Permalink
Cannot retrieve contributors at this time
Fetching contributors…
Cannot retrieve contributors at this time
| #The DD data has some missing values that I've replaced manually. | |
| #un_region_name for Tuvalu should be Polynesia, and un_continent_name should be Oceania; | |
| #un_continent_name for Taiwan should be Asia, and un_region_name should be Eastern Asia | |
| #North and South Vietnam do not have leadership info | |
| #I've also generated a new variable, edate2, from edate. edate2 is the year in which the | |
| #effective head of government took power. | |
| #Reads the manually modified dataset | |
| dd <- read.csv("ddrevisited-data-v1.csv") | |
| #Create factors for easier use | |
| dd$democracy <- factor(dd$democracy,labels=c("Non-democracy","Democracy")) | |
| dd <- transform(dd, regime2 = ifelse(regime < 3, 1, regime)) | |
| dd$regime <- factor(dd$regime,labels=c("Parliamentary Dem","Mixed Dem","Presidential Dem","Civilian Dict","Military Dict","Monarchy")) | |
| dd$regime2 <- factor(dd$regime2,labels=c("Democracy","Civilian Dict","Military Dict","Monarchy")) | |
| dd$comm <- factor(dd$comm,labels=c("Non-communist","Communist")) | |
| #find leadership transitions; this code is slower than it needs to be, I'm sure there are better ways | |
| a <- 2:length(dd$ehead) #a counter variable | |
| b <- 1 #another counter | |
| leaderspellnum <- rep(1,length(dd$ehead)) | |
| for(i in a) { | |
| if (dd$ehead[i-1] != dd$ehead[i]) | |
| { | |
| b <- b + 1 | |
| leaderspellnum[i:length(leaderspellnum)]<-b | |
| } | |
| } | |
| dd <- cbind(dd,leaderspellnum) | |
| #find regime transitions | |
| a <- 2:length(dd$regime) #a counter variable | |
| b <- 1 #another counter | |
| regimespellnum <- rep(1,length(dd$regime)) | |
| for(i in a) { | |
| if (dd$regime[i-1] != dd$regime[i] || dd$ctryname[i-1]!=dd$ctryname[i]) { | |
| b <- b + 1 | |
| regimespellnum[i:length(regimespellnum)]<-b | |
| } | |
| } | |
| dd <- cbind(dd,regimespellnum) | |
| #find democracy transitions | |
| a <- 2:length(dd$democracy) #a counter variable | |
| b <- 1 #another counter | |
| democracyspellnum <- rep(1,length(dd$democracy)) | |
| for(i in a) { | |
| if (dd$democracy[i-1] != dd$democracy[i] || dd$ctryname[i-1]!=dd$ctryname[i]) { | |
| b <- b + 1 | |
| democracyspellnum[i:length(democracyspellnum)]<-b | |
| } | |
| } | |
| dd <- cbind(dd,democracyspellnum) | |
| #fill in the missing dates of entry of leaders and adjust tenure | |
| library(plyr) | |
| #Create a "leader spell" identifier for each leader/democracy/entry date triad | |
| dd <- ddply(dd,.(leaderspellnum,democracy,ctryname),transform,leaderspelldem = paste(ehead,ctryname,min(year),max(year),democracy,sep=".")) | |
| dd <- dd[ order(dd$order), ] | |
| #Add a variable for the end date of the leader spell | |
| dd <- ddply(dd,.(leaderspelldem),transform,enddateleaderdem = max(year)) | |
| dd <- dd[ order(dd$order), ] | |
| #Add a variable for the beginning date of the leader spell | |
| dd <- ddply(dd,.(leaderspelldem),transform,begindateleaderdem = ifelse(enddateleaderdem-tenure08+1 < entryy, enddateleaderdem-tenure08+1, min(year))) | |
| dd <- dd[ order(dd$order), ] | |
| #Create a new tenure variable for the democratic leader spell | |
| dd <- transform(dd, tenuredem = enddateleaderdem - begindateleaderdem + 1) | |
| #Create a "leader spell" identifier for each leader/regime/entry date triad | |
| dd <- ddply(dd,.(leaderspellnum,regime,ctryname),transform,leaderspellreg = paste(ehead,ctryname,min(year),max(year),regime,sep=".")) | |
| dd <- dd[ order(dd$order), ] | |
| #Same thing for regime2 | |
