issueirt: Measuring Issue Specific Ideal Points

Author:

• Sooahn Shin

Reference:

• Shin, Sooahn (2024). “Measuring Issue Specific Ideal Points from Roll Call Votes.” Working Paper.
  • Please cite the paper when using the package.

Installation

You can install the development version of issueirt with:

# install.packages("devtools")
devtools::install_github("sooahnshin/issueirt", build_vignettes = TRUE, dependencies = TRUE)
library(issueirt)
vignette("using_issueirt")
vignette("dynamic_issueirt")

Example: U.S. House of Representatives (1891 - 1896)

Check vignette("dynamic_issueirt") for more details.

Visualization of Issue Specific Axis

library(issueirt)
p_ls <- plot_issueaxis(stan_input        = issueirt:::dynamic_stan_input, 
                       posterior_summary = issueirt:::posterior_summary_pp,
                       group             = issueirt:::posterior_df_pp$group, 
                       p.title           = "IssueIRT",
                       breaks.group      = c("Southern Democrat", "Republican", "Democrat", 
                                             "Ind. Democrat", "Populist", "Silver"), 
                       values.shape      = c(19, 15, 19, 19, 17, 17), 
                       values.color      = c("#C4B0A6", "#D1A4A2", "#B0AECF", 
                                             "#B8B5CF", "#C2D7C1", "#777777"),
                       plot.each.congress = TRUE)

library(patchwork)
library(ggplot2)
theme_set(theme_classic(base_size = 15) + 
          theme(legend.position = "bottom", plot.title = element_text(hjust = 0.5)))

## Monetary Policy Axis: 52nd to 54th H.R. (1891 - 1896)
(p_ls[["H52_Monetary"]]   + labs(title = "52nd H.R.") + theme(legend.position = "none")) +
  (p_ls[["H53_Monetary"]] + labs(title = "53rd H.R.") + theme(legend.position = "bottom")) +
  (p_ls[["H54_Monetary"]] + labs(title = "54th H.R.") + theme(legend.position = "none")) +
  plot_layout(ncol = 3, guides = "collect") &
  plot_annotation(title = "Monetary Policy Axis")

• Points: multidimensional ideal points
  • small light grey: legislators that were not present in that term
• Grey lines: monetary policy related roll-call axes
• Black line: monetary policy axis

IssueIRT is a hierarchical Item Response Theory (IRT) model that estimates an issue-specific axis representing a continuum extending from left to right positions on the issue using (1) roll-call votes and (2) user-supplied issue labels. We can project ideal points on the issue-specific axis to generate single-dimensional, issue-specific ideal points as visualized below.

Visualization of Issue Specific Ideal Points

issue_irt <- get_ideal_points(stan_fit    = issueirt:::fit_dynamic, 
                              issue_label = issueirt:::dynamic_stan_input$misc$issue_code_df$codebook$label,
                              legis_label = issueirt:::dynamic_stan_input$misc$legis_term, 
                              legis_group = issueirt:::posterior_df_pp$group)
monetary_irt <- issue_irt |> 
                dplyr::filter(issue_label %in% c("H52_Monetary", "H53_Monetary", "H54_Monetary")) |>
                dplyr::mutate(issue_index = as.character(issue_index),
                              legis_index = as.character(legis_index)) |>
                dplyr::mutate(dplyr::across(dplyr::where(is.numeric), ~ - .x)) |>
                dplyr::mutate(issue_label = stringr::str_remove(issue_label, "_Monetary"))

## Monetary Policy Ideal Points: 52nd to 54th H.R. (1891 - 1896)
plot_issueirt(issueirt      = monetary_irt, 
              p.title       = "Monetary Policy Specific Ideal Points",
              breaks.group  = c("Southern Democrat", "Republican", "Democrat", 
                                "Ind. Democrat", "Populist", "Silver"), 
              values.shape  = c(19, 15, 19, 19, 17, 17), 
              values.color  = c("#e36934", "#DE0100", "#1405BD", "#635FBD", "#62bd6d", "#777777"),
              alpha         = 0.5, 
              size          = 3) +
  labs(y = "Congress")