2 Capacity Aligned - Infection Cases Regression Table

capacity_aligned_cases_reg.tex

#> 
#> -----------------------------------------------------------------------
#>                             Wave 1     Wave 2     Wave 3      Wave 4   
#>                               (1)       (2)         (3)         (4)    
#> -----------------------------------------------------------------------
#> Leader Gender[female]      -14.849** -97.462*** -109.530*** -124.752***
#>                             (5.744)   (24.187)   (22.933)    (32.636)  
#> Leader Age                  0.598**    1.130      -1.646*    -4.295*** 
#>                             (0.201)   (0.802)     (0.720)     (1.099)  
#> Stringency Index (delayed)   0.096    1.629***   3.037***    -3.699*** 
#>                             (0.060)   (0.478)     (0.506)     (0.783)  
#> New Vaccination (delayed)                         -0.003     -0.024*** 
#>                                                   (0.003)     (0.004)  
#> Democracy Score            -10.203*  -71.794*** -91.054***  -94.504*** 
#>                             (4.754)   (19.489)   (17.987)    (27.630)  
#> Population Size            0.126***    0.183      -0.201      -0.086   
#>                             (0.033)   (0.153)     (0.142)     (0.218)  
#> Urbanization Percentage    0.655***  -3.301***    1.794**      0.541   
#>                             (0.149)   (0.637)     (0.587)     (0.891)  
#> Immigration Percentage       0.092     -1.984    -3.568**     -0.003   
#>                             (0.315)   (1.334)     (1.215)     (1.816)  
#> State capacity              -6.556   107.662***   -22.833     -4.853   
#>                             (5.238)   (19.890)   (18.219)    (29.635)  
#> N                             594      1,029       1,029       1,029   
#> R2                           0.126     0.094       0.120       0.119   
#> Adjusted R2                  0.114     0.086       0.112       0.111   
#> -----------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 14.85 units assuming all other covariates stay constant. This represents a change of more than 0.3612 standard deviations (SDNew Infection Cases= 41.12).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age increases New Infection Cases by 0.5978 units, assuming all other covariates stay constant. This represents a change of more than 0.01454 standard deviations (SDNew Infection Cases= 41.12).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 10.2 units, assuming all other covariates stay constant. This represents a change of more than 0.2482 standard deviations (SDNew Infection Cases= 41.12).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Infection Cases by 0.1258 units, assuming all other covariates stay constant. This represents a change of more than 0.003059 standard deviations (SDNew Infection Cases= 41.12).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 0.6545 units, assuming all other covariates stay constant. This represents a change of more than 0.01592 standard deviations (SDNew Infection Cases= 41.12).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 97.46 units assuming all other covariates stay constant. This represents a change of more than 0.3946 standard deviations (SDNew Infection Cases= 247).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 1.629 units, assuming all other covariates stay constant. This represents a change of more than 0.006596 standard deviations (SDNew Infection Cases= 247).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 71.79 units, assuming all other covariates stay constant. This represents a change of more than 0.2907 standard deviations (SDNew Infection Cases= 247).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Infection Cases by 3.301 units, assuming all other covariates stay constant. This represents a change of more than 0.01337 standard deviations (SDNew Infection Cases= 247).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity increases New Infection Cases by 107.7 units, assuming all other covariates stay constant. This represents a change of more than 0.4359 standard deviations (SDNew Infection Cases= 247).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 109.5 units assuming all other covariates stay constant. This represents a change of more than 0.4849 standard deviations (SDNew Infection Cases= 225.9).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 1.646 units, assuming all other covariates stay constant. This represents a change of more than 0.007285 standard deviations (SDNew Infection Cases= 225.9).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Infection Cases by 3.037 units, assuming all other covariates stay constant. This represents a change of more than 0.01344 standard deviations (SDNew Infection Cases= 225.9).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 91.05 units, assuming all other covariates stay constant. This represents a change of more than 0.4031 standard deviations (SDNew Infection Cases= 225.9).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Infection Cases by 1.794 units, assuming all other covariates stay constant. This represents a change of more than 0.007941 standard deviations (SDNew Infection Cases= 225.9).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Infection Cases by 3.568 units, assuming all other covariates stay constant. This represents a change of more than 0.0158 standard deviations (SDNew Infection Cases= 225.9).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Infection Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Infection Cases decreases by 124.8 units assuming all other covariates stay constant. This represents a change of more than 0.3699 standard deviations (SDNew Infection Cases= 337.3).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Infection Cases by 4.295 units, assuming all other covariates stay constant. This represents a change of more than 0.01273 standard deviations (SDNew Infection Cases= 337.3).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Infection Cases by 3.699 units, assuming all other covariates stay constant. This represents a change of more than 0.01097 standard deviations (SDNew Infection Cases= 337.3).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Infection Cases by 0.02371 units, assuming all other covariates stay constant. This represents a change of more than 7.03e-05 standard deviations (SDNew Infection Cases= 337.3).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Infection Cases by 94.5 units, assuming all other covariates stay constant. This represents a change of more than 0.2802 standard deviations (SDNew Infection Cases= 337.3).

