add 23

23/paper.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af1909814ddf0ce5ad9c91bd3cdfae0676eb304e8c92c438d56fd9e8d77a407f
+size 2076097

23/replication_package/Code/Manekin_Mitts_Effective_for_Whom_Replication.R

@@ -0,0 +1,128 @@
+# ----------------------------------------------------------------------
+# Replication for "Effective for Whom? Ethnic Identity and Nonviolent Resistance"
+# Devorah Manekin and Tamar Mitts
+# American Political Science Review (2021)
+# ----------------------------------------------------------------------
+# This code replicates the tables figures in the article and the appendix.
+# It requires the packages listed below. If not already installed, 
+# use the function install.packages() to install them.
+# ----------------------------------------------------------------------
+
+rm(list=ls())
+
+library(lemon)
+library(stargazer)
+library(xtable)
+library(fastDummies)
+library(tidyr)
+library(gridExtra)
+library(stm)
+library(quanteda)
+library(ggplot2)
+library(dotwhisker)
+library(ggeffects)
+library(estimatr)
+library(corrplot)
+
+# SET WORKING DIRECTORY
+setwd("/Users/tamarmitts/Dropbox (Mitts)/Projects/Non-violence/Replication/")
+
+# LOAD DATASETS
+load("Data/EBCR_EPR_NAVCO2.rdata")
+load("Data/NAVCO2_EPR.rdata")
+load("Data/NAVCO2.rdata")
+load("Data/us_survey_wave1.rdata")
+load("Data/isr_survey_wave1.rdata")
+load("Data/us_survey_wave2.rdata")
+load("Data/isr_survey_wave2.rdata")
+load("Data/us_survey_text_analysis.rdata")
+load("Data/isr_survey_text_analysis_eth.rdata")
+load("Data/isr_survey_text_analysis_arab.rdata")
+
+##--##--##--##
+## Manuscript
+##--##--##--##
+
+# Figure 1
+source("Code/replicate_fig_1.R")
+# Figure 3
+source("Code/replicate_fig_3.R")
+# Figure 4
+source("Code/replicate_fig_4.R")
+# Figure 6
+source("Code/replicate_fig_6.R")
+# Figure 7
+source("Code/replicate_fig_7.R")
+# Figure 8
+source("Code/replicate_fig_8.R")
+# Figure 9
+source("Code/replicate_fig_9.R")
+# Figure 10
+source("Code/replicate_fig_10.R")
+
+# Table 1
+source("Code/replicate_table_1.R")
+# Table 4
+source("Code/replicate_table_4.R")
+
+##--##--##--##
+## Appendix
+##--##--##--##
+
+# Figure A1
+source("Code/replicate_fig_A1.R")
+# Figure A2
+source("Code/replicate_fig_A2.R")
+# Figure A3
+source("Code/replicate_fig_A3.R")
+# Figure A6
+source("Code/replicate_fig_A6.R")
+# Figure A7
+source("Code/replicate_fig_A7.R")
+# Figure A8
+source("Code/replicate_fig_A8.R")
+# Figure A9
+source("Code/replicate_fig_A9.R")
+# Figure A10
+source("Code/replicate_fig_A10.R")
+# Figure A11
+source("Code/replicate_fig_A11.R")
+# Figure A11
+source("Code/replicate_fig_A12.R")
+
+# Table A1
+source("Code/replicate_table_A1.R")
+# Table A2
+source("Code/replicate_table_A2.R")
+# Table A3
+source("Code/replicate_table_A3.R")
+# Table A4
+source("Code/replicate_table_A4.R")
+# Table A5
+source("Code/replicate_table_A5.R")
+# Table A6
+source("Code/replicate_table_A6.R")
+# Table A7
+source("Code/replicate_table_A7.R")
+# Table A8
+source("Code/replicate_table_A8.R")
+# Table A9
+source("Code/replicate_table_A9.R")
+# Table A10
+source("Code/replicate_table_A10.R")
+# Table A10
+source("Code/replicate_table_A11.R")
+# Table A13
+source("Code/replicate_table_A13.R")
+# Table A14
+source("Code/replicate_table_A14.R")
+# Table A15
+source("Code/replicate_table_A15.R")
+# Table A16
+source("Code/replicate_table_A16.R")
+# Table A17
+source("Code/replicate_table_A17.R")
+# Table A18
+source("Code/replicate_table_A18.R")
+
+## END

23/replication_package/Code/replicate_fig_1.R

@@ -0,0 +1,62 @@
+##--##--##--##
+## Figure 1
+##--##--##--##
+
+# Status (excluded, not excluded)
+nv_status = lm(success ~  EPR_STATUS_EXCL + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_NV_ONSET==1,])  
+v_status = lm(success ~  EPR_STATUS_EXCL + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT  + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_V_ONSET==1,])  
+
+preds_nv = ggpredict(nv_status, terms = c("EPR_STATUS_EXCL"))
+preds_nv$tactic = "Non-violent"
+preds_v = ggpredict(v_status, terms = c("EPR_STATUS_EXCL"))
+preds_v$tactic = "Violent"
+
+preds = rbind(preds_nv, preds_v)
+preds$tactic <- factor(preds$tactic, levels = c("Violent", "Non-violent"))
+preds$Status = "Minority/disadvantaged"
+preds$Status[preds$x==0] = "Majority/dominant"
+
+xtable(preds[,-1], digits=2)
+preds = as.data.frame(preds)
+preds$Status = as.factor(preds$Status)
+
+status = ggplot(preds, aes(x = tactic, y = predicted, group=Status)) +
+  geom_line(aes(color=Status))+ geom_errorbar(aes(color = Status), width = 0, ymin=preds$conf.low, ymax=preds$conf.high)+
+  geom_point(aes(color=Status, shape=Status), size=3) + ylim(-0.1,0.5) +  theme_bw() +
+  scale_colour_grey(start = 0.1, end = 0.65) + geom_hline(yintercept=0, linetype="dashed")+
+  ylab("Pr(Campaign Success)") + xlab("Tactic") +ggtitle("(A) Group Status") +
+  theme(legend.position="none")
+
+## Size (above and below the mean of the distribution)
+EBCR_EPR_NAVCO2$small_size = NA
+EBCR_EPR_NAVCO2$small_size[EBCR_EPR_NAVCO2$EPR_GROUPSIZE < mean(EBCR_EPR_NAVCO2$EPR_GROUPSIZE, na.rm=T)] = 1
+EBCR_EPR_NAVCO2$small_size[EBCR_EPR_NAVCO2$EPR_GROUPSIZE >= mean(EBCR_EPR_NAVCO2$EPR_GROUPSIZE, na.rm=T)] = 0
+
+nv_size = lm(success ~  small_size  + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT  + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_NV_ONSET==1,])  
+v_size = lm(success ~  small_size  + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_V_ONSET==1,])  
+
+preds_nv = ggpredict(nv_size, terms = c("small_size"))
+preds_nv$tactic = "Non-violent"
+preds_v = ggpredict(v_size, terms = c("small_size"))
+preds_v$tactic = "Violent"
+
+preds = rbind(preds_nv, preds_v)
+preds$tactic <- factor(preds$tactic, levels = c("Violent", "Non-violent"))
+preds$Status = "Minority/disadvantaged"
+preds$Status[preds$x==0] = "Majority/dominant"
+
+preds = as.data.frame(preds)
+preds$Status = as.factor(preds$Status)
+
+xtable(preds[,-1], digits=2)
+
+size = ggplot(preds, aes(x = tactic, y = predicted, group=Status)) +
+  geom_line(aes(color=Status))+ geom_errorbar(aes(color = Status), width = 0, ymin=preds$conf.low, ymax=preds$conf.high)+
+  geom_point(aes(color=Status, shape=Status), size=3) + ylim(-0.1,0.5) + theme_bw() +
+  #scale_color_manual(values=c("#E69F00", "#56B4E9")) + geom_hline(yintercept=0, linetype="dashed")+
+  scale_colour_grey(start = 0.1, end = 0.65) + geom_hline(yintercept=0, linetype="dashed")+
+  ylab("Pr(Campaign Success)") + xlab("Tactic")  +ggtitle("(B) Group Size") +
+  theme(legend.position="none")
+
+fig1 = grid_arrange_shared_legend(status, size, ncol = 2, nrow = 1)
+ggsave(file="Figures/fig_1.pdf", fig1, width=8, height=4)

23/replication_package/Code/replicate_fig_10.R

@@ -0,0 +1,36 @@
+##--##--##--##
+## Figure 10
+##--##--##--##
+
+isr_survey_eth_content_covars <- estimateEffect(1:10 ~ identity_protesters_eth, isr_survey_text_analysis_eth$stm_topics, meta = docvars(isr_survey_text_analysis_eth$nv_dfm_eth), uncertainty = "Global")
+isr_survey_eth_sum_content_covars = summary(isr_survey_eth_content_covars)
+
+term_translated = c("protest, right, democracy, ethiopians",
+                    "discrimination, justice, understand, frustration",
+                    "equal, rights, treatment, deserve",
+                    "racism, violence, protest",
+                    "judge, physical, right, treatment",
+                    "economic, crisis, unemployment, justified",
+                    "support, people, understand",
+                    "voice, protest, express",
+                    "ethiopian, community, racism",
+                    "violence, express, furious")
+
+
+estimate = rep(NA, 10)
+std.error = rep(NA, 10)
+for(i in 1:length(isr_survey_eth_sum_content_covars$tables)){
+  estimate[i] = isr_survey_eth_sum_content_covars$tables[[i]][2,1]
+  std.error[i] = isr_survey_eth_sum_content_covars$tables[[i]][2,2]
+}
+
+isr_survey_eth_stm_results = data.frame("term" = term_translated, "estimate" = estimate, "std.error"= std.error)
+
+isr_survey_eth_stm_results = isr_survey_eth_stm_results[order(isr_survey_eth_stm_results$estimate, decreasing = T),]
+
+dwplot(isr_survey_eth_stm_results, 
+       vline = geom_vline(xintercept = 0, colour = "grey60", linetype = 2), dot_args = list(size = 2.5)) +
+  theme_bw() + xlab("\nIsraeli white protesters . . . . . . . . . . . . . Israeli Ethiopian protesters") +
+  theme(text = element_text(size=12), axis.text.x = element_text(angle = 0, hjust = 1), legend.position = "none") +
+  scale_colour_grey(start = 0, end = 0)
+ggsave(file="Figures/fig_10.pdf", width=8, height=6)

23/replication_package/Code/replicate_fig_3.R

@@ -0,0 +1,204 @@
+##--##--##--##
+## Figure 3
+##--##--##--##
+
+us_survey_wave1$identity_protesters_fac = as.factor(us_survey_wave1$identity_protesters)
+levels(us_survey_wave1$identity_protesters_fac) = c("White", "Black")
+us_survey_wave1$tactic_fac = as.factor(us_survey_wave1$tactic)
+levels(us_survey_wave1$tactic_fac) = c("March in streets", "Shut down traffic", "Destroy police cars")
+
+isr_survey_wave1$identity_protesters_fac = as.factor(isr_survey_wave1$identity_protesters)
+levels(isr_survey_wave1$identity_protesters_fac) = c("White", "Ethiopian", "Arab")
+isr_survey_wave1$tactic_fac = as.factor(isr_survey_wave1$tactic)
+levels(isr_survey_wave1$tactic_fac) = c("March in streets", "Shut down traffic", "Destroy garbage cans")
+
+
+## US sample
+us_survey_wave1$degree_violence_std = scale(us_survey_wave1$degree_violence)
+us_survey_wave1$police_action_required_std = scale(us_survey_wave1$police_action_required)
+us_survey_wave1$recall_violence2_std = scale(us_survey_wave1$recall_violence2)
+
+march1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+shut1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==1,], weights=weight)
+destroy1 = lm_robust(degree_violence_std ~ identity_protesters_fac , data=us_survey_wave1[us_survey_wave1$tactic==2,], weights=weight)
+
+march2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+shut2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==1,], weights=weight)
+destroy2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==2,], weights=weight)
+
+march3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+shut3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==1,], weights=weight)
+destroy3 = lm_robust(recall_violence2_std ~ identity_protesters_fac , data=us_survey_wave1[us_survey_wave1$tactic==2,], weights=weight)
+
+# Plot differences
+
+term_degree_violence = c(march1$coefficients[2], shut1$coefficients[2], destroy1$coefficients[2])
+se_degree_violence = c(march1$std.error[2], shut1$std.error[2], destroy1$std.error[2])
+statistic_degree_violence = c(march1$statistic[2], shut1$statistic[2], destroy1$statistic[2])
+pval_degree_violence = c(march1$p.value[2], shut1$p.value[2], destroy1$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+term_police_action_required = c(march2$coefficients[2], shut2$coefficients[2], destroy2$coefficients[2])
+se_police_action_required = c(march2$std.error[2], shut2$std.error[2], destroy2$std.error[2])
+statistic_police_action_required = c(march2$statistic[2], shut2$statistic[2], destroy2$statistic[2])
+pval_police_action_required = c(march2$p.value[2], shut2$p.value[2], destroy2$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+
+term_recall_violence = c(march3$coefficients[2], shut3$coefficients[2], destroy3$coefficients[2])
+se_recall_violence = c(march3$std.error[2], shut3$std.error[2], destroy3$std.error[2])
+statistic_recall_violence = c(march3$statistic[2], shut3$statistic[2], destroy3$statistic[2])
+pval_recall_violence = c(march3$p.value[2], shut3$p.value[2], destroy3$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+
+differences_us = rbind(degree_violence, police_action_required, recall_violence)
+
+
+## Israel respondents
+isr_survey_wave1$degree_violence_std = scale(isr_survey_wave1$degree_violence)
+isr_survey_wave1$police_action_required_std = scale(isr_survey_wave1$police_action_required)
+isr_survey_wave1$recall_violence2_std = scale(isr_survey_wave1$recall_violence2)
+
+march1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+shut1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==1,], weights=weight)
+destroy1 = lm_robust(degree_violence_std ~ identity_protesters_fac , data=isr_survey_wave1[isr_survey_wave1$tactic==2,], weights=weight)
+
+march2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+shut2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==1,], weights=weight)
+destroy2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==2,], weights=weight)
+
+march3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+shut3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==1,], weights=weight)
+destroy3 = lm_robust(recall_violence2_std ~ identity_protesters_fac , data=isr_survey_wave1[isr_survey_wave1$tactic==2,], weights=weight)
+
+
+# Ethiopian protesters
+term_degree_violence = c(march1$coefficients[2], shut1$coefficients[2], destroy1$coefficients[2])
+se_degree_violence = c(march1$std.error[2], shut1$std.error[2], destroy1$std.error[2])
+statistic_degree_violence = c(march1$statistic[2], shut1$statistic[2], destroy1$statistic[2])
+pval_degree_violence = c(march1$p.value[2], shut1$p.value[2], destroy1$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+term_police_action_required = c(march2$coefficients[2], shut2$coefficients[2], destroy2$coefficients[2])
+se_police_action_required = c(march2$std.error[2], shut2$std.error[2], destroy2$std.error[2])
+statistic_police_action_required = c(march2$statistic[2], shut2$statistic[2], destroy2$statistic[2])
+pval_police_action_required = c(march2$p.value[2], shut2$p.value[2], destroy2$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+
+term_recall_violence = c(march3$coefficients[2], shut3$coefficients[2], destroy3$coefficients[2])
+se_recall_violence = c(march3$std.error[2], shut3$std.error[2], destroy3$std.error[2])
+statistic_recall_violence = c(march3$statistic[2], shut3$statistic[2], destroy3$statistic[2])
+pval_recall_violence = c(march3$p.value[2], shut3$p.value[2], destroy3$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+
+differences_isr_black = rbind(degree_violence, police_action_required, recall_violence)
+
+term_degree_violence = c(march1$coefficients[3], shut1$coefficients[3], destroy1$coefficients[3])
+se_degree_violence = c(march1$std.error[3], shut1$std.error[3], destroy1$std.error[3])
+statistic_degree_violence = c(march1$statistic[3], shut1$statistic[3], destroy1$statistic[3])
+pval_degree_violence = c(march1$p.value[3], shut1$p.value[3], destroy1$p.value[3])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+term_police_action_required = c(march2$coefficients[3], shut2$coefficients[3], destroy2$coefficients[3])
+se_police_action_required = c(march2$std.error[3], shut2$std.error[3], destroy2$std.error[3])
+statistic_police_action_required = c(march2$statistic[3], shut2$statistic[3], destroy2$statistic[3])
+pval_police_action_required = c(march2$p.value[3], shut2$p.value[3], destroy2$p.value[3])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+
+term_recall_violence = c(march3$coefficients[3], shut3$coefficients[3], destroy3$coefficients[3])
+se_recall_violence = c(march3$std.error[3], shut3$std.error[3], destroy3$std.error[3])
+statistic_recall_violence = c(march3$statistic[3], shut3$statistic[3], destroy3$statistic[3])
+pval_recall_violence = c(march3$p.value[3], shut3$p.value[3], destroy3$p.value[3])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+
+differences_isr_arab = rbind(degree_violence, police_action_required, recall_violence)
+
+## Result tables for plot
+differences_us2 = differences_us
+differences_us2$submodel = differences_us2$term
+differences_us2$term = "Perception of Blacks \n(United States)"
+
+differences_isr_black2 = differences_isr_black
+differences_isr_black2$submodel = differences_isr_black2$term
+differences_isr_black2$term = "Perception of Ethiopians \n(Israel)"
+
+differences_isr_arab2 = differences_isr_arab
+differences_isr_arab2$submodel = differences_isr_arab2$term
+differences_isr_arab2$term = "Perception of Arabs \n(Israel)"
+
+diffs_sm = rbind(differences_us2, differences_isr_arab2, differences_isr_black2)
+rownames(diffs_sm) = 1:nrow(diffs_sm)
+
+results_df <- data.frame(term = diffs_sm$term,
+                         estimate = diffs_sm$estimate,
+                         std.error = diffs_sm$std.error,
+                         model = diffs_sm$model,
+                         submodel = as.character(diffs_sm$submodel),
+                         stringsAsFactors = FALSE)
+
+results_df$submodel[results_df$submodel=="1) Minority: March in streets"] = "March in streets"
+results_df$submodel[results_df$submodel=="2) Minority: Shut down traffic"] = "Shut down traffic"
+results_df$submodel[results_df$submodel=="3) Minority: Destroy police cars / garbage cans"] = "Destroy property"
+
+results_df$model[results_df$model=="Perceived degree of violence"] = "1. Perceived degree \nof violence"
+results_df$model[results_df$model=="Police action required"] = "3. Police action \nrequired"
+results_df$model[results_df$model=="Recall violence"] = "2. Recall \nviolence"
+
+small_multiple(results_df, dot_args = list(aes(shape = submodel))) + 
+  ylab("Coefficient Estimate (Std. Dev. Units)") + theme_bw()+
+  geom_hline(yintercept = 0, colour = "grey60", linetype = 2) +
+  #theme(plot.margin=unit(c(0.4,0.1,2.6,1),"cm"))+
+  theme(legend.position = "bottom", 
+        legend.justification=c(0, 0),
+        # legend.background = element_rect(color="white"),
+        legend.spacing = unit(-4, "pt"),
+        legend.key.size = unit(10, "pt"))+
+  scale_shape_discrete(name = "Protester Tactic") + 
+  scale_color_manual(values=c("#999999", "#E69F00", "#56B4E9"), name = "Protester Tactic")
+ggsave(filename="Figures/fig_3.pdf", width=6, height=6)
+
+table_a14 = results_df

