Not the most elegant, albeit still fairly "tidy" script to make Freedom House's "Freedom of the Press" data actually usable.

tidy-freedom-of-the-press-data.R

library(tidyverse)
library(stringr)

# Freedom of the Press data ---------
# Oh god, Freedom House, why the sam hell do you do these things...

FP <- readxl::read_excel("~/Dropbox/data/freedom-house/FOTP1980-FOTP2017_Public-Data.xlsx", sheet=2) %>% 
  repair_names()

# Ugh, all right. Let's do this...
FP <- slice(FP, -1:-3)

colnames(FP) <- sapply(FP[1,], as.character)

colnames(FP) <- str_c(colnames(FP), seq(1,length(colnames(FP))))

colnames(FP)[1] <- "country"
FP <- slice(FP, -1)

# I'm assuming you're interested in the "Total Score"...
FP %>%
  select(country,contains("Total")) -> FP

# I'm still not the most fluent in sapply within tidyverse framework. Comments here welcome.
FP[,2:ncol(FP)] <- sapply(FP[,2:ncol(FP)],
                                function(x)ifelse(x == "-",NA, as.numeric(x)))

# Total Scores start for the 1993 coverage year (i.e. the 1994 report).
colnames(FP)[2:dim(FP)[2]] <- seq(1993,1993+dim(FP)[2]-2)

FP %>%
  gather(year, fh, `1993`:`2016`) %>%
  mutate(year = as.numeric(year)) -> FP