Authors: John O'Loughlin*, University of Colorado, Andrew Linke, Department of Geography, University of Utah, Frank Witmer*, University of Alaska - Anchorage
Topics: Political Geography, Quantitative Methods, Eurasia
Keywords: Surveys, post Soviet space, multilevel models, contexts
Session Type: Virtual Paper
Start / End Time: 1:30 PM / 2:45 PM
Room: Virtual 50
Presentation File: No File Uploaded
Geographers believe that where people live (their social contexts) shape their political and other behaviors. Most analyses concentrate on demographic characteristics and other personal social variables in understanding political beliefs. The divide between political geographers and political scientists on the relative importance of context has been evident for nearly a century. Studies that are limited in scope and geographic range hardly move the discussion. To tackle this limitation, we add an ambitiously large multilevel analysis of national survey data for over 12000 respondents whose home locations are identified f1 countries and territories of the former Soviet Union (Kazakhstan, Georgia, Armenia, Moldova, Belarus, Ukraine, Crimea, Transnistria, Nagorno-Karabakh, Abkhazia and South Ossetia). We fit multilevel models (for local, regional, and national scales) for a key survey outcome, answers to a question about the Soviet legacy. The responses to the simple binary question “Was the end of the Soviet Union a right or wrong step?” has been shown to be instrumental in predicting many other contemporary beliefs and attitudes. We evaluate a range of demographic (age, gender, education, material status, mood), cultural values (Russian as the home language and a sense of belonging to Russian culture), traditional attitudes (religiosity and beliefs about who should be allowed to marry), media sources for news (television, internet and social media), and domestic political views (direction of the country and democracy as the best political system) in a multilevel model that indicates the relative importance of geographic and other predictors for the aggregated sample and each country.