REPRO-Bench / 29 /replication_package /Code /Figure 8, SI19.R
anonymous-submission-acl2025's picture
add 29
0369477
raw
history blame
3.98 kB
# File name: figure8_SIfigure19.R
# In:
# - /Data/france_data.RData: Data on 2020 municipal elections in France, linked with local-level cases and deaths
# Out:
# - /Figures/figure8.pdf
# - /Figures/SI_figure19.pdf
require(lfe)
require(ggridges)
require(tidyverse)
require(gridExtra)
rm(list =ls())
gc()
####################################################################################################################### Loading functions
source('./helper_functions.R')
load('../Data/france_data.RData')
# Figure 8
p1 <- toanal %>%
mutate(mainstream = ifelse(mainstream == 0,'Non-Mainstream','Mainstream'),
period = factor(ifelse(date == as.Date('2020-03-15'),'March Elections','June Elections'),
levels = c('March Elections','June Elections'))) %>%
group_by(mainstream,period) %>%
summarise(votes = mean(votes)) %>%
ungroup() %>%
filter(complete.cases(.)) %>%
ggplot(aes(x = period,y = votes,fill = factor(mainstream))) +
stat_summary(fun.y = mean,geom = "bar",position = position_dodge(width = .9),
size = 3,alpha = .7) +
scale_fill_manual(name = '',values = c('Mainstream' = 'grey80','Non-Mainstream' = 'grey30')) +
theme_ridges() + xlab('Period') + ylab('Aggregate Vote Share')
p2 <- toanal %>%
mutate(mainstream = ifelse(mainstream == 0,'Non-Mainstream','Mainstream'),
period = factor(ifelse(date == as.Date('2020-03-15'),'March Elections','June Elections'),
levels = c('March Elections','June Elections')),
logcases = log(deaths+1)) %>%
filter(!is.na(period)) %>%
ggplot(aes(x = logcases,y = votes,color = mainstream)) +
geom_point(alpha = .7,size = 2) +
geom_smooth(method = 'lm') +
scale_color_manual(name = "",values = c('Mainstream' = 'grey80',"Non-Mainstream" = "grey30")) +
theme_ridges() + xlab("Deaths (logged)") + ylab('Aggregate Vote Share')
pdf('../figures/figure8.pdf',width = 7,height = 4)
grid.arrange(p1 + theme(legend.position = 'bottom'),
p2 + theme(legend.position = 'bottom'),ncol = 2)
dev.off()
# SI Figure 19
toanal$post <- ifelse(toanal$date == '2020-03-15',0,1)
toanal$antiEst <- (toanal$mainstream - 1)^2
summary(mod1 <- felm(votes ~ antiEst*post | dpt_code | 0 | dpt_code,toanal))
summary(mod2 <- felm(votes ~ antiEst*log(deaths+1) | post + dpt_code | 0 | dpt_code,toanal))
toplot <- data.frame(interaction_plot_continuous(mod1,plot = T,pointsplot = T,colr = 'black',alph = 200,num_points = 2,
xlabel = "Election Wave",ylabel = 'Marginal Coefficient of Mainstream')) %>%
mutate(type = 'Temporal')
toplot <- bind_rows(toplot,data.frame(interaction_plot_continuous(mod2,plot = T,pointsplot = T,
colr = 'black',alph = 200,num_points = 20,
xlabel = "Election Wave",ylabel = 'Marginal Coefficient of Mainstream')) %>%
mutate(type = 'Geographic'))
p1 <- toplot %>%
filter(type == 'Temporal') %>%
mutate(x = factor(ifelse(x_2 == 0,'March Elections','June Elections'),levels = c('March Elections','June Elections'))) %>%
ggplot(aes(x = x,y = delta_1)) +
geom_point() +
geom_errorbar(aes(ymin = lb,ymax = ub),width = .1) +
geom_hline(yintercept = 0,linetype = 'dashed') +
ylab('Margnial Coefficient on Non-Mainstream') + xlab('') +
theme_ridges() +
ggtitle('Temporal Variation')
p2 <- toplot %>%
filter(type == 'Geographic') %>%
ggplot(aes(x = x_2,y = delta_1)) +
geom_point() +
geom_errorbar(aes(ymin = lb,ymax = ub),width = .1) +
geom_hline(yintercept = 0,linetype = 'dashed') +
ylab('Margnial Coefficient on Non-Mainstream') + xlab('Logged Deaths') +
theme_ridges() +
ggtitle('Geographic Variation')
pdf('../Figures/SI_figure19.pdf',width = 8,height = 4)
grid.arrange(p1,p2,ncol = 2)
dev.off()