7.2.3 Capacity Aligned - Infection Cases (Waves 1-4) Regression Coefficients

capacity_aligned_cases_waves_reg.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/capacity_aligned_cases_waves_reg.png", plot = p, height=130)

7.2.4 Capacity Aligned - New Death Cases Regression models

dep <- wl_dependents()[2]
cov_34 <- wl_covariates(cov_group = "Capacity covariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_1AD", "Wave_2AD"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_3AD", "Wave_4AD"))

7.2.5 Capacity Aligned - Deaths Cases Regression Table

capacity_aligned_deaths_reg.tex

#> 
#> -------------------------------------------------------------------
#>                             Wave 1    Wave 2     Wave 3    Wave 4  
#>                               (1)       (2)       (3)        (4)   
#> -------------------------------------------------------------------
#> Leader Gender[female]      -1.637*** -1.640*** -2.008***   -1.128  
#>                             (0.324)   (0.441)   (0.415)    (0.576) 
#> Leader Age                  -0.016    -0.010    -0.039**  -0.056** 
#>                             (0.011)   (0.015)   (0.013)    (0.020) 
#> Stringency Index (delayed) -0.011**  0.037***   0.061***    0.001  
#>                             (0.004)   (0.008)   (0.009)    (0.015) 
#> New Vaccination (delayed)                      -0.0002*** -0.00005 
#>                                                 (0.0001)  (0.0001) 
#> Democracy Score            -0.769**   -0.687     -0.314   2.671*** 
#>                             (0.266)   (0.354)   (0.331)    (0.542) 
#> Population Size            0.010***  0.011***    0.006*    -0.006  
#>                             (0.002)   (0.003)   (0.003)    (0.004) 
#> Urbanization Percentage     0.027**  -0.074***   0.026*    -0.010  
#>                             (0.009)   (0.012)   (0.011)    (0.015) 
#> Immigration Percentage      0.043*   -0.074**  -0.125***  -0.137***
#>                             (0.017)   (0.024)   (0.022)    (0.040) 
#> State capacity              -0.407   1.951***    -0.229    1.361*  
#>                             (0.274)   (0.360)   (0.334)    (0.548) 
#> N                             675      1,029     1,029       582   
#> R2                           0.141     0.101     0.146      0.131  
#> Adjusted R2                  0.131     0.094     0.139      0.118  
#> -------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.637 units assuming all other covariates stay constant. This represents a change of more than 0.6587 standard deviations (SDNew Death Cases= 2.486).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases New Death Cases by 0.01093 units, assuming all other covariates stay constant. This represents a change of more than 0.004399 standard deviations (SDNew Death Cases= 2.486).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases New Death Cases by 0.7691 units, assuming all other covariates stay constant. This represents a change of more than 0.3094 standard deviations (SDNew Death Cases= 2.486).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.009635 units, assuming all other covariates stay constant. This represents a change of more than 0.003876 standard deviations (SDNew Death Cases= 2.486).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Death Cases by 0.02717 units, assuming all other covariates stay constant. This represents a change of more than 0.01093 standard deviations (SDNew Death Cases= 2.486).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage increases New Death Cases by 0.04317 units, assuming all other covariates stay constant. This represents a change of more than 0.01737 standard deviations (SDNew Death Cases= 2.486).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 1.64 units assuming all other covariates stay constant. This represents a change of more than 0.3692 standard deviations (SDNew Death Cases= 4.442).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.03739 units, assuming all other covariates stay constant. This represents a change of more than 0.008418 standard deviations (SDNew Death Cases= 4.442).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.01084 units, assuming all other covariates stay constant. This represents a change of more than 0.00244 standard deviations (SDNew Death Cases= 4.442).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases New Death Cases by 0.07368 units, assuming all other covariates stay constant. This represents a change of more than 0.01659 standard deviations (SDNew Death Cases= 4.442).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.07421 units, assuming all other covariates stay constant. This represents a change of more than 0.01671 standard deviations (SDNew Death Cases= 4.442).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity increases New Death Cases by 1.951 units, assuming all other covariates stay constant. This represents a change of more than 0.4391 standard deviations (SDNew Death Cases= 4.442).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, New Death Cases decreases by 2.008 units assuming all other covariates stay constant. This represents a change of more than 0.476 standard deviations (SDNew Death Cases= 4.218).
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.03863 units, assuming all other covariates stay constant. This represents a change of more than 0.009156 standard deviations (SDNew Death Cases= 4.218).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) increases New Death Cases by 0.06126 units, assuming all other covariates stay constant. This represents a change of more than 0.01452 standard deviations (SDNew Death Cases= 4.218).
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases New Death Cases by 0.0001786 units, assuming all other covariates stay constant. This represents a change of more than 4.233e-05 standard deviations (SDNew Death Cases= 4.218).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases New Death Cases by 0.005755 units, assuming all other covariates stay constant. This represents a change of more than 0.001364 standard deviations (SDNew Death Cases= 4.218).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases New Death Cases by 0.02568 units, assuming all other covariates stay constant. This represents a change of more than 0.006088 standard deviations (SDNew Death Cases= 4.218).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.1254 units, assuming all other covariates stay constant. This represents a change of more than 0.02972 standard deviations (SDNew Death Cases= 4.218).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on New Death Cases.
The Leader Age covariate has statistically significant effect. An increase of 1 unit in Leader Age decreases New Death Cases by 0.05605 units, assuming all other covariates stay constant. This represents a change of more than 0.0143 standard deviations (SDNew Death Cases= 3.921).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases New Death Cases by 2.671 units, assuming all other covariates stay constant. This represents a change of more than 0.6812 standard deviations (SDNew Death Cases= 3.921).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases New Death Cases by 0.1369 units, assuming all other covariates stay constant. This represents a change of more than 0.03492 standard deviations (SDNew Death Cases= 3.921).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity increases New Death Cases by 1.361 units, assuming all other covariates stay constant. This represents a change of more than 0.3473 standard deviations (SDNew Death Cases= 3.921).

