Search
Browse By Day
Browse By Time
Browse By Person
Browse By Mini-Conference
Browse By Division
Browse By Session or Event Type
Browse Sessions by Fields of Interest
Browse Papers by Fields of Interest
Search Tips
Virtual Exhibit Hall
Change Preferences / Time Zone
Sign In
X (Twitter)
Building on widely accepted forecasting models, this paper is an exploratory and preliminary attempt to develop and employ a cultural variable (the “partisan fashion index,” or PFI) to forecast the incumbent party’s two-party share of the US presidential vote from 1976-2020 (11 elections). (This paper builds on previous work that was to be presented at the APSA in 2012, but because of the hurricane that year, never was presented.) The reason for incorporating a cultural variable that relies upon fashion trends is that unlike second quarter GDP growth or presidential approval (or disapproval), fashion trends are by definition prospective as opposed to retrospective indicators. Researchers generally do not associate fashion trends with political decision-making. Yet, we argue that fashion norms provide a window into voters’ social preferences which they exhibit in subsequent voting behavior. To estimate our forecast model, we construct a variable operationalizing fashion trends to provide a unique way to investigate the social identities of partisans and the relative, emerging appeal of the major parties in electorate. Thinking about the intertwining of culture and presidential forecasts also provides an avenue into a larger analysis/understanding of how the cultural and societal landscape is scattered with indicators about politics that are often ignored or left unengaged when they may tell us much about political dynamics and voters. We test our models using Ordinary Least Squares (OLS) regression.
Why fashion, beyond its prospective approach? Fashion visually, albeit both ambiguously and temporally, communicates not only our collectively-ascribed social status (class, gender, and sexuality), but also our “social identities,” including “the configuration of attributes and attitudes persons seek to and actually do communicate about themselves” (Davis 1996: 16). It is thus, another former of democratic expression, similar to, but distinct from casting a ballot, or answering a pollster’s questions. Fashion communicates both the cultural “Zeitgeist” (the “spirit of the times” or what a society believes about class, gender, and sexuality at a specific moment) and an individual’s relationship (e.g., accepting, rejecting, trivializing, or mocking) to it. Viken (2005: 14-17) trenchantly elucidated the political aspects of fashion: “As much as [fashion] may be condemned for its indifference to ethical and political purposes, fashion remains a disruption of the political and, as such, cannot be other than political.”
Given multi-faceted understandings of fashion as much more than merely clothes, fashion seems well-suited to be used in political science as an empirically-observed expression of both aggregate-level political culture (i.e., the shifting power relationships surrounding class, gender, and sexuality) and individual-level partisan attachments (i.e., how “attributes and attitudes” are related to one’s political identity). Thus, this paper integrates a “new” variable to consider in election forecasting—and the integration of this variable might just tell us more about voters in the course of an election cycle than the usual forecasting tools.
We will utilize OLS regression to estimate our empirical models. OLS provides statistical efficiency in the presence of relatively small sample sizes (which characterizes US Presidential elections), a characteristic that is not necessarily matched by time series methods (e.g., Lewis-Beck, 2005).