Reg_assignment.R

dd <- read.csv("progresa.csv")
dim(dd)

dd <- na.omit(dd)
dim(dd)

### QUESTION 1
# An example of what I'm looking for

par(mfrow = c(2,1))
hist(dd$votos1994[dd$treatment==1], col = "red",
     main = "Distribution of Votos 1995 for the treatment group", 
     xlab = "Votos 1995", xlim = c(0, 1500), breaks = 25)

hist(dd$votos1994[dd$treatment==0], col = "blue",
     main = "Distribution of Votos 1995 for the control group", 
     xlab = "Votos 1995", xlim = c(0, 1500), breaks = 25)


### QUESTION 2
# there was a typo in the question... it referred to
#"prd1994" but it should have said "prd1994v"
# so either reg1 or reg2 is fine

reg1 <- lm(t2000r ~ treatment + avgpoverty + I(log(pobtot1994))
           + t1994r + pri1994v + pan1994v + prd1994,
           data = dd)

summary(reg1)

reg2 <- lm(t2000r ~ treatment + avgpoverty + I(log(pobtot1994))
           + t1994r + pri1994v + pan1994v + prd1994v,
           data = dd)

summary(reg2)

# Question 2 question also suggested an additional regression with an 
# addtional outcome variable:
# pri2000s -- so some students may have ALSO run either this regression:

### reg3 <- 
### lm(pri2000v ~ treatment + avgpoverty + I(log(pobtot1994)) + 
### t1994r + pri1994v + pan1994v + prd1994, data = dd)

#or this one

### reg4 <- 
### lm(pri2000v ~ treatment + avgpoverty + I(log(pobtot1994)) + 
### t1994r + pri1994v + pan1994v + prd1994v, data = dd)

# All the estimates are rather small and do not provide strong
# support for the hypothesis that the CCT program increases turnout and
# support for the incumbent party.

#########