23/replication_package/Code/replicate_fig_4.R

@@ -0,0 +1,94 @@
+##--##--##--##
+## Figure 4
+##--##--##--##
+
+march1 = lm(degree_violence ~ identity_protesters, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+march1_maj = lm(degree_violence ~ identity_protesters, data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race == "White",], weights=weight)
+march1_min = lm(degree_violence ~ identity_protesters, data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race == "Black",], weights=weight)
+
+march2 = lm(police_action_required ~ identity_protesters, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+march2_maj = lm(police_action_required ~ identity_protesters, data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race == "White",], weights=weight)
+march2_min = lm(police_action_required ~ identity_protesters, data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race == "Black",], weights=weight)
+
+march1_isr = lm(degree_violence ~ identity_protesters, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+march1_isr_maj = lm(degree_violence ~ identity_protesters, data=isr_survey_wave1[isr_survey_wave1$tactic==0 & isr_survey_wave1$ethnicity %in% c("Ashkenazi", "Mizrachi", "Mixed", "Soviet Union"),], weights=weight)
+march1_isr_min = lm(degree_violence ~ identity_protesters, data=isr_survey_wave1[isr_survey_wave1$tactic==0 & isr_survey_wave1$ethnicity == "Arab",], weights=weight)
+
+march2_isr = lm(police_action_required ~ identity_protesters, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+march2_isr_maj = lm(police_action_required ~ identity_protesters, data=isr_survey_wave1[isr_survey_wave1$tactic==0 & isr_survey_wave1$ethnicity %in% c("Ashkenazi", "Mizrachi", "Mixed", "Soviet Union"),], weights=weight)
+march2_isr_min = lm(police_action_required ~ identity_protesters, data=isr_survey_wave1[isr_survey_wave1$tactic==0 & isr_survey_wave1$ethnicity == "Arab",], weights=weight)
+
+# Predicted values
+perceptions_of_majority_us = c(ggpredict(march1_maj)$identity_protesters[,2],
+                               ggpredict(march2_maj)$identity_protesters[,2])
+
+perceptions_of_minority_us = c(ggpredict(march1_min)$identity_protesters[,2],
+                               ggpredict(march2_min)$identity_protesters[,2])
+
+perceptions_of_majority_isr = c(ggpredict(march1_isr_maj)$identity_protesters[c(1,3),2],
+                                ggpredict(march2_isr_maj)$identity_protesters[c(1,3),2])
+
+perceptions_of_minority_isr = c(ggpredict(march1_isr_min)$identity_protesters[c(1,3),2],
+                                ggpredict(march2_isr_min)$identity_protesters[c(1,3),2])
+
+# Std. errors
+
+se_of_majority_us = c(ggpredict(march1_maj)$identity_protesters[,3],
+                      ggpredict(march2_maj)$identity_protesters[,3])
+
+se_of_minority_us = c(ggpredict(march1_min)$identity_protesters[,3],
+                      ggpredict(march2_min)$identity_protesters[,3])
+
+se_of_majority_isr = c(ggpredict(march1_isr_maj)$identity_protesters[c(1,3),3],
+                       ggpredict(march2_isr_maj)$identity_protesters[c(1,3),3])
+
+se_of_minority_isr = c(ggpredict(march1_isr_min)$identity_protesters[c(1,3),3],
+                       ggpredict(march2_isr_min)$identity_protesters[c(1,3),3])
+
+
+majority_us = data.frame("value" = perceptions_of_majority_us,
+                         "se" = se_of_majority_us,
+                         "Protesters" = rep(c("Majority", "Minority"),2),
+                         "outcome" = c(rep("Perceived degree of violence", 2),
+                                       rep("Police action required", 2)))
+majority_us$respondents = "White \nrespondents"
+majority_us$survey = "U.S.A"
+
+minority_us = data.frame("value" = perceptions_of_minority_us,
+                         "se" = se_of_minority_us,
+                         "Protesters" = rep(c("Majority", "Minority"),2),
+                         "outcome" = c(rep("Perceived degree of violence", 2),
+                                       rep("Police action required", 2)))
+minority_us$respondents = "Black \nrespondents"
+minority_us$survey = "U.S.A"
+
+majority_isr = data.frame("value" = perceptions_of_majority_isr,
+                          "se" = se_of_majority_isr,
+                          "Protesters" = rep(c("Majority", "Minority"),2),
+                          "outcome" = c(rep("Perceived degree of violence", 2),
+                                        rep("Police action required", 2)))
+majority_isr$respondents = "White Jewish \nrespondents"
+majority_isr$survey = "Israel"
+
+minority_isr = data.frame("value" = perceptions_of_minority_isr,
+                          "se" = se_of_minority_isr,
+                          "Protesters" = rep(c("Majority", "Minority"),2),
+                          "outcome" = c(rep("Perceived degree of violence", 2),
+                                        rep("Police action required", 2)))
+minority_isr$respondents = "Arab \nrespondents"
+minority_isr$survey = "Israel"
+
+
+df_all = rbind(majority_us, minority_us, majority_isr, minority_isr)
+df_all = df_all[df_all$outcome != "Recall violence",]
+df_all$respondents = factor(df_all$respondents, levels=c("White \nrespondents", "White Jewish \nrespondents", "Black \nrespondents", "Arab \nrespondents"))
+
+ggplot(df_all, aes(x = respondents, y = value, group=Protesters)) +
+  geom_point(aes(color=Protesters, shape=Protesters), position = position_dodge(0.3))+
+  geom_errorbar(aes(ymin=value-se, ymax=value+se, color=Protesters), width=0,
+                position=position_dodge(0.3))+
+  facet_wrap(~outcome + survey, nrow=2, scales = "free") + theme_bw() +
+  scale_shape_discrete(name = "Protesters") + 
+  theme(legend.position = "bottom") + xlab("") + ylab("") +
+  scale_color_manual(values=c("#E69F00", "#56B4E9"))
+ggsave("Figures/fig_4.pdf", width=6, height=6)

23/replication_package/Code/replicate_fig_6.R

@@ -0,0 +1,70 @@
+##--##--##--##
+## Figure 6
+##--##--##--##
+
+goal_white_usa_pdv = lm_robust(degree_violence_std ~ group_goal, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="White",])
+goal_black_usa_pdv = lm_robust(degree_violence_std ~ group_goal, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="Black",])
+goal_white_usa_rv = lm_robust(recall_violence_std ~ group_goal, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="White",])
+goal_black_usa_rv = lm_robust(recall_violence_std ~ group_goal, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="Black",])
+goal_white_usa_paq = lm_robust(police_action_required_std ~ group_goal, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="White",])
+goal_black_usa_paq = lm_robust(police_action_required_std ~ group_goal, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="Black",])
+
+## Plot interaction of goal and identity
+term_degree_violence = c(goal_white_usa_pdv$coefficients[2], goal_black_usa_pdv$coefficients[2])
+se_degree_violence = c(goal_white_usa_pdv$std.error[2], goal_black_usa_pdv$std.error[2])
+statistic_degree_violence = c(goal_white_usa_pdv$statistic[2], goal_black_usa_pdv$statistic[2])
+pval_degree_violence = c(goal_white_usa_pdv$p.value[2], goal_black_usa_pdv$p.value[2])
+term = c("Majority group protesters", "Minority group protesters")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree \nof violence"
+
+term_recall_violence = c(goal_white_usa_rv$coefficients[2], goal_black_usa_rv$coefficients[2])
+se_recall_violence = c(goal_white_usa_rv$std.error[2], goal_black_usa_rv$std.error[2])
+statistic_recall_violence = c(goal_white_usa_rv$statistic[2], goal_black_usa_rv$statistic[2])
+pval_recall_violence = c(goal_white_usa_rv$p.value[2], goal_black_usa_rv$p.value[2])
+term = c("Majority group protesters", "Minority group protesters")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall \nviolence"
+
+term_police_action_required = c(goal_white_usa_paq$coefficients[2], goal_black_usa_paq$coefficients[2])
+se_police_action_required = c(goal_white_usa_paq$std.error[2], goal_black_usa_paq$std.error[2])
+statistic_police_action_required = c(goal_white_usa_paq$statistic[2], goal_black_usa_paq$statistic[2])
+pval_police_action_required = c(goal_white_usa_paq$p.value[2], goal_black_usa_paq$p.value[2])
+term = c("Majority group protesters", "Minority group protesters")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action \nrequired"
+
+differences_us = rbind(degree_violence, recall_violence, police_action_required)
+
+## Result tables for plot
+
+differences_us2 = differences_us
+differences_us2$submodel = differences_us2$term
+differences_us2$term = "Protesting against police brutality"
+diffs_sm = differences_us2
+results_df = data.frame(term = diffs_sm$term,
+                        estimate = diffs_sm$estimate,
+                        std.error = diffs_sm$std.error,
+                        model = diffs_sm$model,
+                        submodel = as.character(diffs_sm$submodel),
+                        stringsAsFactors = FALSE)
+results_df$term = factor(results_df$model, levels=c("Perceived degree \nof violence",
+                                                    "Recall \nviolence",
+                                                    "Police action \nrequired",
+                                                    "Does not support \nprotest"))
+results_df$model[results_df$submodel=="Majority group protesters"] = "White protesters"
+results_df$model[results_df$submodel=="Minority group protesters"] = "Black protesters"
+
+dwplot(results_df, vline = geom_vline(xintercept = 0, colour = "grey60", linetype = 2), dot_args = list(size = 2.5)) + 
+  theme_bw() + xlab("Coefficient Estimate (Std. Dev. Units)") + theme(legend.position = "bottom", axis.text=element_text(size=9), legend.title = element_blank()) +
+  scale_colour_grey(start = .1, end = .7) 
+ggsave(filename="Figures/fig_6.pdf", width=5, height=4)

23/replication_package/Code/replicate_fig_7.R

@@ -0,0 +1,76 @@
+##--##--##--##
+## Figure 7
+##--##--##--##
+
+commitment_white_usa_pdv = lm_robust(degree_violence_std ~ commitment, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="White",])
+commitment_black_usa_pdv = lm_robust(degree_violence_std ~ commitment, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="Black",])
+
+commitment_white_israel_pdv = lm_robust(degree_violence_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==0,])
+commitment_ethiopian_israel_pdv = lm_robust(degree_violence_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==1,])
+commitment_arab_israel_pdv = lm_robust(degree_violence_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==2,])
+
+term_degree_violence = c(commitment_white_usa_pdv$coefficients[2], commitment_black_usa_pdv$coefficients[2])
+se_degree_violence = c(commitment_white_usa_pdv$std.error[2], commitment_black_usa_pdv$std.error[2])
+statistic_degree_violence = c(commitment_white_usa_pdv$statistic[2], commitment_black_usa_pdv$statistic[2])
+pval_degree_violence = c(commitment_white_usa_pdv$p.value[2], commitment_black_usa_pdv$p.value[2])
+term = c("Majority group", "Minority group")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+differences_us = rbind(degree_violence)
+
+term_degree_violence = c(commitment_white_israel_pdv$coefficients[2], commitment_arab_israel_pdv$coefficients[2])
+se_degree_violence = c(commitment_white_israel_pdv$std.error[2], commitment_arab_israel_pdv$std.error[2])
+statistic_degree_violence = c(commitment_white_israel_pdv$statistic[2], commitment_arab_israel_pdv$statistic[2])
+pval_degree_violence = c(commitment_white_israel_pdv$p.value[2], commitment_arab_israel_pdv$p.value[2])
+term = c("Majority group", "Minority group")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+differences_isr_arab = degree_violence
+
+term_degree_violence = c(commitment_white_israel_pdv$coefficients[2], commitment_ethiopian_israel_pdv$coefficients[2])
+se_degree_violence = c(commitment_white_israel_pdv$std.error[2], commitment_ethiopian_israel_pdv$std.error[2])
+statistic_degree_violence = c(commitment_white_israel_pdv$statistic[2], commitment_ethiopian_israel_pdv$statistic[2])
+pval_degree_violence = c(commitment_white_israel_pdv$p.value[2], commitment_ethiopian_israel_pdv$p.value[2])
+term = c("Majority group", "Minority group")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+differences_isr_ethiopian = degree_violence
+
+## Result tables for plot
+differences_us2 = differences_us
+differences_us2$submodel = differences_us2$term
+differences_us2$term = "Perception of Blacks \n(United States)"
+
+differences_isr_black2 = differences_isr_ethiopian
+differences_isr_black2$submodel = differences_isr_black2$term
+differences_isr_black2$term = "Perception of Ethiopians \n(Israel)"
+
+differences_isr_arab2 = differences_isr_arab
+differences_isr_arab2$submodel = differences_isr_arab2$term
+differences_isr_arab2$term = "Perception of Arabs \n(Israel)"
+
+diffs_sm = rbind(differences_us2, differences_isr_black2, differences_isr_arab2)
+rownames(diffs_sm) = 1:nrow(diffs_sm)
+
+results_df <- data.frame(term = diffs_sm$term,
+                         estimate = diffs_sm$estimate,
+                         std.error = diffs_sm$std.error,
+                         model =  as.character(diffs_sm$submodel),
+                         stringsAsFactors = FALSE)
+
+dwplot(results_df, vline = geom_vline(xintercept = 0, colour = "grey60", linetype = 2), dot_args = list(size = 2.5)) + 
+  theme_bw() + xlab("Coefficient Estimate (Std. Dev. Units)") + theme(legend.position = "bottom", axis.text=element_text(size=9), legend.title = element_blank()) +
+  scale_colour_grey(start = .1, end = .7) +xlim(-0.35,0.1)
+ggsave(filename="Figures/fig_7.pdf", width=5.5, height=4)

