R Packages

hansard

mnis

nomisr

parlitools

What do MPs talk about?

Research question

Do female Labour MPs selected through all women shortlists differ from other female Labour MPs in how they speak or what they speak about?

Tools

• R packages: quanteda, stm

• Hansard data (from the 1997 – 2017 General Elections)

Labour vs Conservatives

Labour MPs

Female Labour MPs

Topic Models

The hansard package

Early Day Motions

library(hansard)
library(ggplot2)
library(dplyr)

edm <- early_day_motions(start_date = "2010-01-01")

edm_sum <- edm %>% group_by(number_of_signatures) %>% summarise(n=n())

theme_set(theme_bw())

p1 <- ggplot(edm_sum) + 
  geom_line(aes(x = number_of_signatures, y = n, colour = n), 
            size = 1.1, alpha = 0.85) + 
  scale_y_sqrt(breaks = c(1,5,10,20,50,100,200,300)) + 
  scale_x_sqrt(breaks = c(1,5,10,20,50,100,200,300)) + 
  scale_colour_viridis_c() + 
  theme(legend.position = "none") + 
  labs(x = "Number of Signatures", y = "Number of Early Day Motions")

p1

EDM Signatures

EDM Title Wordcloud Code

library(quanteda)

title_dfm <- edm$title %>%
  tokens(remove_punct = TRUE, remove_numbers =TRUE) %>% 
  dfm(remove = c(stopwords()))

textplot_wordcloud(title_dfm, max_words = 350, rotation=0,
                   color = viridis::viridis_pal(option = "C")(30))

EDM Title Wordcloud

parlitools

library(parlitools)
library(sf)
library(cartogram)
library(leaflet)

seats19 <- parlitools::bes_2019 %>% 
  left_join(parlitools::party_colour, by = c("winner_19" ="party_name")) 

gb_hex_map <- parlitools::west_hex_map %>% 
  left_join(seats19, by = c("gss_code"="ons_const_id"))

gb_hex_map <- gb_hex_map %>% 
  mutate(majority_19_perc = (majority_19/total_vote_19)*100)

gb_hex_map <- st_transform(gb_hex_map, "+init=epsg:3395")

gb_hex_scaled <- cartogram_cont(gb_hex_map, 'majority_19_perc', itermax = 5)

gb_hex_scaled <- st_transform(gb_hex_scaled, "+init=epsg:4326")

# Creating map labels
labels <- paste0(
  "Constituency: ", gb_hex_scaled$constituency_name.y, "</br>",
  "Winner: ", gb_hex_scaled$winner_19, "</br>",
  "Majority (votes): ", gb_hex_scaled$majority_19,"</br>",
  "Majority (% points): ", round(gb_hex_scaled$majority_19_perc,1),"</br>",
  "Turnout: ", round(gb_hex_scaled$turnout_19,1), "%"
) %>% lapply(htmltools::HTML)

leaflet(options=leafletOptions(
  dragging = FALSE, zoomControl = FALSE, tap = FALSE,
  minZoom = 6, maxZoom = 6, maxBounds = list(list(2.5,-7.75),list(58.25,50.0)),
  attributionControl = FALSE),
  gb_hex_scaled) %>%
  addPolygons(
    color = "lightgrey",
    weight=0.75,
    opacity = 0.5,
    fillOpacity = 1,
    fillColor = ~party_colour,
    label=labels) %>% 
  htmlwidgets::onRender(
    "function(x, y) {
        var myMap = this;
        myMap._container.style['background'] = '#fff';
    }")%>% 
  mapOptions(zoomToLimits = "first")

How safe are seats?

Notes

• Documentation on all this is available at https://docs.evanodell.com

• My version of Hansard speeches is at: https://evanodell.com/datasets/hansard-data/