7.2.6 Capacity Aligned - Deaths Cases (Waves 1-4) Regression Coefficients

capacity_aligned_deaths.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/capacity_aligned_deaths.png", plot = p, height=130)

7.2.7 Capacity Aligned - Excess Mortality Regression models

dep <- wl_dependents()[3]
cov_34 <- wl_covariates(cov_group = "Capacity covariates")
cov_12 <- wl_setdiff(cov_34, c("new_vaccinations_per_mil_delayed"))

m_12 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_12, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_1AD", "Wave_2AD"))

m_34 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_34, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "aligned", waves = c("Wave_3AD", "Wave_4AD"))

7.2.8 Capacity Aligned - Excess Mortality Regression Table

capacity_aligned_excess_reg.tex

#> 
#> -------------------------------------------------------------------
#>                              Wave 1    Wave 2     Wave 3    Wave 4 
#>                               (1)        (2)       (3)       (4)   
#> -------------------------------------------------------------------
#> Leader Gender[female]      -12.149***  -3.710    -4.490*    -2.690 
#>                             (2.791)    (2.598)   (1.956)   (3.249) 
#> Leader Age                   0.018      0.176     0.038     -0.256 
#>                             (0.099)    (0.096)   (0.069)   (0.135) 
#> Stringency Index (delayed) -0.103***  -0.274***  -0.157**   -0.179 
#>                             (0.029)    (0.056)   (0.053)   (0.102) 
#> New Vaccination (delayed)                        -0.0002   -0.002**
#>                                                  (0.0002)  (0.001) 
#> Democracy Score            -10.200***  -0.777    -4.323*   10.397**
#>                             (2.395)    (2.371)   (1.778)   (3.563) 
#> Population Size             0.054***   0.042*     0.023     -0.017 
#>                             (0.016)    (0.017)   (0.012)   (0.026) 
#> Urbanization Percentage     0.265**   -0.605***   0.044    -0.313**
#>                             (0.090)    (0.088)   (0.066)   (0.111) 
#> Immigration Percentage       -0.093   -0.396**  -0.602***  -0.548* 
#>                             (0.146)    (0.142)   (0.109)   (0.243) 
#> State capacity               -2.778    -5.444*  -15.320***  -7.338 
#>                             (2.862)    (2.668)   (2.081)   (4.178) 
#> N                             499        748       736       367   
#> R2                           0.132      0.215     0.287     0.326  
#> Adjusted R2                  0.118      0.207     0.278     0.309  
#> -------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
Interpretation

Wave 1 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 12.15 units assuming all other covariates stay constant. This represents a change of more than 0.6587 standard deviations (SDExcess Mortality= 18.44).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1035 units, assuming all other covariates stay constant. This represents a change of more than 0.00561 standard deviations (SDExcess Mortality= 18.44).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Excess Mortality by 10.2 units, assuming all other covariates stay constant. This represents a change of more than 0.5531 standard deviations (SDExcess Mortality= 18.44).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.05353 units, assuming all other covariates stay constant. This represents a change of more than 0.002902 standard deviations (SDExcess Mortality= 18.44).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage increases Excess Mortality by 0.2645 units, assuming all other covariates stay constant. This represents a change of more than 0.01434 standard deviations (SDExcess Mortality= 18.44).