23/replication_package/Code/replicate_fig_8.R

@@ -0,0 +1,36 @@
+##--##--##--##
+## Figure 8
+##--##--##--##
+
+us_survey_content_covars <- estimateEffect(1:10 ~ identity_protesters, us_survey_text_analysis$stm_topics, meta = docvars(us_survey_text_analysis$nv_dfm), uncertainty = "Global")
+us_survey_sum_content_covars = summary(us_survey_content_covars)
+
+topics = labelTopics(us_survey_text_analysis$stm_topics, n=20)
+
+term = c(paste(topics$prob[1,1:5], collapse = ", "), 
+         paste(topics$prob[2,c(1,2,7,8,15)], collapse = ", "),
+         paste(topics$prob[3,1:5], collapse = ", "),
+         paste(topics$prob[4,4:9], collapse = ", "),
+         paste(topics$prob[5,c(1,2,3,5,15)], collapse = ", "),
+         paste(topics$prob[6,1:5], collapse = ", "),
+         paste(topics$prob[7,1:5], collapse = ", "),
+         paste(topics$prob[8,1:5], collapse = ", "),
+         paste(topics$prob[9,c(3,4,9,15,17)], collapse = ", "),
+         paste(topics$prob[10,c(8,12,13,15,16)], collapse = ", "))
+
+estimate = rep(NA, 10)
+std.error = rep(NA, 10)
+for(i in 1:length(us_survey_sum_content_covars$tables)){
+  estimate[i] = us_survey_sum_content_covars$tables[[i]][2,1]
+  std.error[i] = us_survey_sum_content_covars$tables[[i]][2,2]
+}
+
+us_survey_stm_results = data.frame("term" = term, "estimate" = estimate, "std.error"= std.error)
+us_survey_stm_results = us_survey_stm_results[order(us_survey_stm_results$estimate, decreasing = T),]
+
+dwplot(us_survey_stm_results, 
+       vline = geom_vline(xintercept = 0, colour = "grey60", linetype = 2), dot_args = list(size = 2.5)) +
+  theme_bw() + xlab("\nWhite protesters . . . . . . . . . . . . . . . . . . Black protesters") +
+  theme(text = element_text(size=12), axis.text.x = element_text(angle = 0, hjust = 1), legend.position = "none") +
+  scale_colour_grey(start = 0, end = 0)
+ggsave(file="Figures/fig_8.pdf", width=8, height=6)

23/replication_package/Code/replicate_fig_9.R

@@ -0,0 +1,35 @@
+##--##--##--##
+## Figure 9
+##--##--##--##
+
+isr_survey_arab_content_covars <- estimateEffect(1:10 ~ identity_protesters_arab, isr_survey_text_analysis_arab$stm_topics, meta = docvars(isr_survey_text_analysis_arab$nv_dfm_arab), uncertainty = "Global")
+isr_survey_arab_sum_content_covars = summary(isr_survey_arab_content_covars)
+
+term_translated_arab = c("violent, protest, israel",
+                         "justified, right, protest",
+                         "violent, express, equal, demands",
+                         "democracy, protest, rights",
+                         "legitimate, support, sympathize, voice",
+                         "difficult, situation, understand, pain",
+                         "violent, protest, against",
+                         "social, justice, protest",
+                         "information, details, reason",
+                         "violence, arab, community, minority")
+
+estimate = rep(NA, 10)
+std.error = rep(NA, 10)
+
+for(i in 1:length(isr_survey_arab_sum_content_covars$tables)){
+  estimate[i] = isr_survey_arab_sum_content_covars$tables[[i]][2,1]
+  std.error[i] = isr_survey_arab_sum_content_covars$tables[[i]][2,2]
+}
+
+isr_survey_arab_stm_results = data.frame("term" = term_translated_arab, "estimate" = estimate, "std.error"= std.error)
+isr_survey_arab_stm_results = isr_survey_arab_stm_results[order(isr_survey_arab_stm_results$estimate, decreasing = T),]
+
+dwplot(isr_survey_arab_stm_results, 
+       vline = geom_vline(xintercept = 0, colour = "grey60", linetype = 2), dot_args = list(size = 2.5)) +
+  theme_bw() + xlab("\nIsraeli white protesters . . . . . . . . . . . . . . . . . . Israeli Arab protesters") +
+  theme(text = element_text(size=12), axis.text.x = element_text(angle = 0, hjust = 1), legend.position = "none") +
+  scale_colour_grey(start = 0, end = 0)
+ggsave(file="Figures/fig_9.pdf", width=8, height=6)

23/replication_package/Code/replicate_fig_A1.R

@@ -0,0 +1,25 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A1
+##--##--##--##--##--##--##
+
+EBCR_EPR_NAVCO2$camp_goals_text = NA
+EBCR_EPR_NAVCO2$camp_goals_text[EBCR_EPR_NAVCO2$camp_goals==0] = "Regime change"
+EBCR_EPR_NAVCO2$camp_goals_text[EBCR_EPR_NAVCO2$camp_goals==1] = "Significant institutional reform"
+EBCR_EPR_NAVCO2$camp_goals_text[EBCR_EPR_NAVCO2$camp_goals==2] = "Policy change"
+EBCR_EPR_NAVCO2$camp_goals_text[EBCR_EPR_NAVCO2$camp_goals==3] = "Territorial secession"
+EBCR_EPR_NAVCO2$camp_goals_text[EBCR_EPR_NAVCO2$camp_goals==4] = "Greater autonomy"
+EBCR_EPR_NAVCO2$camp_goals_text[EBCR_EPR_NAVCO2$camp_goals==5] = "Anti occupation"
+
+EBCR_EPR_NAVCO2$EPR_STATUS = factor(EBCR_EPR_NAVCO2$EPR_STATUS, 
+                                    levels=c("DISCRIMINATED", "POWERLESS", "SELF-EXCLUSION", 
+                                             "JUNIOR PARTNER", "SENIOR PARTNER", "DOMINANT", 
+                                             "MONOPOLY", "IRRELEVANT", "STATE COLLAPSE"))
+
+goals_status = prop.table(table(EBCR_EPR_NAVCO2$camp_goals_text[EBCR_EPR_NAVCO2$navco1designation==1],  EBCR_EPR_NAVCO2$EPR_STATUS[EBCR_EPR_NAVCO2$navco1designation==1]))
+
+col6 = c("#9ECAE1", "#6BAED6", "#4292C6", "#2171B5", "#08519C", "#08306B")
+
+pdf(file="Figures/fig_A1.pdf", width=8, height=6)
+corrplot(goals_status[,c("DISCRIMINATED", "POWERLESS", "SELF-EXCLUSION",
+                         "JUNIOR PARTNER", "SENIOR PARTNER", "DOMINANT", "MONOPOLY")], method = "circle", cl.lim = c(0,0.15), is.corr = F, col=col6, tl.col="black")
+dev.off()

23/replication_package/Code/replicate_fig_A10.R

@@ -0,0 +1,68 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A10
+##--##--##--##--##--##--##
+
+commitment_white_usa_rv = lm_robust(recall_violence_std ~ commitment, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="White",])
+commitment_black_usa_rv = lm_robust(recall_violence_std ~ commitment, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="Black",])
+commitment_white_israel_rv = lm_robust(recall_violence_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==0,])
+commitment_ethiopian_israel_rv = lm_robust(recall_violence_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==1,])
+commitment_arab_israel_rv = lm_robust(recall_violence_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==2,])
+
+## U.S. 
+term_recall_violence = c(commitment_white_usa_rv$coefficients[2], commitment_black_usa_rv$coefficients[2])
+se_recall_violence = c(commitment_white_usa_rv$std.error[2], commitment_black_usa_rv$std.error[2])
+statistic_recall_violence = c(commitment_white_usa_rv$statistic[2], commitment_black_usa_rv$statistic[2])
+pval_recall_violence = c(commitment_white_usa_rv$p.value[2], commitment_black_usa_rv$p.value[2])
+term = c("Majority group", "Minority group")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+differences_us = recall_violence
+
+## Israel (Arabs)
+term_recall_violence = c(commitment_white_israel_rv$coefficients[2], commitment_arab_israel_rv$coefficients[2])
+se_recall_violence = c(commitment_white_israel_rv$std.error[2], commitment_arab_israel_rv$std.error[2])
+statistic_recall_violence = c(commitment_white_israel_rv$statistic[2], commitment_arab_israel_rv$statistic[2])
+pval_recall_violence = c(commitment_white_israel_rv$p.value[2], commitment_arab_israel_rv$p.value[2])
+term = c("Majority group", "Minority group")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+differences_isr_arab = recall_violence
+
+## Israel (Ethiopian)
+term_recall_violence = c(commitment_white_israel_rv$coefficients[2], commitment_ethiopian_israel_rv$coefficients[2])
+se_recall_violence = c(commitment_white_israel_rv$std.error[2], commitment_ethiopian_israel_rv$std.error[2])
+statistic_recall_violence = c(commitment_white_israel_rv$statistic[2], commitment_ethiopian_israel_rv$statistic[2])
+pval_recall_violence = c(commitment_white_israel_rv$p.value[2], commitment_ethiopian_israel_rv$p.value[2])
+term = c("Majority group", "Minority group")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+differences_isr_ethiopian = recall_violence
+
+## Result tables for plot
+differences_us$term = "Perception of Blacks \n(United States)"
+differences_isr_ethiopian$term = "Perception of Ethiopians \n(Israel)"
+differences_isr_arab$term = "Perception of Arabs \n(Israel)"
+
+diffs_sm = rbind(differences_us, differences_isr_ethiopian, differences_isr_arab)
+diffs_sm$model = rep(c("Majority group", "Minority group"),3)
+rownames(diffs_sm) = 1:nrow(diffs_sm)
+
+results_df <- data.frame(term = diffs_sm$term,
+                         estimate = diffs_sm$estimate,
+                         std.error = diffs_sm$std.error,
+                         model = diffs_sm$model,
+                         stringsAsFactors = FALSE)
+
+dwplot(results_df,vline = geom_vline(xintercept = 0, colour = "grey60", linetype = 2), dot_args = list(size = 3)) + 
+  theme_bw() + xlab("Coefficient Estimate (Std. Dev. Units)") + theme(legend.position = "bottom", axis.text=element_text(size=9), legend.title = element_blank()) +
+  scale_colour_grey(start = .1, end = .7) 
+ggsave(filename="Figures/fig_A10.pdf", width=6, height=5)

23/replication_package/Code/replicate_fig_A11.R

@@ -0,0 +1,68 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A11
+##--##--##--##--##--##--##
+
+commitment_white_usa_paq = lm_robust(police_action_required_std ~ commitment, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="White",])
+commitment_black_usa_paq = lm_robust(police_action_required_std ~ commitment, data=us_survey_wave2[us_survey_wave2$identity_protesters_fac=="Black",])
+commitment_white_israel_paq = lm_robust(police_action_required_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==0,])
+commitment_ethiopian_israel_paq = lm_robust(police_action_required_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==1,])
+commitment_arab_israel_paq = lm_robust(police_action_required_std ~ commitment, data=isr_survey_wave2[isr_survey_wave2$identity_protesters==2,])
+
+## U.S. 
+term_police_action_required = c(commitment_white_usa_paq$coefficients[2], commitment_black_usa_paq$coefficients[2])
+se_police_action_required = c(commitment_white_usa_paq$std.error[2], commitment_black_usa_paq$std.error[2])
+statistic_police_action_required = c(commitment_white_usa_paq$statistic[2], commitment_black_usa_paq$statistic[2])
+pval_police_action_required = c(commitment_white_usa_paq$p.value[2], commitment_black_usa_paq$p.value[2])
+term = c("Majority group", "Minority group")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+differences_us = police_action_required
+
+## Israel (Arabs)
+term_police_action_required = c(commitment_white_israel_paq$coefficients[2], commitment_arab_israel_paq$coefficients[2])
+se_police_action_required = c(commitment_white_israel_paq$std.error[2], commitment_arab_israel_paq$std.error[2])
+statistic_police_action_required = c(commitment_white_israel_paq$statistic[2], commitment_arab_israel_paq$statistic[2])
+pval_police_action_required = c(commitment_white_israel_paq$p.value[2], commitment_arab_israel_paq$p.value[2])
+term = c("Majority group", "Minority group")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+differences_isr_arab = police_action_required
+
+## Israel (Ethiopian)
+term_police_action_required = c(commitment_white_israel_paq$coefficients[2], commitment_ethiopian_israel_paq$coefficients[2])
+se_police_action_required = c(commitment_white_israel_paq$std.error[2], commitment_ethiopian_israel_paq$std.error[2])
+statistic_police_action_required = c(commitment_white_israel_paq$statistic[2], commitment_ethiopian_israel_paq$statistic[2])
+pval_police_action_required = c(commitment_white_israel_paq$p.value[2], commitment_ethiopian_israel_paq$p.value[2])
+term = c("Majority group", "Minority group")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+differences_isr_ethiopian = police_action_required
+
+## Result tables for plot
+differences_us$term = "Perception of Blacks \n(United States)"
+differences_isr_ethiopian$term = "Perception of Ethiopians \n(Israel)"
+differences_isr_arab$term = "Perception of Arabs \n(Israel)"
+
+diffs_sm = rbind(differences_us, differences_isr_ethiopian, differences_isr_arab)
+diffs_sm$model = rep(c("Majority group", "Minority group"),3)
+rownames(diffs_sm) = 1:nrow(diffs_sm)
+
+results_df <- data.frame(term = diffs_sm$term,
+                         estimate = diffs_sm$estimate,
+                         std.error = diffs_sm$std.error,
+                         model = diffs_sm$model,
+                         stringsAsFactors = FALSE)
+
+dwplot(results_df,vline = geom_vline(xintercept = 0, colour = "grey60", linetype = 2), dot_args = list(size = 3)) + 
+  theme_bw() + xlab("Coefficient Estimate (Std. Dev. Units)") + theme(legend.position = "bottom", axis.text=element_text(size=9), legend.title = element_blank()) +
+  scale_colour_grey(start = .1, end = .7) 
+ggsave(filename="Figures/fig_A11.pdf", width=6, height=5)

