Communication is a central component of representation (Mansbridge, 2003; Disch, 2012). Legislators invest time and resources in crafting speeches in Congress, composing press releases to send to newspapers, and distributing messages directly to their constituents (Yiannakis, 1982; Quinn et al., 2010; Lipinski, 2004; Grimmer, 2013). Indeed, the primary problem in studying the role of communication in representation is that legislators communicate so much that analysts are quickly overwhelmed. Traditional hand-coding is simply unable to keep pace with the staggering amount of text that members of Congress produce each year.
In this chapter I use a text as data method and a collection of press releases to measure how legislators present their work to constituents (Grimmer and Stewart, 2013). Specifically, I measure legislators’ expressed priorities: the attention they allocate to topics and issues when communicating with constituents (Grimmer, 2010). Using the measures of legislators’ expressed priorities, I characterize how Republicans respond to the drastic change in institutional and electoral context after the 2008 election. Not only did the Republican party lose the White House but also the Tea Party movement mobilized and articulated conservative objections to particularistic spending. Replicating a finding from Grimmer, Westwood, and Messing (2014) with alternative measures, I show that Republicans abandon credit claiming. Instead, Republicans articulate criticisms of the Democratic party, the Obama administration, and Democratic policy proposals. In contrast, Democrats embrace credit claiming and defend Democratic policies – though less vocally than Republicans criticize those same proposals. In spite of the shifts in rhetoric, though, I demonstrate that there is a strong year-to-year relationship in legislators’ presentational styles. So, although legislators are responsive at the margin to changing conditions, the basic strategy remains the same.
This chapter contributes to a growing literature that examines legislative speech using automated methods for text analysis (Hillard, Purpura, and Wilkerson, 2008; Monroe, Colaresi, and Quinn, 2008; Quinn et al., 2010; Grimmer, 2013; Cormack, 2014). This literature has demonstrated how computational tools can be successfully used to examine the content of legislation and how the types of bills passed over time have changed (Adler and Wilkerson, 2012). Other studies have demonstrated how text analysis can be used to provide nuanced measures of legislators’ ideal points (Gerrish and Blei, 2012). Still other studies have demonstrated how legislators use communication to create an impression of influence over expenditures (Grimmer, Westwood, and Messing, 2014).