Cognitive load and issue engagement in congressional discourse. (August 2017)

• Record Type: Journal Article
• Authors: Shaffer, Robert
• Is Part Of: Cognitive systems research. Volume 44(2017)
• Page Start: 89
• Page End: 99
• Publication Date: 2017-08
• Subjects: Text-as-data -- Public policy -- Congress -- Financial crisis; Cognition -- Periodicals; Cognitive engineering (System design) -- Periodicals; Artificial intelligence -- Periodicals; 153.05
• DOI: 10.1016/j.cogsys.2017.03.006
• Languages: English
• ISSNs: 1389-0417

Abstract

Abstract: Like all human actors, politicians possess limited cognitive capacity. In ordinary interactions, this limitation discourages political decision-makers from addressing high-dimensional policy problems unless incentivized to do so by exogenous "focusing events." Public policy researchers have documented this pattern extensively, and have argued that cognitive constraints help explain the "stick–slip" dynamics that characterize macro-level policymaking. However, data and measurement limitations have prevented these studies from examining individual-level information processing patterns. In this paper, I develop a text-based approach designed to measure diversity of attention at an individual level, which I apply to an original dataset of Congressional hearing transcripts surrounding the 2008–2009 Financial Crisis. I find that individual speakers engaged with a more diverse set of topics during the crisis than before its onset, and became more focused as the crisis subsided.