23/replication_package/Code/replicate_fig_A12.R

@@ -0,0 +1,50 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A12
+##--##--##--##--##--##--##
+
+## U.S.
+us_survey_wave1_dv = lm(degree_violence ~  identity_protesters*interest_news +age +female + education + income  + ideology + race, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)  
+us_survey_wave1_dv_all = ggpredict(us_survey_wave1_dv, terms = c("interest_news"))
+us_survey_wave1_dv_all$interest_news = c("Hardly at all", "Only now \nand then", "Some of \nthe time", "Most of \nthe time")
+us_survey_wave1_dv_all$interest_news = factor(us_survey_wave1_dv_all$interest_news, levels=c("Hardly at all", "Only now \nand then", "Some of \nthe time", "Most of \nthe time"))
+us_survey_wave1_dv_all = as.data.frame(us_survey_wave1_dv_all)
+
+news_usa_study1 = ggplot(us_survey_wave1_dv_all, aes(x = interest_news, y = predicted, group="all")) +
+  geom_line()+ geom_errorbar(width = 0, ymin=us_survey_wave1_dv_all$conf.low, ymax=us_survey_wave1_dv_all$conf.high)+
+  geom_point(size=3) + ylim(-0.05,10) +  theme_bw() +
+  scale_colour_grey(start = 0.1, end = 0.65) + geom_hline(yintercept=0, linetype="dashed")+
+  ylab("Perceived degree violence") + xlab("\nHow often do you follow the news?") + ggtitle("Perceptions of Black Protesters in the U.S.")
+
+## Israel
+isr_survey_wave1$interest_news2 = isr_survey_wave1$interest_news-1
+
+isr_survey_dv_eth = lm(degree_violence ~  identity_protesters*interest_news2 + age +female + education + income  + ideology + ethnicity, data=isr_survey_wave1[isr_survey_wave1$identity_protesters %in% c(0,1) & isr_survey_wave1$tactic==0,], weights=weight)  
+isr_survey_dv_all_eth = ggpredict(isr_survey_dv_eth, terms = c("interest_news2"))
+isr_survey_dv_all_eth$interest_news = c("Hardly at all", "Only now \nand then", "Some of \nthe time", "Most of \nthe time")
+isr_survey_dv_all_eth$interest_news = factor(isr_survey_dv_all_eth$interest_news, levels=c("Hardly at all", "Only now \nand then", "Some of \nthe time", "Most of \nthe time"))
+isr_survey_dv_all_eth = as.data.frame(isr_survey_dv_all_eth)
+isr_survey_dv_all_eth$Minority = "Ethiopians"
+
+isr_survey_dv_arab = lm(degree_violence ~  identity_protesters*interest_news2 + age +female + education + income  + ideology + ethnicity, data=isr_survey_wave1[isr_survey_wave1$identity_protesters %in% c(0,2) & isr_survey_wave1$tactic==0,], weights=weight)  
+isr_survey_dv_all_arab = ggpredict(isr_survey_dv_arab, terms = c("interest_news2"))
+isr_survey_dv_all_arab$interest_news = c("Hardly at all", "Only now \nand then", "Some of \nthe time", "Most of \nthe time")
+isr_survey_dv_all_arab$interest_news = factor(isr_survey_dv_all_arab$interest_news, levels=c("Hardly at all", "Only now \nand then", "Some of \nthe time", "Most of \nthe time"))
+isr_survey_dv_all_arab = as.data.frame(isr_survey_dv_all_arab)
+isr_survey_dv_all_arab$Minority = "Arabs"
+
+news_israel_study1_eth = ggplot(isr_survey_dv_all_eth, aes(x = interest_news, y = predicted, group="all")) +
+  geom_line()+ geom_errorbar(width = 0, ymin=isr_survey_dv_all_eth$conf.low, ymax=isr_survey_dv_all_eth$conf.high)+
+  geom_point(size=3) + ylim(-0.05,10) +  theme_bw() +
+  scale_colour_grey(start = 0.1, end = 0.65) + geom_hline(yintercept=0, linetype="dashed")+
+  ylab("Perceived degree violence") + xlab("\nHow often do you follow the news?") + ggtitle("Perception of Ethiopian Protesters in Israel")
+
+news_israel_study1_arab = ggplot(isr_survey_dv_all_arab, aes(x = interest_news, y = predicted, group="all")) +
+  geom_line()+ geom_errorbar(width = 0, ymin=isr_survey_dv_all_arab$conf.low, ymax=isr_survey_dv_all_arab$conf.high)+
+  geom_point(size=3) + ylim(-0.05,10) +  theme_bw() +
+  scale_colour_grey(start = 0.1, end = 0.65) + geom_hline(yintercept=0, linetype="dashed")+
+  ylab("Perceived degree violence") + xlab("\nHow often do you follow the news?") + ggtitle("Perception of Arab Protesters in Israel")
+
+
+grid_study1 = grid.arrange(news_usa_study1, news_israel_study1_eth, news_israel_study1_arab, nrow = 1)
+ggsave(filename= "Figures/fig_A12.pdf", plot=grid_study1, width=13, height=4, units="in")
+

23/replication_package/Code/replicate_fig_A2.R

@@ -0,0 +1,36 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A2
+##--##--##--##--##--##--##
+
+mean_sucess_anti_occupation_minority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Anti occupation") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==1 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+mean_sucess_anti_occupation_majority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Anti occupation") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==0 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+
+mean_sucess_greater_autonomy_minority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Greater autonomy") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==1 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+mean_sucess_greater_autonomy_majority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Greater autonomy") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==0 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+
+mean_sucess_policy_change_minority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Policy change") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==1 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+mean_sucess_policy_change_majority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Policy change") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==0 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+
+mean_sucess_regime_change_minority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Regime change") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==1 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+mean_sucess_regime_change_majority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Regime change") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==0 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+
+mean_sucess_inst_reform_minority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Significant institutional reform") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==1 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+mean_sucess_inst_reform_majority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Significant institutional reform") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==0 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+
+mean_sucess_territorial_secession_minority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Territorial secession") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==1 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+mean_sucess_territorial_secession_majority = mean(EBCR_EPR_NAVCO2$success[EBCR_EPR_NAVCO2$camp_goals_text %in% c("Territorial secession") & EBCR_EPR_NAVCO2$EPR_STATUS_EXCL==0 & EBCR_EPR_NAVCO2$navco1designation==1], na.rm=T)
+
+success_props = data.frame("prop_success" = c(mean_sucess_anti_occupation_minority, mean_sucess_anti_occupation_majority,
+                                              mean_sucess_greater_autonomy_minority, mean_sucess_greater_autonomy_majority,
+                                              mean_sucess_policy_change_minority, mean_sucess_policy_change_majority,
+                                              mean_sucess_regime_change_minority, mean_sucess_regime_change_majority,
+                                              mean_sucess_inst_reform_minority, mean_sucess_inst_reform_majority,
+                                              mean_sucess_territorial_secession_minority, mean_sucess_territorial_secession_majority),
+                           "Status" =rep(c("Excluded", "Included"), 6), "goal" = c(rep("Anti \noccupation", 2), rep("Greater \nautonomy", 2), rep("Policy \nchange", 2),
+                                                                                   rep("Regime \nchange", 2), rep("Significant \ninstitutional \nreform", 2), rep("Territorial \nsecession", 2)))
+
+
+ggplot(data=success_props, aes(x=goal, y=prop_success, fill=Status)) +
+  geom_bar(stat="identity", position=position_dodge()) + theme_bw() + 
+  ylab("Proportion of successful campaigns") + xlab("Campaign goal") + scale_fill_brewer(palette="Paired")
+ggsave(file="Figures/fig_A2.pdf", width=6.5, height=4)

23/replication_package/Code/replicate_fig_A3.R

@@ -0,0 +1,11 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A3
+##--##--##--##--##--##--##
+
+status_names = c("DISCRIMINATED", "POWERLESS", "SELF-EXCLUSION", "JUNIOR PARTNER", "SENIOR PARTNER", "DOMINANT", "MONOPOLY")
+
+ggplot(EBCR_EPR_NAVCO2, aes(x=EPR_GROUPSIZE, y=EPR_STATUS_ORD)) + 
+  geom_point(alpha=0.2, color="gray50") + geom_smooth(color="black") +
+  theme_bw() + xlab("Group size") + ylab("Group status") + 
+  scale_y_continuous(breaks = c(1,2,3,4,5,6,7), labels = status_names) 
+ggsave("Figures/fig_A3.pdf", width=6, height=5)

23/replication_package/Code/replicate_fig_A6.R

@@ -0,0 +1,110 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A6
+##--##--##--##--##--##--##
+
+us_survey_wave1$education_numeric = as.numeric(us_survey_wave1$education)
+us_survey_wave1$income_numeric = as.numeric(us_survey_wave1$income)
+us_survey_wave1$ideology_numeric = as.numeric(us_survey_wave1$ideology)
+
+us_survey_wave1 = dummy_cols(us_survey_wave1, select_columns = c("partyID", "race"))
+vars_us_survey_wave1 = c("age", "female", "education_numeric", "income_numeric", "partyID_D", "partyID_I", "partyID_R", "ideology_numeric", "race_White", "race_Black")
+
+means_black_march = means_white_march = means_black_shut = means_white_shut = means_black_destroy = means_white_destroy =  rep(NA, length(vars_us_survey_wave1))
+se_black_march = se_white_march = se_black_shut = se_white_shut = se_black_destroy = se_white_destroy =  rep(NA, length(vars_us_survey_wave1))
+
+for(i in 1:length(vars_us_survey_wave1)){
+  means_black_march[i] = mean(us_survey_wave1[us_survey_wave1$identity_protesters==1 & us_survey_wave1$tactic==0, vars_us_survey_wave1[i]], na.rm=T)
+  means_white_march[i] = mean(us_survey_wave1[us_survey_wave1$identity_protesters==0 & us_survey_wave1$tactic==0, vars_us_survey_wave1[i]], na.rm=T)
+  means_black_shut[i] = mean(us_survey_wave1[us_survey_wave1$identity_protesters==1 & us_survey_wave1$tactic==1, vars_us_survey_wave1[i]], na.rm=T)
+  means_white_shut[i] = mean(us_survey_wave1[us_survey_wave1$identity_protesters==0 & us_survey_wave1$tactic==1, vars_us_survey_wave1[i]], na.rm=T)
+  means_black_destroy[i] = mean(us_survey_wave1[us_survey_wave1$identity_protesters==1 & us_survey_wave1$tactic==2, vars_us_survey_wave1[i]], na.rm=T)
+  means_white_destroy[i] = mean(us_survey_wave1[us_survey_wave1$identity_protesters==0 & us_survey_wave1$tactic==2, vars_us_survey_wave1[i]], na.rm=T)
+  
+  se_black_march[i] = sd(us_survey_wave1[us_survey_wave1$identity_protesters==1 & us_survey_wave1$tactic==0, vars_us_survey_wave1[i]], na.rm=T)
+  se_white_march[i] = sd(us_survey_wave1[us_survey_wave1$identity_protesters==0 & us_survey_wave1$tactic==0, vars_us_survey_wave1[i]], na.rm=T)
+  se_black_shut[i] = sd(us_survey_wave1[us_survey_wave1$identity_protesters==1 & us_survey_wave1$tactic==1, vars_us_survey_wave1[i]], na.rm=T)
+  se_white_shut[i] = sd(us_survey_wave1[us_survey_wave1$identity_protesters==0 & us_survey_wave1$tactic==1, vars_us_survey_wave1[i]], na.rm=T)
+  se_black_destroy[i] = sd(us_survey_wave1[us_survey_wave1$identity_protesters==1 & us_survey_wave1$tactic==2, vars_us_survey_wave1[i]], na.rm=T)
+  se_white_destroy[i] = sd(us_survey_wave1[us_survey_wave1$identity_protesters==0 & us_survey_wave1$tactic==2, vars_us_survey_wave1[i]], na.rm=T)
+  
+}
+
+us_survey_wave1_balance = cbind(means_black_march, se_black_march,
+                                means_white_march, se_white_march, 
+                                means_black_shut, se_black_shut,
+                                means_white_shut, se_white_shut,
+                                means_black_destroy, se_black_destroy,
+                                means_white_destroy, se_white_destroy)
+
+us_survey_wave1_balance = as.data.frame(us_survey_wave1_balance)
+colnames(us_survey_wave1_balance) = c("Blacks, march in streets (mean)", "Blacks, march in streets (sd)",
+                                      "Whites, march in streets (mean)", "Whites, march in streets (sd)",
+                                      "Blacks, shut down traffic (mean)", "Blacks, shut down traffic (sd)",
+                                      "Whites, shut down traffic (mean)", "Whites, shut down traffic (sd)",
+                                      "Blacks, destroy police cars (mean)", "Blacks, destroy police cars (sd)",
+                                      "Whites, destroy police cars (mean)", "Whites, destroy police cars (sd)")
+rownames(us_survey_wave1_balance) = vars_us_survey_wave1
+
+
+us_survey_wave1_balance_means = as.data.frame(cbind(means_black_march,
+                                                    means_white_march, 
+                                                    means_black_shut,
+                                                    means_white_shut,
+                                                    means_black_destroy,
+                                                    means_white_destroy))
+
+rownames(us_survey_wave1_balance_means) = vars_us_survey_wave1
+us_survey_wave1_balance_means$var = vars_us_survey_wave1
+
+keycol <- "condition"
+valuecol <- "mean"
+gathercols <- c("means_black_march", "means_white_march", "means_black_shut", "means_white_shut",
+                "means_black_destroy", "means_white_destroy")
+
+us_survey_wave1_balance_means_long = gather_(us_survey_wave1_balance_means, keycol, valuecol, gathercols)
+
+
+us_survey_wave1_balance_se = as.data.frame(cbind(se_black_march,
+                                                 se_white_march, 
+                                                 se_black_shut,
+                                                 se_white_shut,
+                                                 se_black_destroy,
+                                                 se_white_destroy))
+
+rownames(us_survey_wave1_balance_se) = vars_us_survey_wave1
+us_survey_wave1_balance_se$var = vars_us_survey_wave1
+
+keycol <- "condition"
+valuecol <- "se"
+gathercols <- c("se_black_march", "se_white_march", "se_black_shut", "se_white_shut",
+                "se_black_destroy", "se_white_destroy")
+
+us_survey_wave1_balance_se_long = gather_(us_survey_wave1_balance_se, keycol, valuecol, gathercols)
+
+us_survey_wave1_balance_long = cbind(us_survey_wave1_balance_means_long, us_survey_wave1_balance_se_long[,"se"])
+
+colnames(us_survey_wave1_balance_long) = c("term", "model", "estimate", "std.error")
+
+us_survey_wave1_balance_long$model[us_survey_wave1_balance_long$model == "means_black_destroy"] = "Black, destroy property"
+us_survey_wave1_balance_long$model[us_survey_wave1_balance_long$model == "means_black_march"] = "Black, march in streets"
+us_survey_wave1_balance_long$model[us_survey_wave1_balance_long$model == "means_black_shut"] = "Black, shut down traffic"
+us_survey_wave1_balance_long$model[us_survey_wave1_balance_long$model == "means_white_destroy"] = "White, destroy property"
+us_survey_wave1_balance_long$model[us_survey_wave1_balance_long$model == "means_white_march"] = "White, march in streets"
+us_survey_wave1_balance_long$model[us_survey_wave1_balance_long$model == "means_white_shut"] = "White, shut down traffic"
+
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "age"] = "Age"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "female"] = "Female"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "education_numeric"] = "Education (1-6 scale)"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "income_numeric"] = "Income (1-17 scale)"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "partyID_D"] = "Party ID: Democrat"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "partyID_I"] = "Party ID: Independent"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "partyID_R"] = "Party ID: Republican"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "ideology_numeric"] = "Ideology (1-6 scale)"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "race_White"] = "Race: White"
+us_survey_wave1_balance_long$term[us_survey_wave1_balance_long$term == "race_Black"] = "Race: Black"
+
+us_survey_wave1_balance_long = us_survey_wave1_balance_long[us_survey_wave1_balance_long$term != "Age",]
+
+dwplot(us_survey_wave1_balance_long) + theme_minimal()  + theme(legend.title=element_blank()) +
+  geom_vline(xintercept = 0, colour = "grey60", linetype = 2) 
+ggsave("Figures/fig_A6.pdf", height=5, width=7)