Wave 2 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.2745 units, assuming all other covariates stay constant. This represents a change of more than 0.01058 standard deviations (SDExcess Mortality= 25.93).
The Population Size covariate has statistically significant effect. An increase of 1 unit in Population Size increases Excess Mortality by 0.04225 units, assuming all other covariates stay constant. This represents a change of more than 0.001629 standard deviations (SDExcess Mortality= 25.93).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.6053 units, assuming all other covariates stay constant. This represents a change of more than 0.02334 standard deviations (SDExcess Mortality= 25.93).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.3961 units, assuming all other covariates stay constant. This represents a change of more than 0.01527 standard deviations (SDExcess Mortality= 25.93).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases Excess Mortality by 5.444 units, assuming all other covariates stay constant. This represents a change of more than 0.2099 standard deviations (SDExcess Mortality= 25.93).

Wave 3 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The Leader Gender[female] covariate has statistically significant effect. When Leader Gender[female] is Female, Excess Mortality decreases by 4.49 units assuming all other covariates stay constant. This represents a change of more than 0.2272 standard deviations (SDExcess Mortality= 19.76).
The Stringency Index (delayed) covariate has statistically significant effect. An increase of 1 unit in Stringency Index (delayed) decreases Excess Mortality by 0.1568 units, assuming all other covariates stay constant. This represents a change of more than 0.007935 standard deviations (SDExcess Mortality= 19.76).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score decreases Excess Mortality by 4.323 units, assuming all other covariates stay constant. This represents a change of more than 0.2188 standard deviations (SDExcess Mortality= 19.76).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.6016 units, assuming all other covariates stay constant. This represents a change of more than 0.03045 standard deviations (SDExcess Mortality= 19.76).
The State capacity covariate has statistically significant effect. An increase of 1 unit in State capacity decreases Excess Mortality by 15.32 units, assuming all other covariates stay constant. This represents a change of more than 0.7753 standard deviations (SDExcess Mortality= 19.76).

Wave 4 interpretation

OLS Linear regression was used to analyze the effects on Excess Mortality.
The New Vaccination (delayed) covariate has statistically significant effect. An increase of 1 unit in New Vaccination (delayed) decreases Excess Mortality by 0.001883 units, assuming all other covariates stay constant. This represents a change of more than 8.035e-05 standard deviations (SDExcess Mortality= 23.44).
The Democracy Score covariate has statistically significant effect. An increase of 1 unit in Democracy Score increases Excess Mortality by 10.4 units, assuming all other covariates stay constant. This represents a change of more than 0.4437 standard deviations (SDExcess Mortality= 23.44).
The Urbanization Percentage covariate has statistically significant effect. An increase of 1 unit in Urbanization Percentage decreases Excess Mortality by 0.3132 units, assuming all other covariates stay constant. This represents a change of more than 0.01336 standard deviations (SDExcess Mortality= 23.44).
The Immigration Percentage covariate has statistically significant effect. An increase of 1 unit in Immigration Percentage decreases Excess Mortality by 0.5481 units, assuming all other covariates stay constant. This represents a change of more than 0.02339 standard deviations (SDExcess Mortality= 23.44).