| dd <- ddply(dd,.(leaderspellnum,regime2,ctryname),transform,leaderspellreg2 = paste(ehead,ctryname,min(year),max(year),regime2,sep=".")) | |
| dd <- dd[ order(dd$order), ] | |
| #Create tenure variables for leader spells per regime | |
| dd <- ddply(dd,.(leaderspellreg),transform,tenurereg = length(year)) | |
| dd <- dd[ order(dd$order), ] | |
| dd <- ddply(dd,.(leaderspellreg2),transform,tenurereg2 = length(year)) | |
| dd <- dd[ order(dd$order), ] | |
| #Create left-censoring indicator | |
| dd <- transform(dd, leftcensored = ifelse(edate2 < entryy,TRUE, FALSE)) | |
| #Add a variable for the entry date of the leader spell for each regime type | |
| dd <- ddply(dd,.(leaderspellreg),transform,edatereg = min(year)) | |
| dd <- dd[ order(dd$order), ] | |
| dd <- ddply(dd,.(leaderspellreg2),transform,edatereg2 = min(year)) | |
| dd <- dd[ order(dd$order), ] | |
| #set up the right-censoring indicator | |
| dd <- ddply(dd,.(leaderspelldem),transform,rightcensdem = min(ecens08,na.rm=TRUE)) | |
| dd <- ddply(dd,.(leaderspellreg),transform,rightcensreg = min(ecens08,na.rm=TRUE)) | |
| dd <- dd[ order(dd$order), ] | |
| #Now we create the sample files | |
| ddsurvival.cols <- c("ctryname","cowcode2","politycode","un_region_name","un_continent_name", | |
| "ehead","democracy","begindateleaderdem","enddateleaderdem","tenuredem","rightcensdem") | |
| ddsurvival <- unique(dd[,ddsurvival.cols]) | |
| rownames(ddsurvival) <- 1:nrow(ddsurvival) | |
| ddsurvival2.cols <- c("ctryname","cowcode2","politycode","un_region_name","un_continent_name", | |
| "ehead","leaderspellreg","democracy","regime","edatereg","tenurereg","rightcensreg") | |
| ddsurvival2 <- unique(dd[,ddsurvival2.cols]) | |
| rownames(ddsurvival2) <- 1:nrow(ddsurvival2) | |
| write.csv(ddsurvival,"ddsurvival.csv") | |
| write.csv(ddsurvival2,"ddsurvival2.csv") | |
| #to create ddcoldwar.csv, we need to split the cases that "straddle" 1989, which we take as the end of the cold war. | |
| #There is probably a much easier and faster way to do this | |
| ddcoldwar.cols <- c("ctryname","cowcode2","politycode","un_region_name","un_continent_name", | |
| "ehead","democracy","leaderspelldem","begindateleaderdem","enddateleaderdem","tenuredem","rightcensdem") | |
| ddsurvival3 <- unique(dd[,ddcoldwar.cols]) | |
| ddsurvival.coldwar <- subset(ddsurvival3,enddateleaderdem <= 1989 & begindateleaderdem > 1945) | |
| ddsurvival.notcoldwar <- subset(ddsurvival3,begindateleaderdem >= 1989) | |
| ddsurvival.coldwar <- transform(ddsurvival.coldwar,coldwar=TRUE) | |
| ddsurvival.notcoldwar <- transform(ddsurvival.notcoldwar,coldwar=FALSE) | |
| straddle.1989 <- subset(ddsurvival3,enddateleaderdem > 1989 & begindateleaderdem < 1989) | |
| straddle.1945 <- subset(ddsurvival3,enddateleaderdem > 1945 & begindateleaderdem < 1946) | |
| before.1989 <- transform(straddle.1989,enddateleaderdem=1989, tenuredem = 1989 - begindateleaderdem + 1, coldwar=TRUE) | |
| after.1989 <- transform(straddle.1989,begindateleaderdem=1990, tenuredem = enddateleaderdem - 1989, coldwar=FALSE) | |
| before.1945 <- transform(straddle.1945,enddateleaderdem=1945, tenuredem = 1945 - begindateleaderdem + 1, coldwar=FALSE) | |
| after.1945 <- transform(straddle.1945,begindateleaderdem=1946, tenuredem = enddateleaderdem - 1945, coldwar=TRUE) | |
| ddcoldwar <- rbind(ddsurvival.coldwar,ddsurvival.notcoldwar,before.1945,after.1945,before.1989,after.1989) | |
| ddcoldwar <- ddcoldwar[ order(ddcoldwar$ctryname,ddcoldwar$begindateleaderdem), ] | |
| rownames(ddcoldwar) <- 1:nrow(ddcoldwar) | |
| write.csv(ddcoldwar,"ddcoldwar.csv") |