library(dplyr) library(ggplot2) library(memisc) library(lme4) library(ggpubr) library(tidyverse) library(sjPlot) library(foreign) library(MASS) library(ltm) library(car) library(QuantPsyc) library(brglm2) library(Zelig) library(logistf) library(ggplot2) library(ggpubr) library(ggsignif) library(sjPlot) library(dplyr) library(lme4) library(memisc) library(Hmisc) library(nnet) library(psych) library(geobr) library(ggrepel) library(dplyr, warn.conflicts = FALSE) library(xlsx) library(foreign) library(xlsReadWrite) library(electionsBR) #######3 ggplot() + geom_sf(data=round1, fill="Lula_percent", color="#FEBF57", size=.15, show.legend = FALSE) + theme_minimal() + labs(subtitle="States", size=8) rlang::last_error() ############ all94 <- vote_mun_zone_fed(1994) get_elections(year=1994, position="President", regional_aggregation="Municipality", political_aggregation="Candidate") all94$candi_name<-as.factor(all94$NOME_CANDIDATO) summary(eneas94) eneas94<-subset(all94,NOME_CANDIDATO=="ENEAS") eneas94$vote_eneas_94<-eneas94$TOTAL_VOTOS all98 <- vote_mun_zone_fed(1998) eneas98<-subset(all98,NUMERO_CAND==56) eneas98$vote_eneas_98<-eneas98$TOTAL_VOTOS bozo18<-subset(all2018,NUMERO_CANDIDATO==17) bozo18$vote_bozo_18<-bozo18$TOTAL_VOTOS bozo18$SIGLA_UF<-bozo18$UF bozo18$NUMERO_ZONA<-bozo18$NUM_ZONA eneas1<- merge(eneas94,bozo18,by=c("SIGLA_UF","NOME_MUNICIPIO","NUMERO_ZONA"), all.x=TRUE) summary(eneas1) eneas2<- merge(eneas98,bozo18,by=c( "SIGLA_UF","NOME_MUNICIPIO","NUMERO_ZONA"), all.x=TRUE) ########################################### scatter_eneas<-ggplot(eneas1, aes(x=log(vote_eneas_94), y= log(vote_bozo_18)))+ geom_point() + labs(y="", x="")+scale_colour_grey()+ theme_bw()+ggtitle("")+ geom_smooth(color="red")+ theme(legend.title = element_blank())+ theme(plot.title = element_text(hjust = 0.5))+ facet_wrap(.~SIGLA_UF) scatter_eneas ########################################## eneas_94<-subset(votacao_secao_1994_BR,V14==56) sum(eneas_94$V15) eneas_new$candidate<-as.factor(eneas_new$V15) summary(eneas_new$candidate) ################### parties$illiberal<-as.numeric(parties$v2xpa_illiberal) summary(parties) ggplot(subset(parties,year==1990|year==2018), aes(v2pariglef_average, v2paind_average,color=prona_psl, label=(v2pashname)))+ geom_point(size=2)+scale_color_gradient(low="grey", high="dark green")+ theme_classic2()+ggtitle("")+ theme(legend.title = element_blank())+ theme(legend.position= "none")+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ geom_label(aes(label = v2pashname),size=3.5, segment.color = 'grey50', point.padding = 0.8)+ labs(y="Personalização do partido", x="Esquerda Direita")+ theme( axis.text.x=element_blank(), axis.ticks.x=element_blank())+ theme( axis.text.y=element_blank(), axis.ticks.y=element_blank()) prona1<-ggplot(subset(parties,year==1990|year==2018&v2pashname!="PSD"), aes(v2pariglef, v2xpa_popul,color=prona_psl, label=(v2pashname)))+ stat_summary(fun="mean",geom="point",fill="grey",colour="black")+ geom_point(size=2)+scale_color_gradient(low="grey69", high="dark green")+ theme_classic2()+ggtitle("")+ theme(legend.title = element_blank())+ theme(legend.position= "none")+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ geom_label(aes(label = v2pashname),size=3.5, segment.color = 'grey69', point.padding = 0.8)+ labs(y="Índice de populismo do partido", x="Esquerda Direita")+ theme( axis.text.x=element_blank(), axis.ticks.x=element_blank())+ theme( axis.text.y=element_blank(), axis.ticks.y=element_blank()) prona2<-ggplot(subset(parties,year==1990|year==2018&v2pashname!="PSD"), aes(v2xpa_illiberal_average, v2xpa_popul_average,color=prona_psl, label=(v2pashname)))+ geom_point(size=2)+scale_color_gradient(low="grey69", high="dark green")+ theme_classic2()+ggtitle("")+ theme(legend.title = element_blank())+ theme(legend.position= "none")+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ geom_label(aes(label = v2pashname),size=3.5, segment.color = 'grey69', point.padding = 0.8)+ labs(y="Índice de populismo do partido", x="Democrata Autoritário")+ theme( axis.text.x=element_blank(), axis.ticks.x=element_blank())+ theme( axis.text.y=element_blank(), axis.ticks.y=element_blank()) ggarrange(prona2,prona1,nrow=2) ######################3 eneas98<-subset(all98,NUMERO_CANDIDATO==56) eneas98$vote_eneas_98<-eneas98$QTDE_VOTOS bozo18<-subset(all18,NUMERO_CANDIDATO==17) bozo18$vote_bozo_18<-bozo18$QTDE_VOTOS eneas<- merge(eneas98,bozo18,by=c("UF","NOME_MUNICIPIO","NUM_ZONA","NUM_TURNO"), all.x=TRUE) eneas$name_muni<-eneas$NOME_MUNICIPIO eneas$abbrev_state<-eneas$UF eneas<- merge(eneas,ibge_14,by=c("abbrev_state","name_muni")) summary(eneas) eneas$bozo_pc<-eneas$vote_bozo_18/eneas$Pop_2014 eneas$eneas_pc<-eneas$vote_eneas_98/eneas$Pop_2014 ####################################### # plot scatter_eneas<-ggplot(eneas, aes(x=log(vote_eneas_98), y= log(vote_bozo_18)))+ geom_point() + labs(y="", x="")+scale_colour_grey()+ theme_bw()+ggtitle("")+ geom_smooth(color="red")+ theme(legend.title = element_blank())+ theme(plot.title = element_text(hjust = 0.5))+ facet_wrap(.~abbrev_state) scatter_eneas ################# ## MODELS summary(eneas) bozom1<-lm(vote_bozo_18~vote_eneas_98,data = eneas) bozom2<-lm(vote_bozo_18~vote_eneas_98+abbrev_state,data = eneas) bozom3<-lm(vote_bozo_18~vote_eneas_98+gdp_cap_2014 +Pop_2014+ abbrev_state,data = eneas) bozomml1<-lmer(vote_bozo_18~vote_eneas_98+(1|abbrev_state),data = eneas) bozomml2<-lmer(vote_bozo_18~vote_eneas_98+gdp_cap_2014+Pop_2014+(1|abbrev_state),data = eneas) tab_model(bozom1,bozom2,bozom3, show.se = TRUE,collapse.se = T,show.ci = F, p.style = "asterisk") tab_model(bozomml1,bozomml2, show.se = TRUE,collapse.se = T,show.ci = F, p.style = "asterisk") eneasplot<-plot_model(bozom3,type = "pred", terms = "vote_eneas_98")+ scale_colour_grey()+theme_bw()+ labs(y="", x="")+ggtitle("") eneasplot summary(eneas) cor.test(eneas$bozo_pc,eneas$eneas_pc) bozom1pc<-lm(bozo_pc~eneas_pc,data = eneas) bozom2pc<-lm(bozo_pc~eneas_pc+abbrev_state,data = eneas) bozom3pc<-lm(bozo_pc~eneas_pc+log(gdp_cap_2014)+ abbrev_state,data = eneas) tab_model(bozom1pc,bozom2pc,bozom3pc, show.se = TRUE,collapse.se = T,show.ci = F, p.style = "asterisk") ######################################## ####### MAP states <- read_state(year=2019) sates1 <- merge(states,eneas_uf,by=c("abbrev_state")) mapeneas94<-ggplot() + geom_sf(data=sates1, aes(fill=Eneas94_pc), color= NA, size=.15)+ labs(subtitle="Voto Enéas 1994 (%)", size=8) + theme_minimal()+scale_fill_gradient(low='grey', high='black')+ ylab(label="") + xlab("") mapeneas98<-ggplot() + geom_sf(data=sates1, aes(fill=Eneas98_pc), color= NA, size=.15)+ labs(subtitle="Voto Enéas 1998 (%)", size=8) + theme_minimal()+scale_fill_gradient(low='grey', high='black')+ ylab(label="") + xlab("") mapbozo<-ggplot() + geom_sf(data=sates1, aes(fill=Bozo18_pc), color= NA, size=.15)+ labs(subtitle="Voto Bolsonaro 2018 (%)", size=8) + theme_minimal()+scale_fill_gradient(low='grey', high='black')+ ylab(label="") + xlab("") plot_map1<-ggplot(sates1, aes(x=(Eneas94_pc), y= (Bozo18_pc),label=abbrev_state))+ labs(y="Bolsonaro %", x="Enéas 1994 %")+scale_colour_grey()+ theme_bw()+ggtitle("")+ geom_smooth(method=lm,color="red")+ theme(legend.title = element_blank())+ theme(plot.title = element_text(hjust = 0.5))+ geom_text(check_overlap = TRUE) plot_map2<-ggplot(sates1, aes(x=(Eneas98_pc), y= (Bozo18_pc),label=abbrev_state))+ labs(y="Bolsonaro %", x="Enéas 1998 %")+scale_colour_grey()+ theme_bw()+ggtitle("")+ geom_smooth(method=lm,color="red")+ theme(legend.title = element_blank())+ theme(plot.title = element_text(hjust = 0.5))+ geom_text(check_overlap = TRUE) maps<-ggarrange(mapeneas94,mapeneas98,mapbozo,ncol=1) graphs<-ggarrange(plot_map1,plot_map2,ncol = 1) ggarrange(maps,graphs) ############################################################################## ############################################################################## ##### In Sao Paulo ############################################################################## eneas_02<-subset(SP_2002,NUMERO_CANDIDATO==5656) eneas_02$vote_eneas_SP<-eneas_02$QTDE_VOTOS tiririca<-subset(SP_2010,NUMERO_CANDIDATO==2222) tiririca$name_muni<-tiririca$NOME_MUNICIPIO tiririca$tiririca_vote<-tiririca$QTDE_VOTOS eneas_federal<- merge(eneas_02,bozojr_14, by=c("NOME_MUNICIPIO","NUM_ZONA","NUM_TURNO"), all.x=TRUE) eneas_federal_18<- merge(eneas_02,bozojr_18, by=c("NOME_MUNICIPIO","NUM_ZONA","NUM_TURNO"), all.x=TRUE) eneas_federal$name_muni<-eneas_federal$NOME_MUNICIPIO eneas_federal_18$name_muni<-eneas_federal_18$NOME_MUNICIPIO eneas_federal<- merge(eneas_federal,ibge_14_sp, by=c("name_muni"), all.x=TRUE) eneas_federal_18<- merge(eneas_federal_18,ibge_14_sp, by=c("name_muni"), all.x=TRUE) tiririca1<- merge(eneas_federal,tiririca,by=c("NOME_MUNICIPIO","NUM_ZONA","NUM_TURNO"), all.x=TRUE) tiririca2<- merge(eneas_federal_18,tiririca,by=c("NOME_MUNICIPIO","NUM_ZONA","NUM_TURNO"), all.x=TRUE) eneas_sp<- merge(eneas,tiririca1,by=c("NOME_MUNICIPIO","NUM_ZONA","NUM_TURNO"), all.x=TRUE) eneas_sp<-subset(eneas_sp,UF.x="SP") summary(eneas_sp) ### ## MODELS tiririca1<-lm(vote_eneas_98~tiririca_vote,data = eneas_sp) tiririca2<-lm(vote_eneas_98~tiririca_vote+(population)+(gdp_cap_2014.y),data = eneas_sp) bozosp1<-lm(vote_bozo_18~vote_eneas_98,data = eneas_sp) bozosp2<-lm((vote_bozo_18)~(vote_eneas_98)+(population)+(gdp_cap_2014.y),data = eneas_sp) tab_model(tiririca1,tiririca2,bozosp1,bozosp2, show.se = TRUE,collapse.se = T,show.ci = F, p.style = "asterisk") ########### #bid for congress control1<-lm(vote_eneas_SP~tiririca_vote,data = tiririca1) control2<-lm(vote_eneas_SP~tiririca_vote+(population)+(gdp_cap_2014),data = tiririca1) tab_model(control1,control2, show.se = TRUE,collapse.se = T,show.ci = F, p.style = "asterisk") ######## ## only about 15% of Eneas votes for Congress were ideological ############################ ##################### ### MAP SAO PAULO all_muni <- read_municipality( code_muni = "SP", year= 2010) eneas_federal$name_muni<-eneas_federal$NOME_MUNICIPIO eneas_federal_18$name_muni<-eneas_federal_18$NOME_MUNICIPIO SP_14 <- merge(all_muni,eneas_federal,by=c("name_muni")) SP_14 <- merge(SP_14,population_SP,by=c("name_muni")) SP <- merge(all_muni,eneas_federal_18,by=c("name_muni")) SP <- merge(SP,population_SP,by=c("name_muni")) mapeneas_federal<-ggplot() + geom_sf(data=SP, aes(fill=vote_eneas_SP_pc), color= NA, size=.15)+ labs(subtitle="Vote for Enéas 2002 (for Congress)", size=8) + theme_minimal()+scale_fill_gradient(low='grey', high='black')+ ylab(label="") + xlab("") mapbozojr_federal_14<-ggplot() + geom_sf(data=SP_14, aes(fill=vote_bozo_jr_pc), color= NA, size=.15)+ labs(subtitle="Vote for Ed. Bolsonaro 2014 (for Congress)", size=8) + theme_minimal()+scale_fill_gradient(low='grey', high='black')+ ylab(label="") + xlab("") summary(SP_14$vote_bozo_jr_pc) mapbozojr_federal<-ggplot() + geom_sf(data=SP, aes(fill=vote_bozo_jr_18_pc), color= NA, size=.15)+ labs(subtitle="Vote for Ed. Bolsonaro 2018 (for Congress)", size=8) + theme_minimal()+scale_fill_gradient(low='grey', high='black')+ ylab(label="") + xlab("") ggarrange(mapeneas_federal,mapbozojr_federal_14,mapbozojr_federal,nrow=3,ncol = 1) ########################################################################################## ################### Polling data ################################################################################################## ########################################################################################## eneas_poll<- rbind(eneas_polls,eneas_polls2) eneas_poll2<- rbind(eneas_polls, bozo_polls) describe(eneas_poll2) write.csv(eneas_poll2,"/Users/matiaslopez/Downloads/eneas_poll2.csv") ######################################################################3 ## estimating far right appeal in time eneas_poll2$date2<-as.Date(eneas_poll2$date) eneas<-subset(eneas_poll2,year<2018) ggplot(eneas,aes(y=fascist,x=date2))+ stat_summary(fun = "mean", colour = "darkgreen", size = 2, geom = "line")+ theme_classic2() nsamps=1000 date1<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1994-07-05")[,c("fascist","month")],size=300,replace=T)) date2<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1994-07-26")[,c("fascist","month")],size=300,replace=T)) date3<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1994-08-09")[,c("fascist","month")],size=300,replace=T)) date4<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1994-08-18")[,c("fascist","month")],size=300,replace=T)) date5<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1997-12-15")[,c("fascist","month")],size=300,replace=T)) date6<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1998-04-30")[,c("fascist","month")],size=300,replace=T)) date7<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1998-05-28")[,c("fascist","month")],size=300,replace=T)) date8<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1998-07-09")[,c("fascist","month")],size=300,replace=T)) date9<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1998-08-12")[,c("fascist","month")],size=300,replace=T)) date10<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1998-09-02")[,c("fascist","month")],size=300,replace=T)) date11<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1998-09-18")[,c("fascist","month")],size=300,replace=T)) date12<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="1998-10-02")[,c("fascist","month")],size=300,replace=T)) date13<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2001-03-01")[,c("fascist","month")],size=300,replace=T)) date14<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2002-01-03")[,c("fascist","month")],size=300,replace=T)) date15<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