7.2.9 Capacity Aligned - Excess Mortality (Waves 1-4) Regression Coefficients

capacity_aligned_excess.png

p1 <- wl_plot_coef (model = m_12[[1]], covs = cov_12, palette = "Color")
p2 <- wl_plot_coef (model = m_12[[2]], covs = cov_12, palette = "Color")
p3 <- wl_plot_coef (model = m_34[[1]], covs = cov_34, palette = "Color")
p4 <- wl_plot_coef (model = m_34[[2]], covs = cov_34, palette = "Color")

p <- ggarrange(p1, p2, p3, p4, 
               labels = c("Wave 1", "Wave 2", "Wave 3", "Wave 4"), 
               align = "hv", font.label = list(size = 10, face = "bold"),
               vjust = 2.5, hjust = -3.5, ncol = 2, nrow = 2) 

wl_ggsave("figures/capacity_aligned_excess.png", plot = p, height=130)

7.3 Delay Sensitivity

cov_fr1 <- wl_covariates(cov_group = "Main coviariates") %>%
  wl_setdiff(c("stringency_delayed", "new_vaccinations_per_mil_delayed")) %>%
  c("Stringency Index (delayed 4)" = "stringency_delayed_r1",
    "New Vaccination (delayed 4)" = "new_vaccinations_per_mil_delayed_r1")

cov_fr2 <- wl_covariates(cov_group = "Main coviariates") %>%
  wl_setdiff(c("stringency_delayed", "new_vaccinations_per_mil_delayed")) %>%
  c("Stringency Index (delayed 8)" = "stringency_delayed_r2",
    "New Vaccination (delayed 8)" = "new_vaccinations_per_mil_delayed_r2")

dep <- wl_dependents()[1]

m_fcr1 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fr1, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

m_fcr2 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fr2, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "cases", 
  align_type = "calendar", waves = c("All_C"))

dep <- wl_dependents()[2]

m_fdr1 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fr1, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

m_fdr2 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fr2, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

dep <- wl_dependents()[3]

m_fer1 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fr1, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

m_fer2 <- wl_wave_regression(
  db_type = "Weekly", y_name = dep, x_var = cov_fr2, 
  country_list = wl_countries(), model_type = "OLS",   data_type = "deaths", 
  align_type = "calendar", waves = c("All_D"))

7.3.1 Delay of 4 weeks Table

delay_4_reg.tex

#> 
#> ---------------------------------------------------------------------------------
#>                              New Infection Cases New Death Cases Excess Mortality
#>                                      (1)               (2)             (3)       
#> ---------------------------------------------------------------------------------
#> Leader Gender[female]            -54.289***         -0.794***       -3.743***    
#>                                    (8.696)           (0.149)         (0.922)     
#> Leader Age                         -0.412           -0.017***         0.014      
#>                                    (0.303)           (0.005)         (0.035)     
#> Democracy Score                     2.445           0.611***         3.474***    
#>                                    (7.938)           (0.136)         (0.934)     
#> GDP per Capita                      0.490            0.0001           0.042      
#>                                    (0.299)           (0.005)         (0.037)     
#> Health-system Score                -1.284           -0.066***       -0.758***    
#>                                    (0.776)           (0.013)         (0.097)     
#> Population Size                    -0.007             0.002           -0.002     
#>                                    (0.062)           (0.001)         (0.007)     
#> Urbanization Percentage             0.134           -0.011**        -0.124***    
#>                                    (0.243)           (0.004)         (0.033)     
#> Immigration Percentage              0.247           -0.023**        -0.291***    
#>                                    (0.496)           (0.008)         (0.054)     
#> Population Aged 70 or Older       5.015***          0.238***         0.519***    
#>                                    (1.029)           (0.018)         (0.130)     
#> Stringency Index (delayed 4)      0.853***          0.030***        -0.085***    
#>                                    (0.159)           (0.003)         (0.018)     
#> New Vaccination (delayed 4)       0.004***          -0.00003*        -0.00003    
#>                                    (0.001)          (0.00002)        (0.0001)    
#> N                                   4,419             4,935           3,629      
#> R2                                  0.033             0.091           0.113      
#> Adjusted R2                         0.030             0.089           0.111      
#> ---------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001

