voter_id_rerun.R

install.packages("needs")
library(needs)
needs(sjlabelled, tidyverse, haven, magrittr, ggrepel, sf)

# Load in data from all waves
USoc_indresp_1 <- read_dta("/path to wave 1/", encoding = "latin1")
USoc_indresp_2 <- read_dta("/path to wave 2/", encoding = "latin1")
etc

# Join all waves together, keeping the most recent driving licence record for each respondent, and their vote in the last election
USoc_indresp_10 %>%
  dplyr::select(pidp, j_drive, j_vote8, j_indinub_xw) %>%
  left_join(USoc_indresp_1 %>% dplyr::select(pidp, a_drive)) %>%
  left_join(USoc_indresp_2 %>% dplyr::select(pidp, b_drive)) %>%
  left_join(USoc_indresp_3 %>% dplyr::select(pidp, c_drive)) %>%
  left_join(USoc_indresp_4 %>% dplyr::select(pidp, d_drive)) %>%
  left_join(USoc_indresp_5 %>% dplyr::select(pidp, e_drive)) %>%
  left_join(USoc_indresp_6 %>% dplyr::select(pidp, f_drive)) %>%
  left_join(USoc_indresp_7 %>% dplyr::select(pidp, g_drive)) %>%
  left_join(USoc_indresp_8 %>% dplyr::select(pidp, h_drive)) %>%
  left_join(USoc_indresp_9 %>% dplyr::select(pidp, i_drive)) %>%
  mutate(j_vote8 = as_character(j_vote8)) %>%
  gather(wave, drive, c(2,5:ncol(.))) %>%
  group_by(pidp, j_indinub_xw, j_vote8) %>%
  summarise(drive = ifelse(1 %in% drive, 1, 2)) %>%
  ungroup() %>%
  # Weighted counts of vote by driving licence status:
  count(drive, j_vote8, wt = j_indinub_xw) %>% 
  filter(drive > 0) %>% 
  mutate(drive = c("Yes", "No")[drive]) %>% 
  # Get rid of records with no valid answer to the voting question:
  filter(!j_vote8 %in% c("inapplicable", "refusal")) %>%
  group_by(drive) %>%
  # Calculate the voting patterns of licence-holders and non-licence-holders:
  mutate(share = n/sum(n)*100) %>%
  View