2002-02-01")[,c("fascist","month")],size=300,replace=T)) date16<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2002-03-01")[,c("fascist","month")],size=300,replace=T)) date17<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2002-04-09")[,c("fascist","month")],size=300,replace=T)) date18<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2006-05-01")[,c("fascist","month")],size=300,replace=T)) date19<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2018-04-01")[,c("fascist","month")],size=300,replace=T)) date20<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2018-06-01")[,c("fascist","month")],size=300,replace=T)) date21<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2018-08-01")[,c("fascist","month")],size=300,replace=T)) date22<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2018-08-16")[,c("fascist","month")],size=300,replace=T)) date23<-do(nsamps)*lm(fascist~1,data=resample(subset(eneas_poll2,date=="2018-10-05")[,c("fascist","month")],size=300,replace=T)) date1$date<-"1994-07-05" date2$date<-"1994-07-26" date3$date<-"1994-08-09" date4$date<-"1994-08-18" date5$date<-"1997-12-15" date6$date<-"1998-04-30" date7$date<-"1998-05-28" date8$date<-"1998-07-09" date9$date<-"1998-08-12" date10$date<-"1998-09-02" date11$date<-"1998-09-18" date12$date<-"1998-10-02" date13$date<-"2001-03-01" date14$date<-"2002-01-03" date15$date<-"2002-02-01" date16$date<-"2002-03-01" date17$date<-"2002-04-09" date18$date<-"2006-05-01" date19$date<-"2018-04-01" date20$date<-"2018-06-01" date21$date<-"2018-08-01" date22$date<-"2018-08-16" date23$date<-"2018-10-05" date1$year<-"1994" date2$year<-"1994" date3$year<-"1994" date4$year<-"1994" date5$year<-"1997" date6$year<-"1998" date7$year<-"1998" date8$year<-"1998" date9$year<-"1998" date10$year<-"1998" date11$year<-"1998" date12$year<-"1998" date13$year<-"2001" date14$year<-"2002" date15$year<-"2002" date16$year<-"2002" date17$year<-"2002" date18$year<-"2006" date19$year<-"2018" date20$year<-"2018" date21$year<-"2018" date22$year<-"2018" date23$year<-"2018" date1$candidate<-"Enéas" date2$candidate<-"Enéas" date3$candidate<-"Enéas" date4$candidate<-"Enéas" date5$candidate<-"Enéas" date6$candidate<-"Enéas" date7$candidate<-"Enéas" date8$candidate<-"Enéas" date9$candidate<-"Enéas" date10$candidate<-"Enéas" date11$candidate<-"Enéas" date12$candidate<-"Enéas" date13$candidate<-"Enéas" date14$candidate<-"Enéas" date15$candidate<-"Enéas" date16$candidate<-"Enéas" date17$candidate<-"Enéas" date18$candidate<-"Enéas" date19$candidate<-"Bolsonaro" date20$candidate<-"Bolsonaro" date21$candidate<-"Bolsonaro" date22$candidate<-"Bolsonaro" date23$candidate<-"Bolsonaro" shuff_eneas<-rbind(date1,date2,date3,date4,date5,date6,date7,date8,date9,date10, date11,date12,date13,date14,date15,date16,date17,date18,date19, date20,date21,date22,date23) describe(shuff_eneas) shuff_eneas$date2<-as.Date(shuff_eneas$date) shuff_eneas$CI_low<-shuff_eneas$Intercept-1.96*(shuff_eneas$sigma/sqrt(300)) shuff_eneas$CI_upper<-shuff_eneas$Intercept+1.96*(shuff_eneas$sigma/sqrt(300)) plot1<-ggplot(shuff_eneas,aes(y=Intercept,x=date2))+ theme_classic2()+ geom_ribbon(aes(ymin = CI_low, ymax = CI_upper),fill="darkgreen", alpha = 0.5)+ scale_y_continuous(labels = scales::percent, limits=c(0,0.4),breaks = c(0,0.02,0.04,0.06,0.08,0.1,0.2,0.3,0.4))+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ xlab("") + ylab("")+ theme(legend.title = element_blank())+ theme(legend.position = c(0.8, 0.8))+ theme(legend.text = element_text(size=12))+ stat_summary(data=shuff_eneas, aes(y = Intercept), fun=mean, colour="darkgreen", geom="line") plot2 ###################### ### niche eneas_poll2$white_male<-eneas_poll2$male*eneas_poll2$white eneas_poll2$date2<-as.Date(eneas_poll2$date) evangelicos<-subset(eneas_poll2, religion=="Evangélica não pentecostal (Batista, Presbiteriana, Metodist"| religion=="Evangélica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion=="Evangelho Quadrangular"| religion=="Evangélica"| religion=="Evangélica - Não sabe especificar"| religion=="Evangélica Neo Pentecostal"| religion=="Evangélica Tradicional"| religion=="Igreja Internacional da Graça"| religion=="Outras Evangélicas"| religion="Outras Evangélicas específicas"| religion=="Renascer em Cristo"| religion=="Universal do Reino de Deus") evangelicos<-subset(eneas_poll2,evangelico==1) no_evangelicos<-subset(eneas_poll2,white_male==1&education_self>5&age<26&household_income>=3&date=="1994-08-18"&&( religion!="Evang?lica n?o pentecostal (Batista, Presbiteriana, Metodist"| religion!="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion!="Evangelho Quadrangular "| religion!="Evang��lica"| religion!="Evang��lica - N��o sabe especificar"| religion!="Evang��lica Neo Pentecostal"| religion!="Evang��lica Tradicional"| religion!="Igreja Internacional da Gra��a"| religion!="Outras Evang��licas"| religion!="Renascer em Cristo"| religion!="Sara Nossa Terra"| religion!="Universal do Reino de Deus")) describe(no_evangelicos) fTdate1<-subset(eneas_poll2,white_male==1&education_self>5&age<30&date=="1994-08-18") fTdate2<-subset(eneas_poll2,white_male==1&education_self>5&age<30&date=="1997-12-15") fTdate3<-subset(eneas_poll2,white_male==1&education_self>5&age<30&household_income>=3&date=="2002-02-01") fTdate4<-subset(eneas_poll2,white_male==1&education_self>5&age<30&household_income>3&date=="2002-04-09") fTdate5<-subset(eneas_poll2,white_male==1&education_self>5&age<30&household_income>=3&date=="2018-04-01") fTdate6<-subset(eneas_poll2,white_male==1&education_self>5&household_income>=3&date=="2018-06-01") fTdate7<-subset(eneas_poll2,white_male==1&education_self>5&age<30&household_income>=3&date=="2018-08-01") fTdate8<-subset(eneas_poll2,white_male==1&education_self>5&age<30&household_income>=3&date=="2018-08-16") fTdate9<-subset(eneas_poll2,male==1&education_self>5&age<30&household_income>=3&date=="2018-10-05") fCdate1<-subset(eneas_poll2,(date=="1994-08-18")&(white_male==0|(education_self<=5)|(age>=30)|(household_income<3))) fCdate2<-subset(eneas_poll2,(date=="1997-12-15")&(white_male==0|(education_self<=5)|(age>=30))) fCdate3<-subset(eneas_poll2,(date=="2002-02-01")&(white_male==0|(education_self<=5)|(age>=30)|(household_income<3))) fCdate4<-subset(eneas_poll2,(date=="2002-04-09")&(white_male==0|(education_self<=5)|(age>=30)|(household_income<3))) fCdate5<-subset(eneas_poll2,(date=="2018-04-01")&(white_male==0|(education_self<=5)|(age>=30)|(household_income<3))&(evangelico!=1)) fCdate6<-subset(eneas_poll2,(date=="2018-06-01")&(white_male==0|(education_self<=5)|(household_income<3))&(evangelico!=1)) fCdate7<-subset(eneas_poll2,(date=="2018-08-01")&(white_male==0|(education_self<=5)|(age>=30)|(household_income<3))&(evangelico!=1)) fCdate8<-subset(eneas_poll2,(date=="2018-08-16")&(white_male==0|(education_self<=5)|(age>=30)|(household_income<3))&(evangelico!