7.3.2 Delay of 8 weeks Table

delay_8_reg.tex

#> 
#> ---------------------------------------------------------------------------------
#>                              New Infection Cases New Death Cases Excess Mortality
#>                                      (1)               (2)             (3)       
#> ---------------------------------------------------------------------------------
#> Leader Gender[female]            -61.801***         -1.115***       -4.837***    
#>                                    (9.021)           (0.154)         (0.914)     
#> Leader Age                         -0.361           -0.015**          0.047      
#>                                    (0.315)           (0.005)         (0.035)     
#> Democracy Score                     2.709           0.682***         3.333***    
#>                                    (8.257)           (0.141)         (0.929)     
#> GDP per Capita                      0.481            -0.004           0.017      
#>                                    (0.311)           (0.005)         (0.037)     
#> Health-system Score                -1.212           -0.060***       -0.778***    
#>                                    (0.807)           (0.014)         (0.096)     
#> Population Size                    -0.015             0.002           -0.001     
#>                                    (0.064)           (0.001)         (0.007)     
#> Urbanization Percentage             0.165           -0.012**        -0.133***    
#>                                    (0.253)           (0.004)         (0.033)     
#> Immigration Percentage              0.222            -0.021*        -0.297***    
#>                                    (0.516)           (0.009)         (0.054)     
#> Population Aged 70 or Older       5.052***          0.232***         0.472***    
#>                                    (1.071)           (0.018)         (0.130)     
#> Stringency Index (delayed 8)       0.458**            0.003         -0.219***    
#>                                    (0.162)           (0.003)         (0.017)     
#> New Vaccination (delayed 8)         0.001          -0.0001***        -0.0001     
#>                                    (0.001)          (0.00002)        (0.0001)    
#> N                                   4,231             4,761           3,493      
#> R2                                  0.025             0.082           0.162      
#> Adjusted R2                         0.022             0.080           0.159      
#> ---------------------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001

7.4 Outlier Sensitivity

outlier_reg.tex

#> 
#> ------------------------------------------------------------------
#>                             Infections   Deaths   Excess Mortality
#>                                (1)        (2)           (3)       
#> ------------------------------------------------------------------
#> Leader Gender[female]       -64.438*** -0.959***     -4.247***    
#>                              (8.690)    (0.151)       (0.899)     
#> Leader Age                    -0.467   -0.018***       0.029      
#>                              (0.308)    (0.005)       (0.035)     
#> Stringency Index (delayed)   0.593***   0.013***     -0.186***    
#>                              (0.158)    (0.003)       (0.017)     
#> New Vaccination (delayed)     0.002    -0.0001***     -0.0002     
#>                              (0.001)   (0.00002)      (0.0001)    
#> Democracy Score               1.367     0.630***      3.144***    
#>                              (8.107)    (0.140)       (0.934)     
#> GDP per Capita                0.510      -0.001        0.021      
#>                              (0.306)    (0.005)       (0.037)     
#> Health-system Score           -1.288   -0.068***     -0.790***    
#>                              (0.789)    (0.014)       (0.097)     
#> Population Size               -0.022     0.002         -0.002     
#>                              (0.063)    (0.001)       (0.007)     
#> Urbanization Percentage       0.259      -0.008       -0.102**    
#>                              (0.242)    (0.004)       (0.031)     
#> Immigration Percentage        0.160     -0.023**     -0.291***    
#>                              (0.506)    (0.009)       (0.054)     
#> Population Aged 70 or Older  4.991***   0.235***      0.539***    
#>                              (1.051)    (0.018)       (0.130)     
#> N                             4,421      4,958         3,665      
#> R2                            0.029      0.075         0.135      
#> Adjusted R2                   0.026      0.073         0.133      
#> ------------------------------------------------------------------
#> *p < .05; **p < .01; ***p < .001
cov_list <- wl_covariates(cov_group = "Timing covariates")
p_list <- c("w1c_peak_week","w2c_peak_week","w1d_peak_week","w2d_peak_week")
country_list <- wl_countries()

ld_db <- wl_ld_prepare_cox_aligned(country_list, c("country", cov_list, p_list))

