By the way you have structured your data, it seems to me that you won't get a real yearly proportion until you summarize by year. If you go prop.table(table-ing rigth away, all you'll find is what each record represent to the whole total (either by total, row or column, as in @Robert's answer).
To summaryze by year:
library(data.table)
setDT(CRDATA)
CRDATA[, sum(Arrest), by = Year][, .(Year, prop.table(V1))]
The first expression in [] totalizes by year, the second one calculates the proportion and displays it along the year.
########## EDIT ########
A dplyr alternative:
library(dplyr)
CRDATA %>% group_by(YEAR) %>% summarize(arr = sum(Arrest)) %>% mutate(prop=arr/sum(arr))