23/replication_package/Code/replicate_fig_A7.R

@@ -0,0 +1,133 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A7
+##--##--##--##--##--##--##
+
+isr_survey_wave1$age_numeric = as.numeric(isr_survey_wave1$age)
+age1 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==0, "age_numeric"], na.rm=T)
+age2 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==1, "age_numeric"], na.rm=T)
+age3 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==2, "age_numeric"], na.rm=T)
+age4 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==0, "age_numeric"], na.rm=T)
+age5 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==1, "age_numeric"], na.rm=T)
+age6 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==2, "age_numeric"], na.rm=T)
+age7 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==0, "age_numeric"], na.rm=T)
+age8 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==1, "age_numeric"], na.rm=T)
+age9 = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==2, "age_numeric"], na.rm=T)
+
+ages_isr_survey_wave1 = c(age1, age2, age3, age4, age5, age6, age7, age8, age9)
+
+
+
+isr_survey_wave1 = dummy_cols(isr_survey_wave1, select_columns = c("partyID", "ethnicity"))
+
+vars_isr_survey_wave1 = c("female", "education", "income", "partyID_C", "partyID_L", "partyID_R", "ideology",
+                          "ethnicity_Ethiopia", "ethnicity_Arab", "ethnicity_Ashkenazi", "ethnicity_Mizrachi", "ethnicity_Soviet Union")
+
+means_black_march = means_arab_march = means_white_march = rep(NA, length(vars_isr_survey_wave1))
+means_black_shut = means_arab_shut = means_white_shut = rep(NA, length(vars_isr_survey_wave1))
+means_black_destroy = means_arab_destroy = means_white_destroy =  rep(NA, length(vars_isr_survey_wave1))
+
+se_black_march = se_arab_march = se_white_march = rep(NA, length(vars_isr_survey_wave1))
+se_black_shut = se_arab_shut = se_white_shut = rep(NA, length(vars_isr_survey_wave1))
+se_black_destroy = se_arab_destroy = se_white_destroy =  rep(NA, length(vars_isr_survey_wave1))
+
+for(i in 1:length(vars_isr_survey_wave1)){
+  
+  means_black_march[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==0, vars_isr_survey_wave1[i]], na.rm=T)
+  se_black_march[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==0, vars_isr_survey_wave1[i]], na.rm=T)
+  means_arab_march[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==0, vars_isr_survey_wave1[i]], na.rm=T)
+  se_arab_march[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==0, vars_isr_survey_wave1[i]], na.rm=T)
+  means_white_march[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==0, vars_isr_survey_wave1[i]], na.rm=T)
+  se_white_march[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==0, vars_isr_survey_wave1[i]], na.rm=T)
+  
+  means_black_shut[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==1, vars_isr_survey_wave1[i]], na.rm=T)
+  se_black_shut[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==1, vars_isr_survey_wave1[i]], na.rm=T)
+  means_arab_shut[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==1, vars_isr_survey_wave1[i]], na.rm=T)
+  se_arab_shut[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==1, vars_isr_survey_wave1[i]], na.rm=T)
+  means_white_shut[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==1, vars_isr_survey_wave1[i]], na.rm=T)
+  se_white_shut[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==1, vars_isr_survey_wave1[i]], na.rm=T)
+  
+  means_black_destroy[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==2, vars_isr_survey_wave1[i]], na.rm=T)
+  se_black_destroy[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==1 & isr_survey_wave1$tactic==2, vars_isr_survey_wave1[i]], na.rm=T)
+  means_arab_destroy[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==2, vars_isr_survey_wave1[i]], na.rm=T)
+  se_arab_destroy[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==2 & isr_survey_wave1$tactic==2, vars_isr_survey_wave1[i]], na.rm=T)
+  means_white_destroy[i] = mean(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==2, vars_isr_survey_wave1[i]], na.rm=T)
+  se_white_destroy[i] = sd(isr_survey_wave1[isr_survey_wave1$identity_protesters==0 & isr_survey_wave1$tactic==2, vars_isr_survey_wave1[i]], na.rm=T)
+  
+}
+
+israel_balance_means = as.data.frame(cbind(means_black_march,
+                                           means_arab_march, 
+                                           means_white_march, 
+                                           means_black_shut,
+                                           means_arab_shut,
+                                           means_white_shut,
+                                           means_black_destroy,
+                                           means_arab_destroy,
+                                           means_white_destroy))
+
+rownames(israel_balance_means) = vars_isr_survey_wave1
+israel_balance_means$var = vars_isr_survey_wave1
+
+keycol <- "condition"
+valuecol <- "mean"
+gathercols <- c("means_black_march", "means_arab_march", "means_white_march", "means_black_shut", "means_arab_shut", "means_white_shut",
+                "means_black_destroy", "means_white_destroy", "means_black_destroy", "means_arab_destroy",
+                "means_white_destroy")
+
+israel_balance_means_long = gather_(israel_balance_means, keycol, valuecol, gathercols)
+
+
+israel_balance_se = as.data.frame(cbind(se_black_march,
+                                        se_arab_march, 
+                                        se_white_march, 
+                                        se_black_shut,
+                                        se_arab_shut,
+                                        se_white_shut,
+                                        se_black_destroy,
+                                        se_arab_destroy,
+                                        se_white_destroy))
+
+rownames(israel_balance_se) = vars_isr_survey_wave1
+israel_balance_se$var = vars_isr_survey_wave1
+
+keycol <- "condition"
+valuecol <- "se"
+gathercols <- c("se_black_march", "se_arab_march", "se_white_march", "se_black_shut", "se_arab_shut", "se_white_shut",
+                "se_black_destroy", "se_white_destroy", "se_black_destroy", "se_arab_destroy",
+                "se_white_destroy")
+
+israel_balance_se_long = gather_(israel_balance_se, keycol, valuecol, gathercols)
+
+israel_balance_long = cbind(israel_balance_means_long, israel_balance_se_long[,"se"])
+
+colnames(israel_balance_long) = c("term", "model", "estimate", "std.error")
+
+israel_balance_long$model[israel_balance_long$model == "means_black_destroy"] = "Black, destroy property"
+israel_balance_long$model[israel_balance_long$model == "means_black_march"] = "Black, march in streets"
+israel_balance_long$model[israel_balance_long$model == "means_black_shut"] = "Black, shut down traffic"
+israel_balance_long$model[israel_balance_long$model == "means_white_destroy"] = "White, destroy property"
+israel_balance_long$model[israel_balance_long$model == "means_white_march"] = "White, march in streets"
+israel_balance_long$model[israel_balance_long$model == "means_white_shut"] = "White, shut down traffic"
+israel_balance_long$model[israel_balance_long$model == "means_arab_destroy"] = "Arab, destroy property"
+israel_balance_long$model[israel_balance_long$model == "means_arab_march"] = "Arab, march in streets"
+israel_balance_long$model[israel_balance_long$model == "means_arab_shut"] = "Arab, shut down traffic"
+
+
+israel_balance_long$term[israel_balance_long$term == "age_numeric"] = "Age (1-6 scale)"
+israel_balance_long$term[israel_balance_long$term == "education"] = "Education (0-6 scale)"
+israel_balance_long$term[israel_balance_long$term == "ethnicity2_Ethiopia"] = "Ethnicity: Jewish Ethiopian"
+israel_balance_long$term[israel_balance_long$term == "female"] = "Female"
+israel_balance_long$term[israel_balance_long$term == "ideology"] = "Ideology (1-7 scale)"
+israel_balance_long$term[israel_balance_long$term == "income"] = "Income (0-8 scale)"
+israel_balance_long$term[israel_balance_long$term == "partyID_C"] = "Party ID: Center"
+israel_balance_long$term[israel_balance_long$term == "partyID_L"] = "Party ID: Left"
+israel_balance_long$term[israel_balance_long$term == "partyID_R"] = "Party ID: Right"
+israel_balance_long$term[israel_balance_long$term == "ethnicity2_Arab"] = "Ethnicity: Arab"
+israel_balance_long$term[israel_balance_long$term == "ethnicity2_Ashkenazi"] = "Ethnicity: Jewish Ashkenazi"
+israel_balance_long$term[israel_balance_long$term == "ethnicity2_Mizrachi"] = "Ethnicity: Jewish Mizrachi"
+israel_balance_long$term[israel_balance_long$term == "ethnicity2_Soviet Union"] = "Ethnicity: Jewish from Soviet Union"
+
+
+dwplot(israel_balance_long) + theme_minimal()  + theme(legend.title=element_blank()) +
+  geom_vline(xintercept = 0, colour = "grey60", linetype = 2)
+ggsave("Figures/fig_A7.pdf", height=5, width=7)

23/replication_package/Code/replicate_fig_A8.R

@@ -0,0 +1,150 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A8
+##--##--##--##--##--##--##
+
+us_survey_wave2$age_numeric = as.numeric(us_survey_wave2$age)
+age1 = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age2 = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+age3 = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age4 = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+age5 = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age6 = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+age7 = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age8 = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+
+ages_us_survey_wave2 = c(age1, age2, age3, age4, age5, age6, age7, age8)
+
+
+us_survey_wave2$education_numeric = as.numeric(us_survey_wave2$education)
+us_survey_wave2$income_numeric = as.numeric(us_survey_wave2$income)
+us_survey_wave2$ideology_numeric = as.numeric(us_survey_wave2$pol_views) +1
+
+us_survey_wave2 = dummy_cols(us_survey_wave2, select_columns = c("partyID", "race"))
+vars_us_survey_wave2 = c("female", "education_numeric", "income_numeric", "partyID_D", "partyID_I", "partyID_R", "ideology_numeric",  "race_White / Caucasian","race_Black / African American")
+
+means_black_generic_nocomm = means_black_generic_comm = means_black_group_nocomm = means_black_group_comm = 
+  means_white_generic_nocomm = means_white_generic_comm = means_white_group_nocomm = means_white_group_comm = rep(NA, length(vars_us_survey_wave2))
+
+se_black_generic_nocomm = se_black_generic_comm = se_black_group_nocomm = se_black_group_comm = 
+  se_white_generic_nocomm = se_white_generic_comm = se_white_group_nocomm = se_white_group_comm = rep(NA, length(vars_us_survey_wave2))
+
+
+for(i in 1:length(vars_us_survey_wave2)){
+  means_black_generic_nocomm[i] = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  means_black_generic_comm[i] = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  means_black_group_nocomm[i] = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  means_black_group_comm[i] = mean(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  means_white_generic_nocomm[i] = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  means_white_generic_comm[i] = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  means_white_group_nocomm[i] = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  means_white_group_comm[i] = mean(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  
+  se_black_generic_nocomm[i] = sd(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  se_black_generic_comm[i] = sd(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  se_black_group_nocomm[i] = sd(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  se_black_group_comm[i] = sd(us_survey_wave2[us_survey_wave2$black==1 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  se_white_generic_nocomm[i] = sd(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  se_white_generic_comm[i] = sd(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==0 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  se_white_group_nocomm[i] = sd(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==0, vars_us_survey_wave2[i]], na.rm=T)
+  se_white_group_comm[i] = sd(us_survey_wave2[us_survey_wave2$black==0 & us_survey_wave2$group_goal==1 & us_survey_wave2$commitment==1, vars_us_survey_wave2[i]], na.rm=T)
+  
+}
+
+us_survey_wave2_balance = cbind(means_black_generic_nocomm, se_black_generic_nocomm,
+                                means_black_generic_comm, se_black_generic_comm, 
+                                means_black_group_nocomm, se_black_group_nocomm,
+                                means_black_group_comm, se_black_group_comm,
+                                means_white_generic_nocomm, se_white_generic_nocomm,
+                                means_white_generic_comm, se_white_generic_comm,
+                                means_white_group_nocomm, se_white_group_nocomm,
+                                means_white_group_comm, se_white_group_comm)
+
+
+
+
+us_survey_wave2_balance = as.data.frame(us_survey_wave2_balance)
+rownames(us_survey_wave2_balance) = vars_us_survey_wave2
+
+
+us_survey_wave2_balance_means = as.data.frame(cbind(means_black_generic_nocomm,
+                                                    means_black_generic_comm, 
+                                                    means_black_group_nocomm,
+                                                    means_black_group_comm,
+                                                    means_white_generic_nocomm,
+                                                    means_white_generic_comm,
+                                                    means_white_group_nocomm,
+                                                    means_white_group_comm))
+
+rownames(us_survey_wave2_balance_means) = vars_us_survey_wave2
+us_survey_wave2_balance_means$var = vars_us_survey_wave2
+
+keycol <- "condition"
+valuecol <- "mean"
+gathercols <- c("means_black_generic_nocomm",
+                "means_black_generic_comm", 
+                "means_black_group_nocomm",
+                "means_black_group_comm",
+                "means_white_generic_nocomm",
+                "means_white_generic_comm",
+                "means_white_group_nocomm",
+                "means_white_group_comm")
+
+us_survey_wave2_balance_means_long = gather_(us_survey_wave2_balance_means, keycol, valuecol, gathercols)
+
+
+us_survey_wave2_balance_se = as.data.frame(cbind(se_black_generic_nocomm,
+                                                 se_black_generic_comm, 
+                                                 se_black_group_nocomm,
+                                                 se_black_group_comm,
+                                                 se_white_generic_nocomm,
+                                                 se_white_generic_comm,
+                                                 se_white_group_nocomm,
+                                                 se_white_group_comm))
+
+rownames(us_survey_wave2_balance_se) = vars_us_survey_wave2
+us_survey_wave2_balance_se$var = vars_us_survey_wave2
+
+keycol <- "condition"
+valuecol <- "se"
+gathercols <- c("se_black_generic_nocomm",
+                "se_black_generic_comm", 
+                "se_black_group_nocomm",
+                "se_black_group_comm",
+                "se_white_generic_nocomm",
+                "se_white_generic_comm",
+                "se_white_group_nocomm",
+                "se_white_group_comm")
+
+us_survey_wave2_balance_se_long = gather_(us_survey_wave2_balance_se, keycol, valuecol, gathercols)
+
+us_survey_wave2_balance_long = cbind(us_survey_wave2_balance_means_long, us_survey_wave2_balance_se_long[,"se"])
+
+colnames(us_survey_wave2_balance_long) = c("term", "model", "estimate", "std.error")
+
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_black_generic_nocomm"] = "Black, generic goal, no commitment"
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_black_generic_comm"] = "Black, generic goal, commitment"
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_black_group_nocomm"] = "Black, group goal, no commitment"
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_black_group_comm"] = "Black, group goal, no commitment"
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_white_generic_nocomm"] = "White, generic goal, no commitment"
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_white_generic_comm"] = "White, generic goal, commitment"
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_white_group_nocomm"] = "White, group goal, no commitment"
+us_survey_wave2_balance_long$model[us_survey_wave2_balance_long$model == "means_white_group_comm"] = "White, group goal, no commitment"
+
+
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "age"] = "Age"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "female"] = "Female"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "education_numeric"] = "Education (1-7 scale)"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "income_numeric"] = "Income (1-6 scale)"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "partyID_D"] = "Party ID: Democrat"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "partyID_I"] = "Party ID: Independent"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "partyID_R"] = "Party ID: Republican"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "ideology_numeric"] = "Ideology (1-7 scale)"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "race_White / Caucasian"] = "Race: White"
+us_survey_wave2_balance_long$term[us_survey_wave2_balance_long$term == "race_Black / African American"] = "Race: Black"
+
+
+us_survey_wave2_balance_long = us_survey_wave2_balance_long[us_survey_wave2_balance_long$term != "Age",]
+
+dwplot(us_survey_wave2_balance_long) + theme_minimal()  + theme(legend.title=element_blank()) +
+  geom_vline(xintercept = 0, colour = "grey60", linetype = 2) 
+ggsave("Figures/fig_A8.pdf", height=5, width=7)