=1)) fCdate9<-subset(eneas_poll2,(date=="2018-10-05")&(white_male==0|(education_self<=5)|(age>=30)|(household_income<3))&(evangelico!=1)) niche<-subset(eneas_poll2,white==1&education_self>5) not_niche<-subset(eneas_poll2,(white==0|(education_self<=5))) niche$niche<-1 not_niche$niche<-0 a<-rbind(niche,not_niche) ggplot(a,aes(y=niche,x=year))+stat_summary(fun.y=mean,geom="line")+ xlim(1994,2018) ### final vote of evangelicos from lapop lapop_bozo<-subset(lapop,vb3n==1501&q3cn==2|5) lapop_bozo_no<-subset(lapop,vb3n!=1501&q3cn==2|5) lapop_bozo$fascist<-1 lapop_bozo_no$fascist<-0 evangelicos_outubro<-rbind(lapop_bozo_no,lapop_bozo) evangelicos$date<-as.Date(evangelicos$date) describe(evangelicos_outubro) nsamps<-1000 Tdate1<-do(nsamps)*lm(fascist~1,data=fTdate1[,c("fascist","month")],size=600,replace=T) Tdate2<-do(nsamps)*lm(fascist~1,data=fTdate2[,c("fascist","month")],size=600,replace=T) Tdate3<-do(nsamps)*lm(fascist~1,data=fTdate3[,c("fascist","month")],size=600,replace=T) Tdate4<-do(nsamps)*lm(fascist~1,data=fTdate4[,c("fascist","month")],size=600,replace=T) Tdate5<-do(nsamps)*lm(fascist~1,data=fTdate5[,c("fascist","month")],size=600,replace=T) Tdate6<-do(nsamps)*lm(fascist~1,data=fTdate6[,c("fascist","month")],size=600,replace=T) Tdate7<-do(nsamps)*lm(fascist~1,data=fTdate7[,c("fascist","month")],size=600,replace=T) Tdate8<-do(nsamps)*lm(fascist~1,data=fTdate8[,c("fascist","month")],size=600,replace=T) Tdate9<-do(nsamps)*lm(fascist~1,data=fTdate9[,c("fascist","month")],size=600,replace=T) Tdate10<-do(nsamps)*lm(fascist~1,data=fTdate10[,c("fascist","month")],size=600,replace=T) Cdate1<-do(nsamps)*lm(fascist~1,data=fCdate1[,c("fascist","month")],size=600,replace=T) Cdate2<-do(nsamps)*lm(fascist~1,data=fCdate2[,c("fascist","month")],size=600,replace=T) Cdate3<-do(nsamps)*lm(fascist~1,data=fCdate3[,c("fascist","month")],size=600,replace=T) Cdate4<-do(nsamps)*lm(fascist~1,data=fCdate4[,c("fascist","month")],size=600,replace=T) Cdate5<-do(nsamps)*lm(fascist~1,data=fCdate5[,c("fascist","month")],size=600,replace=T) Cdate6<-do(nsamps)*lm(fascist~1,data=fCdate6[,c("fascist","month")],size=600,replace=T) Cdate7<-do(nsamps)*lm(fascist~1,data=fCdate7[,c("fascist","month")],size=600,replace=T) Cdate8<-do(nsamps)*lm(fascist~1,data=fCdate8[,c("fascist","month")],size=600,replace=T) Cdate9<-do(nsamps)*lm(fascist~1,data=fCdate9[,c("fascist","month")],size=600,replace=T) Cdate10<-do(nsamps)*lm(fascist~1,data=fCdate10[,c("fascist","month")],size=600,replace=T) Evangelicos1<-do(nsamps)*lm(fascist~1,data=subset(evangelicos,date=="1994-08-18")[,c("fascist","month")],size=600,replace=T) Evangelicos2<-do(nsamps)*lm(fascist~1,data=subset(evangelicos,date=="2018-04-01")[,c("fascist","month")],size=600,replace=T) Evangelicos3<-do(nsamps)*lm(fascist~1,data=subset(evangelicos,date=="2018-06-01")[,c("fascist","month")],size=600,replace=T) Evangelicos4<-do(nsamps)*lm(fascist~1,data=subset(evangelicos,date=="2018-08-01")[,c("fascist","month")],size=600,replace=T) Evangelicos5<-do(nsamps)*lm(fascist~1,data=subset(evangelicos,date=="2018-08-16")[,c("fascist","month")],size=600,replace=T) Evangelicos6<-do(nsamps)*lm(fascist~1,data=evangelicos_outubro[,c("fascist","fecha")],size=600,replace=T) Tdate2$Intercept<-Tdate2$Intercept+(Tdate2$Intercept/2.5) #adding the effect of being white (odds ratio 1.18) Tdate9$Intercept<-Tdate9$Intercept+(Tdate9$Intercept/2.5) Tdate1$date<-"1994-08-18" Tdate2$date<-"1997-12-15" Tdate3$date<-"2002-02-01" Tdate4$date<-"2002-04-09" Tdate5$date<-"2018-04-01" Tdate6$date<-"2018-06-01" Tdate7$date<-"2018-08-01" Tdate8$date<-"2018-08-16" Tdate9$date<-"2018-10-05" Cdate1$date<-"1994-08-18" Cdate2$date<-"1997-12-15" Cdate3$date<-"2002-02-01" Cdate4$date<-"2002-04-09" Cdate5$date<-"2018-04-01" Cdate6$date<-"2018-06-01" Cdate7$date<-"2018-08-01" Cdate8$date<-"2018-08-16" Cdate9$date<-"2018-10-05" Evangelicos1$date<-"1994-08-18" Evangelicos2$date<-"2018-04-01" Evangelicos3$date<-"2018-06-01" Evangelicos4$date<-"2018-08-01" Evangelicos5$date<-"2018-08-16" Evangelicos6$date<-"2018-10-05" Evangelicos1$group<-"Evangélicos" Evangelicos2$group<-"Evangélicos" Evangelicos3$group<-"Evangélicos" Evangelicos4$group<-"Evangélicos" Evangelicos5$group<-"Evangélicos" Evangelicos6$group<-"Evangélicos" Tdate1$group<-"HJBER" Tdate2$group<-"HJBER" Tdate3$group<-"HJBER" Tdate4$group<-"HJBER" Tdate5$group<-"HJBER" Tdate6$group<-"HJBER" Tdate7$group<-"HJBER" Tdate8$group<-"HJBER" Tdate9$group<-"HJBER" Cdate1$group<-"Outros" Cdate2$group<-"Outros" Cdate3$group<-"Outros" Cdate4$group<-"Outros" Cdate5$group<-"Outros" Cdate6$group<-"Outros" Cdate7$group<-"Outros" Cdate8$group<-"Outros" Cdate9$group<-"Outros" shuff_eneas2<-rbind(Tdate1,Tdate2,Tdate3,Tdate4,Tdate5,Tdate6,Tdate8,Tdate9, Cdate1,Cdate2,Cdate3,Cdate4,Cdate5,Cdate6,Cdate8,Cdate9, Evangelicos1,Evangelicos2,Evangelicos3,Evangelicos4,Evangelicos5,Evangelicos6) shuff_eneas2$date2<-as.Date(shuff_eneas2$date) shuff_eneas2$CI_low<-shuff_eneas2$Intercept-1.96*(shuff_eneas2$sigma/sqrt(600)) shuff_eneas2$CI_upper<-shuff_eneas2$Intercept+1.96*(shuff_eneas2$sigma/sqrt(600)) eneassh<-subset(shuff_eneas2,date2<"2018-04-01") bozosh<-subset(shuff_eneas2,date2>="2018-04-01") plot1<-ggplot(eneassh,aes(y=Intercept,x=date2))+ theme_classic2()+ geom_ribbon(data=subset(eneassh,group=="Outros"),aes(ymin = CI_low, ymax = CI_upper),fill="grey", alpha = 0.5)+ geom_ribbon(data=subset(eneassh,group=="Evangélicos"),aes(ymin = CI_low, ymax = CI_upper),fill="lightblue", alpha = 0.6)+ geom_ribbon(data=subset(eneassh,group=="HJBER"),aes(ymin = CI_low, ymax = CI_upper),fill="darkgreen", alpha = 0.6)+ scale_y_continuous(labels = scales::percent, limits=c(0,0.57),breaks = c(0,0.05,0.1,0.15,0.2,0.25,0.3,0.35,0.4,0.45,0.5))+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ xlab("Voto em Enéas") + ylab("")+ theme(legend.title = element_blank())+ theme(legend.position = c(0.2, 0.9))+ theme(legend.text = element_text(size=12))+ stat_summary(data=subset(eneassh,group=="Outros"), aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ stat_summary(data=subset(eneassh,group=="Evangélicos"), aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ stat_summary(data=subset(eneassh,group=="HJBER"), aes(y = Intercept), fun=mean, colour="darkgreen",size=1, geom="line")+ geom_point(aes(colour=factor(group),shape=factor(group)),size=3) plot2<-ggplot(bozosh,aes(y=Intercept,x=date2))+ theme_classic2()+ geom_ribbon(data=subset(bozosh,group=="Outros"),aes(ymin = CI_low, ymax = CI_upper),fill="grey", alpha = 0.5)+ geom_ribbon(data=subset(bozosh,group=="Evangélicos"),aes(ymin = CI_low, ymax = CI_upper),fill="lightblue", alpha = 0.6)+ geom_ribbon(data=subset(bozosh,group=="HJBER"),aes(ymin = CI_low, ymax = CI_upper),fill="darkgreen", alpha = 0.6)+ scale_y_continuous(labels = scales::percent, limits=c(0,0.57),breaks = c(0,0.05,0.1,0.15,0.2,0.25,0.3,0.35,0.4,0.45,0.5))+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ xlab("Voto em Bolsonaro") + ylab("")+ theme(legend.title = element_blank())+ theme(legend.position = "none")+ theme(legend.text = element_text(size=12))+ stat_summary(data=subset(bozosh,group=="Outros"), aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ stat_summary(data=subset(bozosh,group=="Evangélicos"), aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ stat_summary(data=subset(bozosh,group=="HJBER"), aes(y = Intercept), fun=mean, colour="darkgreen",size=1, geom="line")+ geom_point(aes(colour=factor(group),shape=factor(group)),size=3) ggarrange(plot1,plot2,ncol=2) all<-ggplot(shuff_eneas2,aes(y=Intercept,x=date2))+ theme_classic2()+ geom_ribbon(data=subset(shuff_eneas2,group=="Outros"),aes(ymin = CI_low, ymax = CI_upper),fill="grey", alpha = 0.5)+ geom_ribbon(data=subset(shuff_eneas2,group=="Evangélicos"),aes(ymin = CI_low, ymax = CI_upper),fill="lightblue", alpha = 0.6)+ geom_ribbon(data=subset(shuff_eneas2,group=="HJBEA"),aes(ymin = CI_low, ymax = CI_upper),fill="darkgreen", alpha = 0.6)+ scale_y_continuous(labels = scales::percent, limits=c(0,0.6),breaks = c(0,0.05,0.1,0.15,0.2,0.25,0.3,0.35,0.4,0.45,0.5))+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ xlab("Voto em Bolsonaro") + ylab("")+ theme(legend.title = element_blank())+ theme(legend.position = "none")+ theme(legend.text = element_text(size=12))+ stat_summary(data=subset(shuff_eneas2,group=="Outros"), aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ stat_summary(data=subset(shuff_eneas2,group=="Evangélicos"), aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ stat_summary(data=subset(shuff_eneas2,group=="HJBEA"), aes(y = Intercept), fun=mean, colour="darkgreen",size=1, geom="line")+ geom_point(aes(colour=factor(group),shape=factor(group)),size=3) ###########3 # effect of income summary(eneas_poll2$date) eneas_poll2$white_male<-eneas_poll2$white*eneas_poll2$male eneas_poll2$date<-as.Date(eneas_poll2$date) TE1<-subset(eneas_poll2,white_male==1&year<2018&age<30&education_self>5&household_income==1) TE2<-subset(eneas_poll2,white_male==1&year<2018&age<30&education_self>5&household_income==2) TE3<-subset(eneas_poll2,white_male==1&year<2018&age<30&education_self>5&household_income==3) TE4<-subset(eneas_poll2,white_male==1&year<2018&age<30&education_self>5&household_income==4) TE5<-subset(eneas_poll2,white_male==1&year<2018&age<30&education_self>5&household_income==5) TE6<-subset(eneas_poll2,white_male==1&year<2018&age<30&education_self>5&household_income==6) TE7<-subset(eneas_poll2,white_male==1&year<2018&age<30&education_self>5&household_income==7) T1<-do(nsamps)*lm(fascist~1,data=TE1[,c("fascist","month")],size=600,replace=T) T2<-do(nsamps)*lm(fascist~1,data=TE2[,c("fascist","month")],size=600,replace=T) T3<-do(nsamps)*lm(fascist~1,data=TE3[,c("fascist","month")],size=600,replace=T) T4<-do(nsamps)*lm(fascist~1,data=TE4[,c("fascist","month")],size=600,replace=T) T5<-do(nsamps)*lm(fascist~1,data=TE5[,c("fascist","month")],size=600,replace=T) T6<-do(nsamps)*lm(fascist~1,data=TE6[,c("fascist","month")],size=600,replace=T) T7<-do(nsamps)*lm(fascist~1,data=TE7[,c("fascist","month")],size=600,replace=T) T1$income<-1 T2$income<-2 T3$income<-3 T4$income<-4 T5$income<-5 T6$income<-6 T7$income<-7 treats_eneas<-rbind(T1,T2,T3,T4,T5,T6,T7) summary(eneas_poll2$date) BE1<-subset(eneas_poll2,white_male==1&date=="2018-04-01"&age<30&education_self>5&household_income==1) BE2<-subset(eneas_poll2,white_male==1&date=="2018-04-01"&age<30&education_self>5&household_income==2) BE3<-subset(eneas_poll2,white_male==1&date=="2018-04-01"&age<30&education_self>5&household_income==3) BE4<-subset(eneas_poll2,white_male==1&date=="2018-04-01"&age<30&education_self>5&household_income==4) BE5<-subset(eneas_poll2,white_male==1&date=="2018-04-01"&age<30&education_self>5&household_income==5) BE6<-subset(eneas_poll2,white_male==1&date=="2018-04-01"&age<30&education_self>5&household_income==6) BE7<-subset(eneas_poll2,white_male==1&date=="2018-04-01"&age<30&education_self>5&household_income==7) B1<-do(nsamps)*lm(fascist~1,data=BE1[,c("fascist","month")],size=600,replace=T) B2<-do(nsamps)*lm(fascist~1,data=BE2[,c("fascist","month")],size=600,replace=T) B3<-do(nsamps)*lm(fascist~1,data=BE3[,c("fascist","month")],size=600,replace=T) B4<-do(nsamps)*lm(fascist~1,data=BE4[,c("fascist","month")],size=600,replace=T) B5<-do(nsamps)*lm(fascist~1,data=BE5[,c("fascist","month")],size=600,replace=T) B6<-do(nsamps)*lm(fascist~1,data=BE6[,c("fascist","month")],size=600,replace=T) B7<-do(nsamps)*lm(fascist~1,data=BE7[,c("fascist","month")],size=600,replace=T) B1$income<-1 B2$income<-2 B3$income<-3 B4$income<-4 B5$income<-5 B6$income<-6 B7$income<-7 treats_bozo<-rbind(B1,B2,B3,B4,B5,B6,B7) treats_eneas$CI_low<-treats_eneas$Intercept-1.96*(treats_eneas$sigma/sqrt(600)) treats_eneas$CI_upper<-treats_eneas$Intercept+1.96*(treats_eneas$sigma/sqrt(600)) treats_bozo$CI_low<-treats_bozo$Intercept-1.96*(treats_bozo$sigma/sqrt(600)) treats_bozo$CI_upper<-treats_bozo$Intercept+1.96*(treats_bozo$sigma/sqrt(600)) renda_eneas<-ggplot(treats_eneas,aes(x=income,y=Intercept))+ stat_summary(data=treats_eneas, aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ geom_ribbon(data=treats_eneas,aes(ymin = CI_low, ymax = CI_upper),fill="grey", alpha = 0.5)+ theme_classic2()+ scale_y_continuous(labels = scales::percent, limits=c(0,0.65),breaks = c(0,0.05,0.1,0.15,0.2,0.25,0.3,0.35,0.4,0.45,0.5,0.55,0.6,0.65,0.7))+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ xlab("Faixas de renda")+ylab("Voto em Enéas")+ geom_vline(xintercept = 3)+ annotate(geom="text", x=4, y=0.55, label="Mediana", color="black") renda_bozo<-ggplot(treats_bozo,aes(x=income,y=Intercept))+ stat_summary(data=treats_bozo, aes(y = Intercept), fun=mean, colour="black",linetype="dashed", geom="line")+ geom_ribbon(data=treats_bozo,aes(ymin = CI_low, ymax = CI_upper),fill="grey", alpha = 0.5)+ theme_classic2()+ scale_y_continuous(labels = scales::percent, limits=c(0,0.65),breaks = c(0,0.05,0.1,0.15,0.2,0.25,0.3,0.35,0.4,0.45,0.5,0.55,0.6,0.65,0.7))+ theme(axis.text.x = element_text( size = 12, face = "bold",colour="black" ), axis.text.y = element_text( size = 12, face = "bold",colour="black" ))+ xlab("Faixas de renda")+ylab("Voto em Bolsonaro")+ geom_vline(xintercept = 2)+ annotate(geom="text", x=3, y=0.55, label="Mediana", color="black") ggarrange(renda_eneas,renda_bozo) ########################### BBm4<-do(nsamps)*lm(fascist~1,data=resample(Bm4[,c("fascist","month")],size=300,replace=T)) BBm6<-do(nsamps)*lm(fascist~1,data=resample(Bm6[,c("fascist","month")],size=300,replace=T)) BBm8<-do(nsamps)*lm(fascist~1,data=resample(Bm8[,c("fascist","month")],size=300,replace=T)) BBm10<-do(nsamps)*lm(fascist~1,data=resample(Bm10[,c("fascist","month")],size=300,replace=T)) confint(BBm4,level=0.99) confint(BBm6,level=0.99) confint(BBm8,level=0.99) confint(BBm10,level=0.99) confint(yy94,level=0.99) ## making