ggparliament: Parliament Plots

title: 'ggparliament: A ggplot2 extension for parliament plots in R' tags: - government statistics - data visualization - social science - political science authors: - name: Zoe Meers affiliation: 1 orcid: 0000-0001-8045-6531 - name: Robert Hickman affiliation: 2 - name: Thomas J. Leeper affiliation: 3 orcid: 0000-0003-4097-6326 affiliations: - name: United States Studies Centre, University of Sydney index: 1 - name: Department of Physiology, Development and Neuroscience, University of Cambridge index: 2 - name: Department of Methodology, London School of Economics and Political Science index: 3 date: 12 February 2018 bibliography: paper.bib

Summary

ggparliament is a R [@R] package for visualizing legislative chambers. It allows for easy visualization of the party composition of a legislative body or for plotting election results using syntax from the tidyverse, a set of packages for cleaning, modelling, and visualizing data [@tidyverse]. The package is one of the first ggplot2 [@ggplot2] extensions for visualizing political data. It is considerably more detailed than most visualization tools for graphing legislatures in that it is designed to handle many different legislatures irrespective of size, number of parties, or characteristics of legislatures. Legislature visualizations are also known as parliament plots or parliament charts.

Visualizing legislatures can be done in a variety of formats - usually in a JavaScript framework. ggparliament seeks to implement the flexibility from JavaScript in R. R users will find they can use a similar syntax to the popular ggplot2 package to generate a visualization of a legislative body. In particular, ggparliament is a R package because R is a statistical computing language that many quantitative political scientists use in research. (See pscl [@pscl], Amelia [@Amelia], MatchIt [@MatchIt], and margins [@margins] for an incomplete list of R packages created primarily for political science research.)

ggparliament is useful research tool for a variety of social science disciplines, including quantitative political science. It is particularly beneficial for political scientists who research political institutions, such as electoral systems, party politics, or legislative politics. ggparliament provides several layouts, representing different legislative chambers e.g. the United Kingdom's House of Commons, Australia's horseshoe-shaped parliament, or the widely-used semicircle legislative chamber.

The standard syntax for a ggparliament plot is as follows:

us_data <- ggparliament::election_data %>% 
  filter(country == "USA" & year == 2016 & 
           house == "Representatives") %>% 
  parliament_data(election_data = ., 
            parl_rows = 8, 
            party_seats = .$seats,
            type = 'semicircle')

ggplot(us_data, aes(x,y), colour = "black") + 
  geom_parliament_seats(size = 0.8) + 
  theme_ggparliament()

ggparliament has been designed with several key features in mind. It provides multiple layouts for different parliamentary systems, catering to researchers of various political systems. ggparliament items can be more complex than what we have detailed here. For example, it is possible to facet the ggparliament object over numerous elections or parliaments; this work is further explained in the package documentation.

In this paper we present a R package to plot legislatures, also known as parliament plots. Parliament plots provide summary statistics for the distribution of a legislature, allowing users to view the number of seats per party. In addition, ggparliament visualizes descriptive data about legislators or legislative districts as well as election results.