23/replication_package/Code/replicate_fig_A9.R

@@ -0,0 +1,149 @@
+##--##--##--##--##--##--##
+## Appendix: Figure A9
+##--##--##--##--##--##--##
+
+isr_survey_wave2 = dummy_cols(isr_survey_wave2, select_columns = c("partyID", "ethnicity"))
+
+vars_isr_survey_wave2 = c("female", "education", "income", "partyID_L", "partyID_R", "ideology",
+                          "ethnicity_Ethiopia", "ethnicity_Ashkenazi", "ethnicity_Mizrachi", "ethnicity_Soviet Union")
+
+isr_survey_wave2$age_numeric = as.numeric(isr_survey_wave2$age)
+age1 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age2 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+age3 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age4 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+age5 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age6 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+age7 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age8 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+age9 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==0 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, "age_numeric"], na.rm=T)
+age10 = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==0 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, "age_numeric"], na.rm=T)
+
+ages_isr_survey_wave2 = c(age1, age2, age3, age4, age5, age6, age7, age8, age9, age10)
+
+means_eth_generic_nocomm = means_eth_generic_comm = means_eth_group_nocomm = means_eth_group_comm = 
+  means_arab_generic_nocomm = means_arab_generic_comm = means_arab_group_nocomm = means_arab_group_comm = 
+  means_white_generic_nocomm = means_white_generic_comm = rep(NA, length(vars_isr_survey_wave2))
+
+se_eth_generic_nocomm = se_eth_generic_comm = se_eth_group_nocomm = se_eth_group_comm = 
+  se_arab_generic_nocomm = se_arab_generic_comm = se_arab_group_nocomm = se_arab_group_comm = 
+  se_white_generic_nocomm = se_white_generic_comm = rep(NA, length(vars_isr_survey_wave2))
+
+
+for(i in 1:length(vars_isr_survey_wave2)){
+  means_eth_generic_nocomm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  means_eth_generic_comm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  means_eth_group_nocomm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  means_eth_group_comm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  means_arab_generic_nocomm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  means_arab_generic_comm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  means_arab_group_nocomm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  means_arab_group_comm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  means_white_generic_nocomm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==0 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  means_white_generic_comm[i] = mean(isr_survey_wave2[isr_survey_wave2$identity_protesters==0 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  
+  se_eth_generic_nocomm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  se_eth_generic_comm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  se_eth_group_nocomm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  se_eth_group_comm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==1 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  se_arab_generic_nocomm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  se_arab_generic_comm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  se_arab_group_nocomm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  se_arab_group_comm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==2 & isr_survey_wave2$group_goal==1 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  se_white_generic_nocomm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==0 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==0, vars_isr_survey_wave2[i]], na.rm=T)
+  se_white_generic_comm[i] = sd(isr_survey_wave2[isr_survey_wave2$identity_protesters==0 & isr_survey_wave2$group_goal==0 & isr_survey_wave2$commitment==1, vars_isr_survey_wave2[i]], na.rm=T)
+  
+}
+
+
+isr_survey_wave2_balance_means = as.data.frame(cbind(means_eth_generic_nocomm,
+                                                     means_eth_generic_comm, 
+                                                     means_eth_group_nocomm,
+                                                     means_eth_group_comm,
+                                                     means_arab_generic_nocomm,
+                                                     means_arab_generic_comm,
+                                                     means_arab_group_nocomm,
+                                                     means_arab_group_comm,
+                                                     means_white_generic_nocomm,
+                                                     means_white_generic_comm))
+
+rownames(isr_survey_wave2_balance_means) = vars_isr_survey_wave2
+isr_survey_wave2_balance_means$var = vars_isr_survey_wave2
+
+keycol <- "condition"
+valuecol <- "mean"
+gathercols <- c("means_eth_generic_nocomm",
+                "means_eth_generic_comm", 
+                "means_eth_group_nocomm",
+                "means_eth_group_comm",
+                "means_arab_generic_nocomm",
+                "means_arab_generic_comm",
+                "means_arab_group_nocomm",
+                "means_arab_group_comm",
+                "means_white_generic_nocomm",
+                "means_white_generic_comm")
+
+isr_survey_wave2_balance_means_long = gather_(isr_survey_wave2_balance_means, keycol, valuecol, gathercols)
+
+isr_survey_wave2_balance_se = as.data.frame(cbind(se_eth_generic_nocomm,
+                                                  se_eth_generic_comm, 
+                                                  se_eth_group_nocomm,
+                                                  se_eth_group_comm,
+                                                  se_arab_generic_nocomm,
+                                                  se_arab_generic_comm,
+                                                  se_arab_group_nocomm,
+                                                  se_arab_group_comm,
+                                                  se_white_generic_nocomm,
+                                                  se_white_generic_comm))
+
+rownames(isr_survey_wave2_balance_se) = vars_isr_survey_wave2
+isr_survey_wave2_balance_se$var = vars_isr_survey_wave2
+
+keycol <- "condition"
+valuecol <- "se"
+gathercols <- c("se_eth_generic_nocomm",
+                "se_eth_generic_comm", 
+                "se_eth_group_nocomm",
+                "se_eth_group_comm",
+                "se_arab_generic_nocomm",
+                "se_arab_generic_comm",
+                "se_arab_group_nocomm",
+                "se_arab_group_comm",
+                "se_white_generic_nocomm",
+                "se_white_generic_comm")
+
+isr_survey_wave2_balance_se_long = gather_(isr_survey_wave2_balance_se, keycol, valuecol, gathercols)
+
+isr_survey_wave2_balance_long = cbind(isr_survey_wave2_balance_means_long, isr_survey_wave2_balance_se_long[,"se"])
+
+colnames(isr_survey_wave2_balance_long) = c("term", "model", "estimate", "std.error")
+
+
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_eth_generic_nocomm"] = "Ethiopian, generic goal, no commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_eth_generic_comm"] = "Ethiopian, generic goal, commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_eth_group_nocomm"] = "Ethiopian, group goal, no commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_eth_group_comm"] = "Ethiopian, group goal, no commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_arab_generic_nocomm"] = "Arab, generic goal, no commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_arab_generic_comm"] = "Arab, generic goal, commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_arab_group_nocomm"] = "Arab, group goal, no commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_arab_group_comm"] = "Arab, group goal, no commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_white_generic_nocomm"] = "White, generic goal, no commitment"
+isr_survey_wave2_balance_long$model[isr_survey_wave2_balance_long$model == "means_white_generic_comm"] = "White, generic goal, commitment"
+
+
+
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "education"] = "Education (0-6 scale)"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "ethnicity_Ethiopia"] = "Ethnicity: Ethiopian"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "ethnicity_Ashkenazi"] = "Ethnicity: Ashkenazi"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "ethnicity_Mizrachi"] = "Ethnicity: Mizrachi"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "ethnicity_Soviet Union"] = "Ethnicity: Soviet Union"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "female"] = "Female"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "ideology"] = "Ideology (1-7 scale)"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "income"] = "Income (0-8 scale)"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "partyID_L"] = "Party ID: Left"
+isr_survey_wave2_balance_long$term[isr_survey_wave2_balance_long$term == "partyID_R"] = "Party ID: Right"
+
+
+dwplot(isr_survey_wave2_balance_long) + theme_minimal()  + theme(legend.title=element_blank()) +
+  geom_vline(xintercept = 0, colour = "grey60", linetype = 2)
+ggsave("Figures/fig_A9.pdf", height=5, width=7)

23/replication_package/Code/replicate_table_1.R

@@ -0,0 +1,24 @@
+##--##--##--##
+## Table 1
+##--##--##--##
+
+## A. U.S
+degv_yg = lm(degree_violence ~ identity_protesters + tactic  , data=us_survey_wave1, weights=weight)
+police_yg = lm(police_action_required ~ identity_protesters + tactic  , data=us_survey_wave1, weights=weight)
+recall_violence_yg = lm(recall_violence2 ~ identity_protesters + tactic  , data=us_survey_wave1, weights=weight)
+approve_protest_yg = lm(approve_protest_scale ~ identity_protesters + tactic  , data=us_survey_wave1, weights=weight)
+stargazer(degv_yg, recall_violence_yg, police_yg, 
+          covariate.labels = c("Black protesters", "Shut down traffic", 
+                               "Destroy police cars", "Intercept: White protesters, March in streets"),
+          out = "Tables/table_1_panel_A.tex")
+
+## B. Israel
+degv_isr = lm(degree_violence ~ identity_protesters + tactic  , data=isr_survey_wave1, weights=weight)
+police_isr = lm(police_action_required ~ identity_protesters + tactic  , data=isr_survey_wave1, weights=weight)
+recall_violence_isr = lm(recall_violence2 ~ identity_protesters + tactic  , data=isr_survey_wave1, weights=weight)
+approve_protest_isr = lm(approve_protest_scale ~ identity_protesters + tactic  , data=isr_survey_wave1, weights=weight)
+stargazer(degv_isr, recall_violence_isr, police_isr,
+          covariate.labels = c("Ethiopian protesters", "Arab protesters",
+                               "Shut down traffic", "Destroy garbage cans",
+                               "Intercept: White protesters, March in streets"),
+          out = "Tables/table_1_panel_B.tex")

23/replication_package/Code/replicate_table_4.R

@@ -0,0 +1,22 @@
+##--##--##--##
+## Table 4
+##--##--##--##
+
+## A. U.S. sample
+mod1 = lm(degree_violence ~ black + group_goal + commitment , data=us_survey_wave2)
+mod2 = lm(recall_violence2 ~ black + group_goal + commitment, data=us_survey_wave2)
+mod3 = lm(police_action_required ~ black + group_goal + commitment, data=us_survey_wave2)
+stargazer(mod1, mod2, mod3, 
+          covariate.labels = c("Black protesters", "Minority group goal",
+                               "Commitment to nonviolence", "Intercept: White protesters, generic goal, no commitment"),
+          out = "Tables/table_4_panel_A.tex")
+
+## B. Israel sample
+mod1_israel = lm(degree_violence ~ identity_protesters + group_goal + commitment, data=isr_survey_wave2)
+mod2_israel = lm(recall_violence2 ~ identity_protesters + group_goal + commitment, data=isr_survey_wave2)
+mod3_israel = lm(police_action_required ~ identity_protesters + group_goal + commitment, data=isr_survey_wave2)
+stargazer(mod1_israel, mod2_israel, mod3_israel,
+          covariate.labels = c("Ethiopian protesters", "Arab protesters",
+                               "Minority group goal", "Commitment to nonviolence",
+                               "Intercept: White protesters, generic goal, no commitment"),
+          out = "Tables/table_4_panel_B.tex")

23/replication_package/Code/replicate_table_A1.R

@@ -0,0 +1,15 @@
+##--##--##--##--##--##--##
+## Appendix: Table A1
+##--##--##--##--##--##--##
+
+stargazer(EBCR_EPR_NAVCO2[,c("YEAR", "success", "INIT_NV_ONSET", "EPR_GROUPSIZE", "EPR_STATUS_ORD",
+                             "EPR_STATUS_EXCL", "POP_LOG_LAG_EXT", "GDPPC_LOG_LAG_EXT",
+                             "PASTNV", "PASTV", "VDEM_POLYARCHY_LAG", "VDEM_PHYSINT_LAG", 
+                             "EPR_TEK_EGIP", "EPR_DOWNGRADED5", "HORIZ_INEQ", "NVYEARS",
+                             "VYEARS")],
+          covariate.labels = c("Year", "Campaign success", "NV campaign", "EPR group size", "EPR status",
+                               "EPR Status: excluded", "Country population (logged)", "Country GDP per capita (logged)", "Prior participation in nonviolence", "Prior participation in violence",
+                               "Level of democracy", "Physical integrity index", "Neighboring kin in power",
+                               "Downgraded", "Horizontal inequality", "Nonviolent years", "Violent years"),
+          out="Tables/table_A1.tex")
+

23/replication_package/Code/replicate_table_A10.R

@@ -0,0 +1,9 @@
+##--##--##--##--##--##--##
+## Appendix: Table A10
+##--##--##--##--##--##--##
+
+us_wave2_sumstats = us_survey_wave2[,c("age", "female", "education", "income", "partyID", "pol_views", "interest_politics",
+                                       "degree_violence", "police_action_required", "recall_violence2", "race")]
+us_wave2_sumstats = dummy_cols(us_wave2_sumstats, select_columns = c("age", "education", "income", "partyID", "pol_views", "race"))
+table_a10 = stargazer(us_wave2_sumstats, omit.summary.stat=c("p25", "p75"))
+cat(table_a10, sep = '\n', file = "Tables/table_A10.tex")

23/replication_package/Code/replicate_table_A11.R

@@ -0,0 +1,9 @@
+##--##--##--##--##--##--##
+## Appendix: Table A11
+##--##--##--##--##--##--##
+
+isr_wave2_sumstats = isr_survey_wave2[,c("age", "female", "religiosity", "ethnicity", "education", "income", "partyID", "ideology", 
+                                         "degree_violence", "police_action_required", "recall_violence2")]
+isr_wave2_sumstats = dummy_cols(isr_wave2_sumstats, select_columns = c("age", "religiosity", "ethnicity", "education", "income", "partyID"))
+table_a11 = stargazer(isr_wave2_sumstats, omit.summary.stat=c("p25", "p75"))
+cat(table_a11, sep = '\n', file = "Tables/table_A11.tex")

23/replication_package/Code/replicate_table_A13.R

@@ -0,0 +1,15 @@
+##--##--##--##--##--##--##
+## Appendix: Table A13
+##--##--##--##--##--##--##
+
+## A. U.S. sample
+degv_yg = lm(degree_violence ~ identity_protesters + tactic + age +female + education + income  + ideology + race, data=us_survey_wave1, weights=weight)
+police_yg = lm(police_action_required ~ identity_protesters + tactic  + age +female + education + income  + ideology + race, data=us_survey_wave1, weights=weight)
+recall_violence_yg = lm(recall_violence2 ~ identity_protesters + tactic  + age +female + education + income  + ideology + race, data=us_survey_wave1, weights=weight)
+stargazer(degv_yg, recall_violence_yg, police_yg, out = "Tables/table_A13_panel_A.tex")
+
+## B. Israel sample
+degv_ip = lm(degree_violence ~ identity_protesters + tactic + age +female + education + income  + ideology + ethnicity, data=isr_survey_wave1, weights=weight)
+police_ip = lm(police_action_required ~ identity_protesters + tactic  + age +female + education + income  + ideology + ethnicity , data=isr_survey_wave1, weights=weight)
+recall_violence_ip = lm(recall_violence2 ~ identity_protesters + tactic  + age +female + education + income + ideology  + ethnicity , data=isr_survey_wave1, weights=weight)
+stargazer(degv_ip, recall_violence_ip, police_ip, out = "Tables/table_A13_panel_B.tex")