7.5 Lockdown Timing & Stringency

7.5.1 Lockdown Survival Curves

Wave 1 Surival Curve.png

sfit <- survfit(w1_surv ~ pm_gender, data=ld_db)
p <- wl_surv_plot_fit(sfit, "Wave 1")
wl_ggsave("figures/Wave 1 Surival Curve.png", plot = p)

Wave 2 Surival Curve.png

sfit <- survfit(w2_surv ~ pm_gender, data=ld_db)
p <- wl_surv_plot_fit(sfit, "Wave 2")
wl_ggsave("figures/Wave 2 Surival Curve.png", plot = p)

7.5.2 Lockdown Timing Regression

Lockdown timing regression.tex

#> 
#> -------------------------------------------
#>                             Wave 1  Wave 2 
#>                               (1)     (2)  
#> -------------------------------------------
#> Leader Gender[female]        0.282   0.310 
#>                             (0.710) (0.693)
#> Leader Age                   0.978   0.985 
#>                             (0.020) (0.022)
#> Democracy Score              1.401   1.324 
#>                             (0.525) (0.527)
#> GDP per Capita               1.024   1.001 
#>                             (0.027) (0.028)
#> Health-system Score          0.983   0.946 
#>                             (0.045) (0.051)
#> Population Size              0.999   1.004 
#>                             (0.003) (0.003)
#> Urbanization Percentage      0.988   0.997 
#>                             (0.017) (0.017)
#> Immigration Percentage       0.985  1.091* 
#>                             (0.034) (0.037)
#> Population Aged 70 or Older  0.924   0.966 
#>                             (0.066) (0.072)
#> State capacity               0.938   0.951 
#>                             (0.654) (0.808)
#> N                             50      50   
#> R2                           0.241   0.259 
#> -------------------------------------------
#> *p < .05; **p < .01; ***p < .001           
#> All estimates are exponentiated.

Wave 1 Lockdown Timing Regression.png

p_title <- paste("Wave 1", m_title)  
p <- wl_cox_print_model(scox1, ld_db, p_title)
wl_ggsave("figures/Wave 1 Lockdown Timing Regression.png", plot = p)

Wave 2 Lockdown Timing Regression.png

p_title <- paste("Wave 2", m_title)
p <- wl_cox_print_model(scox2, ld_db, p_title)
wl_ggsave("figures/Wave 2 Lockdown Timing Regression.png", plot = p)

7.5.3 Lockdown Stringency Regression

Lockdown Stringency regression table.tex

#> 
#> ---------------------------------------------
#>                              Wave 1   Wave 2 
#>                               (1)      (2)   
#> ---------------------------------------------
#> Leader Gender[female]       0.341*** 0.309***
#>                             (0.313)  (0.188) 
#> Leader Age                   0.981   0.973***
#>                             (0.011)  (0.007) 
#> Democracy Score             2.182**   1.180  
#>                             (0.284)  (0.154) 
#> GDP per Capita               0.987   1.050***
#>                             (0.014)  (0.008) 
#> Health-system Score          0.943*   0.976  
#>                             (0.028)  (0.015) 
#> Population Size              0.996   1.005***
#>                             (0.002)  (0.001) 
#> Urbanization Percentage      1.015   1.032***
#>                             (0.009)  (0.005) 
#> Immigration Percentage      1.068***  1.017  
#>                             (0.018)  (0.010) 
#> Population Aged 70 or Older  1.045   0.902***
#>                             (0.038)  (0.023) 
#> State capacity               0.406*  0.331***
#>                             (0.408)  (0.227) 
#>                             1.029***  1.000  
#>                             (0.006)  (0.0004)
#>                              1.047   1.202***
#>                             (0.057)  (0.024) 
#> N                             463     1,225  
#> ---------------------------------------------
#> *p < .05; **p < .01; ***p < .001             
#> All estimates are exponentiated.