data for male high edu richer my94<-subset(y94,male==1) my97<-subset(y97,male==1) my98<-subset(y98,male==1) my01<-subset(y01,male==1) my02<-subset(y02,male==1) my06<-subset(y06,male==1) me94<-subset(my94,education_self>5) me97<-subset(my97,education_self>5) me98<-subset(my98,education_self>5) me01<-subset(my01,education_self>5) me02<-subset(my02,education_self>5) me06<-subset(my06,education_self>5) mer94<-subset(me94,household_income>2) mer97<-subset(me97,household_income>2) mer98<-subset(me98,household_income>2) mer01<-subset(me01,household_income>2) mer02<-subset(me02,household_income>2) mer06<-subset(me06,household_income>2) mery94<-subset(mer94,age<31) mery97<-subset(me97,age<35) mery98<-subset(me98,age<35) mery01<-subset(mer01,age<38) mery02<-subset(mer02,age<39) mery06<-subset(mer06,age<43) ## making data for eneas foters among male high edu richer younger wy94<-subset(y94,white_male==1) wy97<-subset(y97,white_male==1) wy02<-subset(y02,white_male==1) we94<-subset(wy94,education_self>5) we97<-subset(wy97,education_self>5) we02<-subset(wy02,education_self>5) wer94<-subset(we94,household_income>2) wer97<-subset(we97,household_income>2) wer02<-subset(we02,household_income>2) wery94<-subset(wer94,age<31) wery97<-subset(we97,age<35) wery02<-subset(wer02,age<39) ## data for bozo niche Bw<-subset(bozo_polls,white_male==1) Bwe<-subset(Bw,education_self>5) Bwer<-subset(Bwe,household_income>3) Bwery<-subset(Bwer,age<31) ## simpler stuff bozopop<-do(nsamps)*lm(fascist~1,data=resample(bozo_polls[,c("fascist","month")],size=300,replace=T)) bozonich<-do(nsamps)*lm(fascist~1,data=resample(Bwery[,c("fascist","month")],size=300,replace=T)) confint(bozopop,level=0.99) confint(bozonich,level=0.99) library(coin) table(bozo_polls$household_income,bozo_polls$year) #white males wyy94<-do(nsamps)*lm(fascist~1,data=resample(wy94[,c("fascist","month")],size=300,replace=T)) wyy97<-do(nsamps)*lm(fascist~1,data=resample(wy97[,c("fascist","month")],size=300,replace=T)) wyy02<-do(nsamps)*lm(fascist~1,data=resample(wy02[,c("fascist","month")],size=300,replace=T)) #upperclass MALES Myy94<-do(nsamps)*lm(fascist~1,data=resample(mery94[,c("fascist","month")],size=300,replace=T)) Myy97<-do(nsamps)*lm(fascist~1,data=resample(mery97[,c("fascist","month")],size=300,replace=T)) Myy98<-do(nsamps)*lm(fascist~1,data=resample(mery98[,c("fascist","month")],size=300,replace=T)) Myy01<-do(nsamps)*lm(fascist~1,data=resample(mery01[,c("fascist","month")],size=300,replace=T)) Myy02<-do(nsamps)*lm(fascist~1,data=resample(mery02[,c("fascist","month")],size=300,replace=T)) Myy06<-do(nsamps)*lm(fascist~1,data=resample(mery06[,c("fascist","month")],size=300,replace=T)) #white educated upper class young males Pyy94<-do(nsamps)*lm(fascist~1,data=resample(wery94[,c("fascist","month")],size=300,replace=T)) Pyy97<-do(nsamps)*lm(fascist~1,data=resample(wery97[,c("fascist","month")],size=300,replace=T)) Pyy02<-do(nsamps)*lm(fascist~1,data=resample(wery02[,c("fascist","month")],size=300,replace=T)) # rest of eneas electorate confint(wyy94,level=0.99) confint(wyy97,level=0.99) confint(wyy02,level=0.99) confint(Pyy94,level=0.99) confint(Pyy97,level=0.99) confint(Pyy02,level=0.99) confint(Myy94,level=0.99) confint(Myy97,level=0.99) confint(Myy98,level=0.99) confint(Myy01,level=0.99) confint(Myy02,level=0.99) confint(Myy06,level=0.99) #ading effect of white for 2006 effectW<-glm(fascist~white+male+year+education_self+ household_income,data=subset(eneas_poll2,year<=2018), family="binomial") tab_model(effectW) confint(yy94, level=0.99) confint(yy97, level=0.99) confint(yy98, level=0.99) confint(yy01, level=0.95) confint(yy02, level=0.95) confint(yy06, level=0.95) ##plot eneas_est<-subset(eneas_estimates,Year<=2007) bozo_est<-subset(eneas_estimates,Year>=2007) plot1<-ggplot(data=subset(eneas_est,group=="População"), aes(x=Year, y=Estimate)) + geom_pointrange(aes(ymin=ci_lower, ymax=ci_upper),colour="darkgreen",size=1)+ geom_line(colour="darkgreen",size=1)+ scale_y_continuous(labels = scales::percent, limits=c(0,0.4),breaks = c(0,0.1,0.2,0.3,0.4))+ scale_x_continuous(limits=c(1994,2006),breaks = c(1994,1997,1998,2001,2002,2006))+ theme_classic()+ geom_pointrange(data=subset(eneas_est,group=="Homens brancos jovens com maior educação e renda"),aes(ymin=ci_lower, ymax=ci_upper),colour="blue",size=1)+ geom_line(data=subset(eneas_est,group=="Homens brancos jovens com maior educação e renda"),colour="blue",size=1)+ xlab("Preferência por Enéas") + ylab("")+ theme(axis.title.x = element_text(color="black", size=12, face="bold"), axis.title.y = element_text(color="black", size=24, face="bold"))+ annotate(geom="text", x=2004, y=0.01, label="Total", color="darkgreen",face="bold")+ annotate(geom="text", x=1997, y=0.2, label="Homens jovens brancos\ncom maior renda e escolaridade ", color="blue",face="bold")+ theme(axis.text.x = element_text( size = 11, face = "bold",colour="black" ), axis.text.y = element_text( size = 15, face = "bold",colour="black" )) plot2<-ggplot(data=subset(bozo_est,group=="População"), aes(x=Year, y=Estimate)) + geom_pointrange(data=subset(bozo_est,group=="Homens brancos jovens com maior educação e renda"),aes(ymin=ci_lower, ymax=ci_upper),colour="blue",size=1)+ geom_line(data=subset(bozo_est, group=="Homens brancos jovens com maior educação e renda"),colour="blue",size=1)+ geom_pointrange(aes(ymin=ci_lower, ymax=ci_upper),colour="darkgreen",size=1)+ geom_line(colour="darkgreen",size=1)+ scale_y_continuous(labels = scales::percent, limits=c(0,0.4),breaks = c(0,0.1,0.2,0.3,0.4))+ theme_classic()+ scale_x_discrete(labels = c("") )+ xlab("Preferência por Bolsonaro 2018") + ylab("")+ theme(axis.title.x = element_text(color="black", size=12, face="bold"), axis.title.y = element_text(color="black", size=24, face="bold"))+ annotate(geom="text", x=2018.4, y=0.01, label="Abril", color="black")+ annotate(geom="text", x=2018.6, y=0.01, label="Junho", color="black")+ annotate(geom="text", x=2018.8, y=0.01, label="Agosto", color="black")+ annotate(geom="text", x=2019, y=0.01, label="Outubro", color="black")+ theme(axis.text.x = element_text( size = 11, face = "bold",colour="black" ), axis.text.y = element_text( size = 15, face = "bold",colour="black" )) ggarrange(plot1,plot2,nrow=2) ######################################################################### library(lubridate) ymd(eneas_poll$date) neas_elections<-subset(eneas_poll,year!=1997) neas_elections<-subset(neas_elections,year!=2001) eneas_poll2$Month<-as.factor(eneas_poll2$month) fascists<-ggplot(eneas_poll2, aes(x=(Month), y= (fascist)))+ labs(y="", x="")+scale_colour_grey()+ stat_summary(fun=mean,geom="point",fill="grey",colour="black")+ stat_summary(fun.data=mean_cl_normal,geom="errorbar",width=.3)+ theme_bw()+ggtitle("")+ geom_smooth(se=T,color="black")+ theme(legend.title = element_blank())+ theme(plot.title = element_text(hjust = 0.5))+ scale_y_continuous(labels = scales::percent)+facet_wrap(.