23/replication_package/Code/replicate_table_A14.R

@@ -0,0 +1,194 @@
+##--##--##--##
+## Table A14
+##--##--##--##
+
+us_survey_wave1$identity_protesters_fac = as.factor(us_survey_wave1$identity_protesters)
+levels(us_survey_wave1$identity_protesters_fac) = c("White", "Black")
+us_survey_wave1$tactic_fac = as.factor(us_survey_wave1$tactic)
+levels(us_survey_wave1$tactic_fac) = c("March in streets", "Shut down traffic", "Destroy police cars")
+
+isr_survey_wave1$identity_protesters_fac = as.factor(isr_survey_wave1$identity_protesters)
+levels(isr_survey_wave1$identity_protesters_fac) = c("White", "Ethiopian", "Arab")
+isr_survey_wave1$tactic_fac = as.factor(isr_survey_wave1$tactic)
+levels(isr_survey_wave1$tactic_fac) = c("March in streets", "Shut down traffic", "Destroy garbage cans")
+
+
+## US sample
+us_survey_wave1$degree_violence_std = scale(us_survey_wave1$degree_violence)
+us_survey_wave1$police_action_required_std = scale(us_survey_wave1$police_action_required)
+us_survey_wave1$recall_violence2_std = scale(us_survey_wave1$recall_violence2)
+
+march1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+shut1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==1,], weights=weight)
+destroy1 = lm_robust(degree_violence_std ~ identity_protesters_fac , data=us_survey_wave1[us_survey_wave1$tactic==2,], weights=weight)
+
+march2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+shut2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==1,], weights=weight)
+destroy2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==2,], weights=weight)
+
+march3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==0,], weights=weight)
+shut3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=us_survey_wave1[us_survey_wave1$tactic==1,], weights=weight)
+destroy3 = lm_robust(recall_violence2_std ~ identity_protesters_fac , data=us_survey_wave1[us_survey_wave1$tactic==2,], weights=weight)
+
+# Plot differences
+
+term_degree_violence = c(march1$coefficients[2], shut1$coefficients[2], destroy1$coefficients[2])
+se_degree_violence = c(march1$std.error[2], shut1$std.error[2], destroy1$std.error[2])
+statistic_degree_violence = c(march1$statistic[2], shut1$statistic[2], destroy1$statistic[2])
+pval_degree_violence = c(march1$p.value[2], shut1$p.value[2], destroy1$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+term_police_action_required = c(march2$coefficients[2], shut2$coefficients[2], destroy2$coefficients[2])
+se_police_action_required = c(march2$std.error[2], shut2$std.error[2], destroy2$std.error[2])
+statistic_police_action_required = c(march2$statistic[2], shut2$statistic[2], destroy2$statistic[2])
+pval_police_action_required = c(march2$p.value[2], shut2$p.value[2], destroy2$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+
+term_recall_violence = c(march3$coefficients[2], shut3$coefficients[2], destroy3$coefficients[2])
+se_recall_violence = c(march3$std.error[2], shut3$std.error[2], destroy3$std.error[2])
+statistic_recall_violence = c(march3$statistic[2], shut3$statistic[2], destroy3$statistic[2])
+pval_recall_violence = c(march3$p.value[2], shut3$p.value[2], destroy3$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+
+differences_us = rbind(degree_violence, police_action_required, recall_violence)
+
+
+## Israel respondents
+isr_survey_wave1$degree_violence_std = scale(isr_survey_wave1$degree_violence)
+isr_survey_wave1$police_action_required_std = scale(isr_survey_wave1$police_action_required)
+isr_survey_wave1$recall_violence2_std = scale(isr_survey_wave1$recall_violence2)
+
+march1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+shut1 = lm_robust(degree_violence_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==1,], weights=weight)
+destroy1 = lm_robust(degree_violence_std ~ identity_protesters_fac , data=isr_survey_wave1[isr_survey_wave1$tactic==2,], weights=weight)
+
+march2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+shut2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==1,], weights=weight)
+destroy2 = lm_robust(police_action_required_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==2,], weights=weight)
+
+march3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==0,], weights=weight)
+shut3 = lm_robust(recall_violence2_std ~ identity_protesters_fac, data=isr_survey_wave1[isr_survey_wave1$tactic==1,], weights=weight)
+destroy3 = lm_robust(recall_violence2_std ~ identity_protesters_fac , data=isr_survey_wave1[isr_survey_wave1$tactic==2,], weights=weight)
+
+
+# Ethiopian protesters
+term_degree_violence = c(march1$coefficients[2], shut1$coefficients[2], destroy1$coefficients[2])
+se_degree_violence = c(march1$std.error[2], shut1$std.error[2], destroy1$std.error[2])
+statistic_degree_violence = c(march1$statistic[2], shut1$statistic[2], destroy1$statistic[2])
+pval_degree_violence = c(march1$p.value[2], shut1$p.value[2], destroy1$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+term_police_action_required = c(march2$coefficients[2], shut2$coefficients[2], destroy2$coefficients[2])
+se_police_action_required = c(march2$std.error[2], shut2$std.error[2], destroy2$std.error[2])
+statistic_police_action_required = c(march2$statistic[2], shut2$statistic[2], destroy2$statistic[2])
+pval_police_action_required = c(march2$p.value[2], shut2$p.value[2], destroy2$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+
+term_recall_violence = c(march3$coefficients[2], shut3$coefficients[2], destroy3$coefficients[2])
+se_recall_violence = c(march3$std.error[2], shut3$std.error[2], destroy3$std.error[2])
+statistic_recall_violence = c(march3$statistic[2], shut3$statistic[2], destroy3$statistic[2])
+pval_recall_violence = c(march3$p.value[2], shut3$p.value[2], destroy3$p.value[2])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+
+differences_isr_black = rbind(degree_violence, police_action_required, recall_violence)
+
+term_degree_violence = c(march1$coefficients[3], shut1$coefficients[3], destroy1$coefficients[3])
+se_degree_violence = c(march1$std.error[3], shut1$std.error[3], destroy1$std.error[3])
+statistic_degree_violence = c(march1$statistic[3], shut1$statistic[3], destroy1$statistic[3])
+pval_degree_violence = c(march1$p.value[3], shut1$p.value[3], destroy1$p.value[3])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+degree_violence = as.data.frame(cbind(term_degree_violence, se_degree_violence, statistic_degree_violence, pval_degree_violence))
+degree_violence = cbind(degree_violence, term)
+colnames(degree_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(degree_violence) = term
+degree_violence$model = "Perceived degree of violence"
+
+term_police_action_required = c(march2$coefficients[3], shut2$coefficients[3], destroy2$coefficients[3])
+se_police_action_required = c(march2$std.error[3], shut2$std.error[3], destroy2$std.error[3])
+statistic_police_action_required = c(march2$statistic[3], shut2$statistic[3], destroy2$statistic[3])
+pval_police_action_required = c(march2$p.value[3], shut2$p.value[3], destroy2$p.value[3])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+police_action_required = as.data.frame(cbind(term_police_action_required, se_police_action_required, statistic_police_action_required, pval_police_action_required))
+police_action_required = cbind(police_action_required, term)
+colnames(police_action_required) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(police_action_required) = term
+police_action_required$model = "Police action required"
+
+term_recall_violence = c(march3$coefficients[3], shut3$coefficients[3], destroy3$coefficients[3])
+se_recall_violence = c(march3$std.error[3], shut3$std.error[3], destroy3$std.error[3])
+statistic_recall_violence = c(march3$statistic[3], shut3$statistic[3], destroy3$statistic[3])
+pval_recall_violence = c(march3$p.value[3], shut3$p.value[3], destroy3$p.value[3])
+term = c("1) Minority: March in streets", "2) Minority: Shut down traffic", "3) Minority: Destroy police cars / garbage cans")
+recall_violence = as.data.frame(cbind(term_recall_violence, se_recall_violence, statistic_recall_violence, pval_recall_violence))
+recall_violence = cbind(recall_violence, term)
+colnames(recall_violence) = c("estimate", "std.error", "statistic", "p.value", "term")
+rownames(recall_violence) = term
+recall_violence$model = "Recall violence"
+
+differences_isr_arab = rbind(degree_violence, police_action_required, recall_violence)
+
+## Result tables for plot
+differences_us2 = differences_us
+differences_us2$submodel = differences_us2$term
+differences_us2$term = "Perception of Blacks \n(United States)"
+
+differences_isr_black2 = differences_isr_black
+differences_isr_black2$submodel = differences_isr_black2$term
+differences_isr_black2$term = "Perception of Ethiopians \n(Israel)"
+
+differences_isr_arab2 = differences_isr_arab
+differences_isr_arab2$submodel = differences_isr_arab2$term
+differences_isr_arab2$term = "Perception of Arabs \n(Israel)"
+
+diffs_sm = rbind(differences_us2, differences_isr_arab2, differences_isr_black2)
+rownames(diffs_sm) = 1:nrow(diffs_sm)
+
+results_df <- data.frame(term = diffs_sm$term,
+                         estimate = diffs_sm$estimate,
+                         std.error = diffs_sm$std.error,
+                         model = diffs_sm$model,
+                         submodel = as.character(diffs_sm$submodel),
+                         stringsAsFactors = FALSE)
+
+results_df$submodel[results_df$submodel=="1) Minority: March in streets"] = "March in streets"
+results_df$submodel[results_df$submodel=="2) Minority: Shut down traffic"] = "Shut down traffic"
+results_df$submodel[results_df$submodel=="3) Minority: Destroy police cars / garbage cans"] = "Destroy property"
+
+results_df$model[results_df$model=="Perceived degree of violence"] = "1. Perceived degree \nof violence"
+results_df$model[results_df$model=="Police action required"] = "3. Police action \nrequired"
+results_df$model[results_df$model=="Recall violence"] = "2. Recall \nviolence"
+
+
+table_a14 = xtable(results_df[,-1], digits=2)
+print(table_a14, file="Tables/table_A14.tex")
+

23/replication_package/Code/replicate_table_A15.R

@@ -0,0 +1,25 @@
+##--##--##--##--##--##--##
+## Appendix: Table A15
+##--##--##--##--##--##--##
+
+## A. U.S. sample
+degv_yg = lm(degree_violence ~ identity_protesters * tactic  , data=us_survey_wave1, weights=weight)
+police_yg = lm(police_action_required ~ identity_protesters * tactic  , data=us_survey_wave1, weights=weight)
+recall_violence_yg = lm(recall_violence2 ~ identity_protesters * tactic  , data=us_survey_wave1, weights=weight)
+stargazer(degv_yg, recall_violence_yg, police_yg, 
+          covariate.labels = c("Black protesters", "Shut down traffic", "Destroy police cars",
+                               "Black protesters x Shut down traffic", 
+                               "Black protesters x Destroy police cars", "Intercept"),
+          out = "Tables/table_A15_panel_A.tex")
+
+## B. Israel sample
+degv_isr = lm(degree_violence ~ identity_protesters * tactic  , data=isr_survey_wave1, weights=weight)
+police_isr = lm(police_action_required ~ identity_protesters * tactic  , data=isr_survey_wave1, weights=weight)
+recall_violence_isr = lm(recall_violence2 ~ identity_protesters * tactic  , data=isr_survey_wave1, weights=weight)
+stargazer(degv_isr, recall_violence_isr, police_isr,
+          covariate.labels = c("Ethiopian protesters", "Arab protesters", "Shut down traffic",
+                               "Destroy garbage cans", "Ethiopian protesters x Shut down traffic",
+                               "Arab protesters x Shut down traffic", "Ethiopian protesters x Destroy garbage cans",
+                               "Arab protesters x Destroy garbage cans", "Intercept"),
+          out = "Tables/table_A15_panel_B.tex")
+

23/replication_package/Code/replicate_table_A16.R

@@ -0,0 +1,102 @@
+##--##--##--##--##--##--##
+## Appendix: Table A16
+##--##--##--##--##--##--##
+
+## A. U.S. 
+degv_yg_nv_maj = lm_robust(degree_violence ~ identity_protesters  , data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race=="White",], weights=weight)
+police_yg_nv_maj = lm_robust(police_action_required ~ identity_protesters  , data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race == "White",], weights=weight)
+degv_yg_nv_min = lm_robust(degree_violence ~ identity_protesters  , data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race=="Black",], weights=weight)
+police_yg_nv_min = lm_robust(police_action_required ~ identity_protesters  , data=us_survey_wave1[us_survey_wave1$tactic==0 & us_survey_wave1$race == "Black",], weights=weight)
+
+# Majority perceptions
+us_mean_degv_maj_white = degv_yg_nv_maj$coefficients[1]
+us_mean_degv_maj_black = degv_yg_nv_maj$coefficients[1] + degv_yg_nv_maj$coefficients[2] 
+us_diff_degv_maj = degv_yg_nv_maj$coefficients[2]
+us_pval_diff_degv_maj = degv_yg_nv_maj$p.value[2]
+us_pct_change_degv_maj = (degv_yg_nv_maj$coefficients[2] /us_mean_degv_maj_white)*100
+us_mean_police_maj_white = police_yg_nv_maj$coefficients[1]
+us_mean_police_maj_black = police_yg_nv_maj$coefficients[1] + police_yg_nv_maj$coefficients[2] 
+us_diff_police_maj = police_yg_nv_maj$coefficients[2]
+us_pval_diff_police_maj = police_yg_nv_maj$p.value[2]
+us_pct_change_police_maj = (police_yg_nv_maj$coefficients[2] /us_mean_police_maj_white)*100
+
+# Minority perceptions
+us_mean_degv_min_white = degv_yg_nv_min$coefficients[1]
+us_mean_degv_min_black = degv_yg_nv_min$coefficients[1] + degv_yg_nv_min$coefficients[2] 
+us_diff_degv_min = degv_yg_nv_min$coefficients[2]
+us_pval_diff_degv_min = degv_yg_nv_min$p.value[2]
+us_pct_change_degv_min = (degv_yg_nv_min$coefficients[2] /us_mean_degv_min_white)*100
+us_mean_police_min_white = police_yg_nv_min$coefficients[1]
+us_mean_police_min_black = police_yg_nv_min$coefficients[1] + police_yg_nv_min$coefficients[2] 
+us_diff_police_min = police_yg_nv_min$coefficients[2]
+us_pval_diff_police_min = police_yg_nv_min$p.value[2]
+us_pct_change_police_min = (police_yg_nv_min$coefficients[2] /us_mean_police_min_white)*100
+
+
+## B. ISRAEL
+degv_isr_nv_maj = lm_robust(degree_violence ~ identity_protesters  , data=isr_survey_wave1[isr_survey_wave1$survey=="isr" & isr_survey_wave1$tactic==0 &  isr_survey_wave1$ethnicity!= "Ethiopia",])
+police_isr_nv_maj = lm_robust(police_action_required ~ identity_protesters  , data=isr_survey_wave1[isr_survey_wave1$survey=="isr" & isr_survey_wave1$tactic==0 & isr_survey_wave1$ethnicity!= "Ethiopia",])
+degv_isr_nv_min = lm_robust(degree_violence ~ identity_protesters  , data=isr_survey_wave1[isr_survey_wave1$survey=="isr_ar" & isr_survey_wave1$tactic==0,])
+police_isr_nv_min = lm_robust(police_action_required ~ identity_protesters  , data=isr_survey_wave1[isr_survey_wave1$survey=="isr_ar" & isr_survey_wave1$tactic==0,])
+
+# Majority perceptions
+israel_mean_degv_maj_whitejew = degv_isr_nv_maj$coefficients[1]
+isr_mean_degv_maj_eth = degv_isr_nv_maj$coefficients[1] + degv_isr_nv_maj$coefficients[2] 
+isr_mean_degv_maj_arab = degv_isr_nv_maj$coefficients[1] + degv_isr_nv_maj$coefficients[3] 
+isr_diff_degv_maj_eth = degv_isr_nv_maj$coefficients[2]
+isr_diff_degv_maj_arab = degv_isr_nv_maj$coefficients[3]
+isr_pval_diff_degv_maj_eth = degv_isr_nv_maj$p.value[2]
+isr_pval_diff_degv_maj_arab = degv_isr_nv_maj$p.value[3]
+isr_pct_change_degv_maj_eth = (degv_isr_nv_maj$coefficients[2] /israel_mean_degv_maj_whitejew)*100
+isr_pct_change_degv_maj_arab = (degv_isr_nv_maj$coefficients[3] /israel_mean_degv_maj_whitejew)*100
+israel_mean_police_maj_whitejew = police_isr_nv_maj$coefficients[1]
+isr_mean_police_maj_eth = police_isr_nv_maj$coefficients[1] + police_isr_nv_maj$coefficients[2] 
+isr_mean_police_maj_arab = police_isr_nv_maj$coefficients[1] + police_isr_nv_maj$coefficients[3] 
+isr_diff_police_maj_eth = police_isr_nv_maj$coefficients[2]
+isr_diff_police_maj_arab = police_isr_nv_maj$coefficients[3]
+isr_pval_diff_police_maj_eth = police_isr_nv_maj$p.value[2]
+isr_pval_diff_police_maj_arab = police_isr_nv_maj$p.value[3]
+isr_pct_change_police_maj_eth = (police_isr_nv_maj$coefficients[2] /israel_mean_police_maj_whitejew)*100
+isr_pct_change_police_maj_arab = (police_isr_nv_maj$coefficients[3] /israel_mean_police_maj_whitejew)*100
+
+## Minority perceptions
+israel_mean_degv_min_whitejew = degv_isr_nv_min$coefficients[1]
+isr_mean_degv_min_eth = degv_isr_nv_min$coefficients[1] + degv_isr_nv_min$coefficients[2] 
+isr_mean_degv_min_arab = degv_isr_nv_min$coefficients[1] + degv_isr_nv_min$coefficients[3] 
+isr_diff_degv_min_eth = degv_isr_nv_min$coefficients[2]
+isr_diff_degv_min_arab = degv_isr_nv_min$coefficients[3]
+isr_pval_diff_degv_min_eth = degv_isr_nv_min$p.value[2]
+isr_pval_diff_degv_min_arab = degv_isr_nv_min$p.value[3]
+isr_pct_change_degv_min_eth = (degv_isr_nv_min$coefficients[2] /israel_mean_degv_min_whitejew)*100
+isr_pct_change_degv_min_arab = (degv_isr_nv_min$coefficients[3] /israel_mean_degv_min_whitejew)*100
+israel_mean_police_min_whitejew = police_isr_nv_min$coefficients[1]
+isr_mean_police_min_eth = police_isr_nv_min$coefficients[1] + police_isr_nv_min$coefficients[2] 
+isr_mean_police_min_arab = police_isr_nv_min$coefficients[1] + police_isr_nv_min$coefficients[3] 
+isr_diff_police_min_eth = police_isr_nv_min$coefficients[2]
+isr_diff_police_min_arab = police_isr_nv_min$coefficients[3]
+isr_pval_diff_police_min_eth = police_isr_nv_min$p.value[2]
+isr_pval_diff_police_min_arab = police_isr_nv_min$p.value[3]
+isr_pct_change_police_min_eth = (police_isr_nv_min$coefficients[2] /israel_mean_police_min_whitejew)*100
+isr_pct_change_police_min_arab = (police_isr_nv_min$coefficients[3] /israel_mean_police_min_whitejew)*100
+
+us_degv_maj = c(us_mean_degv_maj_white, us_mean_degv_maj_black, us_diff_degv_maj, us_pval_diff_degv_maj, us_pct_change_degv_maj)
+us_degv_min = c(us_mean_degv_min_white, us_mean_degv_min_black, us_diff_degv_min, us_pval_diff_degv_min, us_pct_change_degv_min)
+us_police_maj = c(us_mean_police_maj_white, us_mean_police_maj_black, us_diff_police_maj, us_pval_diff_police_maj, us_pct_change_police_maj)
+us_police_min = c(us_mean_police_min_white, us_mean_police_min_black, us_diff_police_min, us_pval_diff_police_min, us_pct_change_police_min)
+
+isr_degv_maj_eth = c(israel_mean_degv_maj_whitejew, isr_mean_degv_maj_eth, isr_diff_degv_maj_eth, isr_pval_diff_degv_maj_eth, isr_pct_change_degv_maj_eth)
+isr_degv_maj_arab = c(israel_mean_degv_maj_whitejew, isr_mean_degv_maj_arab, isr_diff_degv_maj_arab, isr_pval_diff_degv_maj_arab, isr_pct_change_degv_maj_arab)
+isr_degv_min_eth = c(israel_mean_degv_min_whitejew, isr_mean_degv_min_eth, isr_diff_degv_min_eth, isr_pval_diff_degv_min_eth, isr_pct_change_degv_min_eth)
+isr_degv_min_arab = c(israel_mean_degv_min_whitejew, isr_mean_degv_min_arab, isr_diff_degv_min_arab, isr_pval_diff_degv_min_arab, isr_pct_change_degv_min_arab)
+isr_police_maj_eth = c(israel_mean_police_maj_whitejew, isr_mean_police_maj_eth, isr_diff_police_maj_eth, isr_pval_diff_police_maj_eth, isr_pct_change_police_maj_eth)
+isr_police_maj_arab = c(israel_mean_police_maj_whitejew, isr_mean_police_maj_arab, isr_diff_police_maj_arab, isr_pval_diff_police_maj_arab, isr_pct_change_police_maj_arab)
+isr_police_min_eth = c(israel_mean_police_min_whitejew, isr_mean_police_min_eth, isr_diff_police_min_eth, isr_pval_diff_police_min_eth, isr_pct_change_police_min_eth)
+isr_police_min_arab = c(israel_mean_police_min_whitejew, isr_mean_police_min_arab, isr_diff_police_min_arab, isr_pval_diff_police_min_arab, isr_pct_change_police_min_arab)
+
+pct_change_tab = rbind(us_degv_maj, us_degv_min, us_police_maj, us_police_min, 
+                       isr_degv_maj_arab, isr_degv_min_arab, isr_police_maj_arab, isr_police_min_arab,
+                       isr_degv_maj_eth, isr_degv_min_eth, isr_police_maj_eth, isr_police_min_eth) 
+colnames(pct_change_tab) = c("Mean (majority)", "Mean (minority)", "Difference", "P-value", "Percent change")
+
+table_a15 = xtable(pct_change_tab, digits=2)
+print(table_a15, file="Tables/table_A16.tex")