~year) fascists ###################################### #### MODELS ####################################################3 eneas_poll2$Homem<-as.factor(eneas_poll2$male) eneas_poll2$Branco<-as.factor(eneas_poll2$white) eneas_poll2$SES<-(eneas_poll2$education_self*eneas_poll2$household_income) eneas_poll2$SES_log<-log(eneas_poll2$SES) eneas_poll2$white_male<-eneas_poll2$white*eneas_poll2$male eneas_poll2$Homem_branco<-as.factor(eneas_poll2$white_male) eneas_poll2$White_male<-as.factor(eneas_poll2$white_male) eneas_poll2$ano<-as.factor(eneas_poll2$year) evan1<-subset(eneas_poll2,(date2=="1994-08-18")&( religion=="Evang?lica n?o pentecostal (Batista, Presbiteriana, Metodist"| religion=="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion=="Evangelho Quadrangular "| religion=="Evang��lica"| religion=="Evang��lica - N��o sabe especificar"| religion=="Evang��lica Neo Pentecostal"| religion=="Evang��lica Tradicional"| religion=="Igreja Internacional da Gra��a"| religion=="Outras Evang��licas"| religion=="Renascer em Cristo"| religion=="Sara Nossa Terra"| religion=="Universal do Reino de Deus")) summary(evan4) evan2<-subset(eneas_poll2,(date2=="2018-04-01")&( religion=="Evang?lica n?o pentecostal (Batista, Presbiteriana, Metodist"| religion=="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion=="Evangelho Quadrangular "| religion=="Evang��lica"| religion=="Evang��lica - N��o sabe especificar"| religion=="Evang��lica Neo Pentecostal"| religion=="Evang��lica Tradicional"| religion=="Igreja Internacional da Gra��a"| religion=="Outras Evang��licas"| religion=="Renascer em Cristo"| religion=="Sara Nossa Terra"| religion=="Universal do Reino de Deus")) evan3<-subset(eneas_poll2,(date2=="2018-06-01")&( religion=="Evang?lica n?o pentecostal (Batista, Presbiteriana, Metodist"| religion=="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion=="Evangelho Quadrangular "| religion=="Evang��lica"| religion=="Evang��lica - N��o sabe especificar"| religion=="Evang��lica Neo Pentecostal"| religion=="Evang��lica Tradicional"| religion=="Igreja Internacional da Gra��a"| religion=="Outras Evang��licas"| religion=="Renascer em Cristo"| religion=="Sara Nossa Terra"| religion=="Universal do Reino de Deus")) evan4<-subset(eneas_poll2,(date2=="2018-08-01")&( religion=="Evang?lica n?o pentecostal (Batista, Presbiteriana, Metodist"| religion=="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion=="Evangelho Quadrangular "| religion=="Evang��lica"| religion=="Evang��lica - N��o sabe especificar"| religion=="Evang��lica Neo Pentecostal"| religion=="Evang��lica Tradicional"| religion=="Igreja Internacional da Gra��a"| religion=="Outras Evang��licas"| religion=="Renascer em Cristo"| religion=="Sara Nossa Terra"| religion=="Universal do Reino de Deus")) evan5<-subset(eneas_poll2,(date2=="2018-08-16")&( religion=="Evang?lica n?o pentecostal (Batista, Presbiteriana, Metodist"| religion=="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion=="Evangelho Quadrangular "| religion=="Evang��lica"| religion=="Evang��lica - N��o sabe especificar"| religion=="Evang��lica Neo Pentecostal"| religion=="Evang��lica Tradicional"| religion=="Igreja Internacional da Gra��a"| religion=="Outras Evang��licas"| religion=="Renascer em Cristo"| religion=="Sara Nossa Terra"| religion=="Universal do Reino de Deus")) NOevan1<-subset(eneas_poll2,(date2=="1994-08-18")&( religion!="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion!="Evangelho Quadrangular "| religion!="Evang��lica"| religion!="Evang��lica - N��o sabe especificar"| religion!="Evang��lica Neo Pentecostal"| religion!="Evang��lica Tradicional"| religion!="Igreja Internacional da Gra��a"| religion!="Outras Evang��licas"| religion!="Renascer em Cristo"| religion!="Sara Nossa Terra"| religion!="Universal do Reino de Deus")) NOevan2<-subset(eneas_poll2,(date2=="2018-04-01")&( religion!="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion!="Evangelho Quadrangular "| religion!="Evang��lica"| religion!="Evang��lica - N��o sabe especificar"| religion!="Evang��lica Neo Pentecostal"| religion!="Evang��lica Tradicional"| religion!="Igreja Internacional da Gra��a"| religion!="Outras Evang��licas"| religion!="Renascer em Cristo"| religion!="Sara Nossa Terra"| religion!="Universal do Reino de Deus")) NOevan3<-subset(eneas_poll2,(date2=="2018-06-01")&( religion!="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion!="Evangelho Quadrangular "| religion!="Evang��lica"| religion!="Evang��lica - N��o sabe especificar"| religion!="Evang��lica Neo Pentecostal"| religion!="Evang��lica Tradicional"| religion!="Igreja Internacional da Gra��a"| religion!="Outras Evang��licas"| religion!="Renascer em Cristo"| religion!="Sara Nossa Terra"| religion!="Universal do Reino de Deus")) NOevan4<-subset(eneas_poll2,(date2=="2018-08-01")&( religion!="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion!="Evangelho Quadrangular "| religion!="Evang��lica"| religion!="Evang��lica - N��o sabe especificar"| religion!="Evang��lica Neo Pentecostal"| religion!="Evang��lica Tradicional"| religion!="Igreja Internacional da Gra��a"| religion!="Outras Evang��licas"| religion!="Renascer em Cristo"| religion!="Sara Nossa Terra"| religion!="Universal do Reino de Deus")) NOevan5<-subset(eneas_poll2,(date2=="2018-08-16")&( religion!="Evang?lica n?o pentecostal (Batista, Presbiteriana, Metodist"| religion!="Evang?lica pentecostal (Congrega??o Crist? no Brasil, Assemb"| religion!="Evangelho Quadrangular "| religion!="Evang��lica"| religion!="Evang��lica - N��o sabe especificar"| religion!="Evang��lica Neo Pentecostal"| religion!="Evang��lica Tradicional"| religion!="Igreja Internacional da Gra��a"| religion!="Outras Evang��licas"| religion!="Renascer em Cristo"| religion!="Sara Nossa Terra"| religion!="Universal do Reino de Deus")) evan1$Evangelico<-1 evan2$Evangelico<-1 evan3$Evangelico<-1 evan4$Evangelico<-1 evan5$Evangelico<-1 NOevan1$Evangelico<-0 NOevan2$Evangelico<-0 NOevan3$Evangelico<-0 NOevan4$Evangelico<-0 NOevan5$Evangelico<-0 eneas_poll2_relig<-rbind(evan1,evan2,evan3,evan4,evan5,NOevan1,NOevan2,NOevan3,NOevan4,NOevan5) eneas_poll<-subset(eneas_poll2,year<2018) eneas_poll_relig<-subset(eneas_poll2_relig,year<2018) bozo_poll<-subset(eneas_poll2,year==2018) bozo_poll_relig<-subset(eneas_poll2_relig,year==2018) vote1<-glm(vote_eneas~education_self+household_income+ age+ano,family=binomial,data=eneas_poll) vote2<-glm(vote_eneas~Homem_branco+education_self+household_income+ age+ano,family=binomial,data=eneas_poll) vote3<-glm(vote_eneas~Homem_branco*education_self+ Homem_branco*household_income+ Homem_branco*age+Homem_branco+ano,family=binomial,data=eneas_poll) vote4<-glm(vote_eneas~Evangelico+Homem_branco*education_self+ Homem_branco*household_income+ Homem_branco*age,family=binomial,data=eneas_poll_relig) vote5<-glm(fascist~education_self+household_income+ age,family=binomial,data=bozo_poll) vote6<-glm(fascist~Homem_branco+education_self+household_income+ age,family=binomial,data=bozo_poll) vote7<-glm(fascist~Homem_branco*education_self+ Homem_branco*household_income+ Homem_branco*age+Homem_branco,family=binomial,data=bozo_poll) vote8<-glm(vote_eneas~Evangelico+Homem_branco*education_self+ Homem_branco*household_income+ Homem_branco*age+Homem_branco,family=binomial,data=eneas_poll) vote_ideo_eneas<-glm(vote_eneas~favors_death_penalty,family=binomial,data=subset(eneas_poll,year==2006)) vote_ideo_eneas<-lm(fascist~1,data=subset(eneas_poll,favors_death_penalty==1&education_self>5&age<31&household_income>3)) summary(vote_ideo_eneas) eneas_poll02<-subset(eneas_poll,year==2002) eneas_poll06<-subset(eneas_poll,year==2006) table(eneas_poll02$favors_death_penalty,eneas_poll02$fascist) table(eneas_poll06$favors_death_penalty,eneas_poll06$fascist) tab_model(vote1,vote2, vote3,vote4,vote5,vote6,vote7,vote8, show.se = TRUE,collapse.se = T,show.ci = F, p.style = "asterisk") tab_model(vote2) renda_eneas<-plot_model(vote1, type = "pred", terms = c("household_income"))+ ggtitle("Voto em Enéas por faixa de renda")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.2),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8)) educ_eneas<-plot_model(vote1, type = "pred", terms = c("education_self"))+ ggtitle("Voto em Enéas por faixa de escolaridade")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.2),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8)) idade_eneas<-plot_model(vote1, type = "pred", terms = c("age"))+ ggtitle("Voto em Enéas por idade")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.2),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8))+ xlim(16,90) brancos_eneas<-plot_model(vote2, type = "pred", terms = c("Homem_branco"))+ ggtitle("Voto em Enéas entre homens brancos")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.2),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8)) renda_bozo<-plot_model(vote5, type = "pred", terms = c("household_income"))+ ggtitle("Voto em Bolsonaro por faixa de renda")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.8),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8)) educ_bozo<-plot_model(vote5, type = "pred", terms = c("education_self"))+ ggtitle("Voto em Bolsonaro por faixa de escolaridade")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.8),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8)) idade_bozo<-plot_model(vote5, type = "pred", terms = c("age"))+ ggtitle("Voto em Bolsonaro por idade")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.8),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8))+ xlim(16,90) brancos_bozo<-plot_model(vote6, type = "pred", terms = c("Homem_branco"))+ ggtitle("Voto em Bolsonaro entre homens brancos")+xlab("") + ylab("")+ scale_fill_grey()+theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.8),breaks = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8)) bozo_interact1<-plot_model(vote6, type = "pred", terms = c("household_income","Homem_branco"),colors="bw", ci.lvl = .90)+ ggtitle("Voto em Bolsonaro")+xlab("Renda domiciliar") + ylab("")+ theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.5),breaks = c(0,0.1,0.2,0.3,0.4,0.5))+ theme(legend.position = "none")+ xlim(2,6) eneas_interact1<-plot_model(vote3, type = "pred", terms = c("household_income","Homem_branco"),colors="bw", ci.lvl = .90)+ ggtitle("Voto em Enéas")+xlab("Renda domiciliar") + ylab("")+ theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.15),breaks = c(0,0.05,0.1,0.15,0.2))+ theme(legend.position = "none")+ xlim(2,6) bozo_interact2<-plot_model(vote6, type = "pred", terms = c("education_self","Homem_branco"),colors="bw", ci.lvl = .90)+ ggtitle("Voto em Bolsonaro")+xlab("Nível de educação formal") + ylab("")+ theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.5),breaks = c(0,0.1,0.2,0.3,0.4,0.5))+ theme(legend.position = c(0.8, 0.8)) eneas_interact2<-plot_model(vote3, type = "pred", terms = c("education_self","Homem_branco"),colors="bw", ci.lvl = .90)+ ggtitle("Voto em Enéas")+xlab("Nível de educação formal") + ylab("")+ theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.15),breaks = c(0,0.05,0.1,0.15,0.2))+ theme(legend.position = "none") bozo_interact3<-plot_model(vote6, type = "pred", terms = c("age","Homem_branco"),colors="bw", ci.lvl = .90)+ ggtitle("Voto em Bolsonaro")+xlab("Idade") + ylab("")+ theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.5),breaks = c(0,0.1,0.2,0.3,0.4,0.5))+ theme(legend.position = "none") eneas_interact3<-plot_model(vote3, type = "pred", terms = c("age","Homem_branco"),colors="bw", ci.lvl = .90)+ ggtitle("Voto em Enéas")+xlab("Idade") + ylab("")+ theme_classic2()+ theme(plot.title = element_text(hjust=0.5))+ theme(axis.text.x=element_text(size=12, face="bold",colour="black"))+ theme(axis.text.y=element_text(size=12, face="bold",colour="black"))+ scale_y_continuous(labels = scales::percent, limits=c(0,0.15),breaks = c(0,0.05,0.1,0.15,0.2))+ theme(legend.position = "none") ggarrange(eneas_interact3,eneas_interact1,eneas_interact2, bozo_interact3,bozo_interact1,bozo_interact2) ggarrange(renda_eneas,educ_eneas,idade_eneas, renda_bozo,educ_bozo,idade_bozo, ncol = 3,nrow = 2) bozo_polls$Homem<-as.factor(bozo_polls$male) bozo_polls$Branco<-as.factor(bozo_polls$white) bozo_polls$SES<-(bozo_polls$education_self*bozo_polls$household_income) bozo_polls$SES_log<-log(bozo_polls$SES) bozo_polls$white_male<-bozo_polls$white*bozo_polls$male bozo_polls$Homem_branco<-as.factor(bozo_polls$white_male) bozo_polls$White_male<-as.factor(bozo_polls$white_male) vote_bozo<-glm(vote_bozo~SES_log*Homem_branco+ age+Homem_branco*South,family=binomial,data=bozo_polls) vote_bozo_en<-glm(vote_bozo~SES_log*White_male+ age+White_male*South,family=binomial,data=bozo_polls) early_vote_bozo<-glm(vote_bozo~SES*Homem_branco+ age+Homem_branco*South,family=binomial,data=subset(bozo_polls,month==4)) final_vote_bozo<-glm(vote_bozo~SES*Homem_branco+ age+Homem_branco*South,family=binomial,data=subset(bozo_polls,month==8)) tab_model(vote1,vote2,vote3,vote4,vote_bozo, show.se = TRUE,collapse.se = T,show.ci = F, p.style = "asterisk",p.threshold = c(0.10, 0.05, 0.01)) pp_bozo<-plot_model(vote_bozo_en, type = "pred", terms = c("SES_log","White_male"))+ ggtitle("Bolsonaro")+xlab("Socio-Economic Status (log)") + ylab("")+ theme_bw()+ theme(plot.title = element_text(hjust=0.5)) ggarrange(pp1,pp_bozo) pp2<-plot_model(vote0, type = "pred", terms = c("age","Homem_branco"))+ ggtitle("Idade")+xlab("") + ylab("")+ theme_bw()+ theme(plot.title = element_text(hjust=0.5))