23/replication_package/Code/replicate_table_A17.R

@@ -0,0 +1,15 @@
+##--##--##--##--##--##--##
+## Appendix: Table A17
+##--##--##--##--##--##--##
+
+## A. U.S. sample
+mod1 = lm(degree_violence ~ black + group_goal + commitment + age + female +education + income + pol_views + race, data=us_survey_wave2)
+mod2 = lm(recall_violence2 ~ black + group_goal + commitment + age + female +education + income + pol_views + race, data=us_survey_wave2)
+mod3 = lm(police_action_required ~ black + group_goal + commitment + age + female +education + income + pol_views + race, data=us_survey_wave2)
+stargazer(mod1, mod2, mod3, out="Tables/table_A17_panel_A.tex")
+
+## B. Israel sample
+mod1_israel = lm(degree_violence ~ identity_protesters + group_goal + commitment + age + female +education + income + ideology + ethnicity, data=isr_survey_wave2)
+mod2_israel = lm(recall_violence2 ~ identity_protesters + group_goal + commitment + age + female +education + income + ideology + ethnicity, data=isr_survey_wave2)
+mod3_israel = lm(police_action_required ~ identity_protesters + group_goal + commitment + age + female +education + income + ideology + ethnicity, data=isr_survey_wave2)
+stargazer(mod1_israel, mod2_israel, mod3_israel, out="Tables/table_A17_panel_B.tex")

23/replication_package/Code/replicate_table_A18.R

@@ -0,0 +1,11 @@
+##--##--##--##--##--##--##
+## Appendix: Table A18
+##--##--##--##--##--##--##
+
+generic_goal_dv = lm(degree_violence ~ identity_protesters + commitment , data=isr_survey_wave2[isr_survey_wave2$group_goal==0,])
+generic_goal_rv = lm(recall_violence2 ~ identity_protesters + commitment , data=isr_survey_wave2[isr_survey_wave2$group_goal==0,])
+generic_goal_paq = lm(police_action_required ~ identity_protesters + commitment , data=isr_survey_wave2[isr_survey_wave2$group_goal==0,])
+stargazer(generic_goal_dv, generic_goal_rv, generic_goal_paq,
+          covariate.labels = c("Ethiopian protesters", "Arab protesters",
+                               "Commitment to nonviolence", "Intercept: White protesters, generic goal, no commitment"),
+          out = "Tables/table_A18.tex")

23/replication_package/Code/replicate_table_A2.R

@@ -0,0 +1,14 @@
+##--##--##--##--##--##--##
+## Appendix: Table A2
+##--##--##--##--##--##--##
+
+nv_majority = prop.table(table(NAVCO2_EPR$success[NAVCO2_EPR$navco1designation==1 & NAVCO2_EPR$EPR_STATUS_EXCL==0]))
+v_majority = prop.table(table(NAVCO2_EPR$success[NAVCO2_EPR$navco1designation==0 & NAVCO2_EPR$EPR_STATUS_EXCL==0]))
+nv_minority = prop.table(table(NAVCO2_EPR$success[NAVCO2_EPR$navco1designation==1 & NAVCO2_EPR$EPR_STATUS_EXCL==1]))
+v_minority = prop.table(table(NAVCO2_EPR$success[NAVCO2_EPR$navco1designation==0 & NAVCO2_EPR$EPR_STATUS_EXCL==1]))
+
+navco_table = cbind(nv_majority, v_majority, nv_minority, v_minority)
+rownames(navco_table) = c("Campaign failure",   "Campaign success")
+
+table_a2 = xtable(navco_table, digits=2)
+print(table_a2, file="Tables/table_A2.tex")

23/replication_package/Code/replicate_table_A3.R

@@ -0,0 +1,20 @@
+##--##--##--##--##--##--##
+## Appendix: Table A3
+##--##--##--##--##--##--##
+
+tab1 = lm(success ~  EPR_STATUS_EXCL* INIT_NV_ONSET , data= EBCR_EPR_NAVCO2) 
+tab2 = lm(success ~  EPR_STATUS_EXCL *  INIT_NV_ONSET + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + NVYEARS +I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3), data= EBCR_EPR_NAVCO2) 
+tab3 = lm(success ~  EPR_STATUS_EXCL *  INIT_NV_ONSET + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2)  
+tab4 = lm(success ~  EPR_STATUS_ORD* INIT_NV_ONSET , data= EBCR_EPR_NAVCO2) 
+tab5 = lm(success ~  EPR_STATUS_ORD *  INIT_NV_ONSET + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + NVYEARS +I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3), data= EBCR_EPR_NAVCO2) 
+tab6 = lm(success ~  EPR_STATUS_ORD *  INIT_NV_ONSET + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2)  
+
+stargazer(tab4, tab5, tab6, tab1, tab2, tab3, table.placement="H", order = c(3,1,20, 2, 21),
+          covariate.labels = c("NV campaign", "EPR Status", "EPR Status × NV Campaign",
+                               "EPR Status: Excluded", "EPR Status: Excluded × NV Campaign",
+                               "EPR group size", "Country population", "Country GDP per capita",
+                               "Prior participation in NV", "Prior participation in V",
+                               "Level of democracy", "Physical integrity index", "Neighboring kin in power",
+                               "Group status downgraded", "Horizontal inequality"),
+          omit = c("VYEARS", "NVYEARS"),
+          out="Tables/table_A3.tex")

23/replication_package/Code/replicate_table_A4.R

@@ -0,0 +1,17 @@
+##--##--##--##--##--##--##
+## Appendix: Table A4
+##--##--##--##--##--##--##
+
+tab7 = lm(success ~  EPR_GROUPSIZE* INIT_NV_ONSET , data= EBCR_EPR_NAVCO2) 
+tab8 = lm(success ~  EPR_GROUPSIZE *  INIT_NV_ONSET  + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + NVYEARS +I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3), data= EBCR_EPR_NAVCO2) 
+tab9 = lm(success ~  EPR_GROUPSIZE *  INIT_NV_ONSET  + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2)  
+
+stargazer(tab7, tab8, tab9, table.placement="H", order = c(2,1,18),
+          covariate.labels = c("NV campaign", "EPR group size", "EPR group size × NV campaign",
+                               "Country population", "Country GDP per capita",
+                               "Prior participation in NV", "Prior participation in V",
+                               "Level of democracy", "Physical integrity index", 
+                               "Neighboring kin in power", "Group status downgraded",
+                               "Horizontal inequality"),
+          omit = c("VYEARS", "NVYEARS"),
+          out="Tables/table_A4.tex")

23/replication_package/Code/replicate_table_A5.R

@@ -0,0 +1,45 @@
+##--##--##--##--##--##--##
+## Appendix: Table A5
+##--##--##--##--##--##--##
+
+# Status (excluded, not excluded)
+
+nv_status = lm(success ~  EPR_STATUS_EXCL + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_NV_ONSET==1,])  
+v_status = lm(success ~  EPR_STATUS_EXCL + EPR_GROUPSIZE + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_V_ONSET==1,])  
+
+preds_nv = ggpredict(nv_status, terms = c("EPR_STATUS_EXCL"))
+preds_nv$tactic = "Non-violent"
+preds_v = ggpredict(v_status, terms = c("EPR_STATUS_EXCL"))
+preds_v$tactic = "Violent"
+
+preds_status = rbind(preds_nv, preds_v)
+preds_status$tactic <- factor(preds_status$tactic, levels = c("Violent", "Non-violent"))
+preds_status$Status = "Excluded"
+preds_status$Status[preds$x==0] = "Not excluded"
+preds_status$type = "status"
+
+## Size (above and below the mean of the distribution)
+
+EBCR_EPR_NAVCO2$small_size = NA
+EBCR_EPR_NAVCO2$small_size[EBCR_EPR_NAVCO2$EPR_GROUPSIZE < mean(EBCR_EPR_NAVCO2$EPR_GROUPSIZE, na.rm=T)] = 1
+EBCR_EPR_NAVCO2$small_size[EBCR_EPR_NAVCO2$EPR_GROUPSIZE >= mean(EBCR_EPR_NAVCO2$EPR_GROUPSIZE, na.rm=T)] = 0
+
+nv_size = lm(success ~  small_size  + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_NV_ONSET==1,])  
+v_size = lm(success ~  small_size  + POP_LOG_LAG_EXT + GDPPC_LOG_LAG_EXT + PASTNV + PASTV + VDEM_POLYARCHY_LAG + VDEM_PHYSINT_LAG + EPR_TEK_EGIP + EPR_DOWNGRADED5 + HORIZ_INEQ + NVYEARS + I(NVYEARS^2) + I(NVYEARS^3) + VYEARS + I(VYEARS^2) + I(VYEARS^3) , data= EBCR_EPR_NAVCO2[EBCR_EPR_NAVCO2$INIT_V_ONSET==1,])  
+
+preds_nv = ggpredict(nv_size, terms = c("small_size"))
+preds_nv$tactic = "Non-violent"
+preds_v = ggpredict(v_size, terms = c("small_size"))
+preds_v$tactic = "Violent"
+
+preds_size = rbind(preds_nv, preds_v)
+preds_size$tactic <- factor(preds_size$tactic, levels = c("Violent", "Non-violent"))
+preds_size$Status = "Group size < mean"
+preds_size$Status[preds_size$x==0] = "Group size >= mean"
+preds_size$type = "size"
+
+preds_all = as.data.frame(rbind(preds_status, preds_size))
+preds_all = preds_all[,c("predicted", "std.error", "conf.low", "conf.high",
+                         "tactic", "Status")]
+table_a5 = xtable(preds_all, digits=2)
+print(table_a5, file="Tables/table_A5.tex")

23/replication_package/Code/replicate_table_A6.R

@@ -0,0 +1,11 @@
+##--##--##--##--##--##--##
+## Appendix: Table A6
+##--##--##--##--##--##--##
+
+pol_irr1 = lm(success ~  navco1designation , 
+              data= NAVCO2[!(NAVCO2$id %in% NAVCO2_EPR$id),])
+pol_irr2 = lm(success ~  navco1designation , 
+              data= NAVCO2[NAVCO2$id %in% NAVCO2_EPR$id[NAVCO2_EPR$EPR_STATUS == "IRRELEVANT"],])
+
+stargazer(pol_irr1, pol_irr2, covariate.labels = c("Nonviolent campaign", "Constant"),
+          out="Tables/table_A6.tex")

23/replication_package/Code/replicate_table_A7.R

@@ -0,0 +1,8 @@
+##--##--##--##--##--##--##
+## Appendix: Table A7
+##--##--##--##--##--##--##
+
+us_wave1_sumstats = us_survey_wave1[,c("age", "female", "education", "income", "partyID", "ideology", "interest_politics",
+                                      "degree_violence", "police_action_required", "recall_violence2", "race")]
+us_wave1_sumstats = dummy_cols(us_wave1_sumstats, select_columns = c("education", "income", "partyID", "ideology", "race"))
+stargazer(us_wave1_sumstats, omit.summary.stat=c("p25", "p75"), out="Tables/table_A7.tex")

23/replication_package/Code/replicate_table_A8.R

@@ -0,0 +1,10 @@
+##--##--##--##--##--##--##
+## Appendix: Table A8
+##--##--##--##--##--##--##
+
+isr_wave1_jewish = isr_survey_wave1[isr_survey_wave1$survey=="isr",]
+isr_wave1_jewish_sumstats = isr_wave1_jewish[,c("age", "female", "religiosity", "ethnicity", "education", "income", "partyID", "ideology", "interest_news",
+                                                "degree_violence", "police_action_required", "recall_violence2")]
+isr_wave1_jewish_sumstats = dummy_cols(isr_wave1_jewish_sumstats, select_columns = c("age", "religiosity", "ethnicity", "education", "income", "partyID"))
+table_a8 = stargazer(isr_wave1_jewish_sumstats, omit.summary.stat=c("p25", "p75"))
+cat(table_a8, sep = '\n', file = "Tables/table_A8.tex")

23/replication_package/Code/replicate_table_A9.R

@@ -0,0 +1,11 @@
+##--##--##--##--##--##--##
+## Appendix: Table A9
+##--##--##--##--##--##--##
+
+isr_wave1_arab = isr_survey_wave1[isr_survey_wave1$survey=="isr_ar",]
+isr_wave1_arab_sumstats = isr_wave1_arab[,c("age", "female", "religiosity", "ethnicity", "education", "income", "partyID", "ideology", "interest_news",
+                                            "degree_violence", "police_action_required", "recall_violence2")]
+isr_wave1_arab_sumstats = dummy_cols(isr_wave1_arab_sumstats, select_columns = c("age", "religiosity", "ethnicity", "education", "income", "partyID"))
+table_a9 = stargazer(isr_wave1_arab_sumstats, omit.summary.stat=c("p25", "p75"))
+cat(table_a9, sep = '\n', file = "Tables/table_A9.tex")
+

23/replication_package/Data/EBCR_EPR_NAVCO2.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75451734eeb300dd765549b181cb21c1d4c91c9b0702562c19e1094898fa9764
+size 49172

23/replication_package/Data/NAVCO2.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6722c3f6703ec017e192f3b5bab9e7125b705c1bca159dffb55c7a944ca7260
+size 28901

23/replication_package/Data/NAVCO2_EPR.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f10808bec84e5e5d50b30f973373892b0ef570903808560cfcda645babffb6ea
+size 8510

23/replication_package/Data/isr_survey_text_analysis_arab.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f11a984db34d2bd6549168ffae4f373eed2c0ebcf8c49d52118ced6ecf986392
+size 791146

23/replication_package/Data/isr_survey_text_analysis_eth.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45c2f8f014182cef1752de744d235802ac03fda2410afe47e2571180bb7ffbbb
+size 844628

23/replication_package/Data/isr_survey_wave1.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:800e60783ecf8e9fac913544d37489598f2b767cf17d4f09c29badc851889138
+size 81215

23/replication_package/Data/isr_survey_wave2.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bee0683b7b5aa05998fad0bad2a6f9ca82b3c2729f640898f8af7e0f89184c1
+size 92753

23/replication_package/Data/us_survey_text_analysis.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b03739389a1428c180e0e3929d037876bf34c4d893167054dac97ce5b2d01c9
+size 686955

23/replication_package/Data/us_survey_wave1.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d02a0157e3113d52e5429679e908d35e2ece406fa07fedc57b7a8289df587a46
+size 47487

23/replication_package/Data/us_survey_wave2.rdata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac163b849b38a58dac943fc0954f31127da273d8f2e247f5ca323c523a5930e4
+size 94383

23/replication_package/Figures/fig_1.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fe315acebbd890225496f07ebc7b1b292fbf2b4d02bd1369e487b623132d66f
+size 5597