Socioeconomic determinants of conservative vote share in the 2016 U.S. presidential election
Xinyao Fu
School of Mathematics and Statistics, University of Glasgow
DOI: https://doi.org/10.59429/bam.v7i4.12409
Keywords: county-level analysis; generalized additive models;income and education;model comparison; nonlinear effects;conservative vote share
Abstract
This study examines the socioeconomic determinants of Conservative voting outcomes in the 2016 U.S. presidential election using county-level data. Combining election returns with indicators of income, educational attainment, poverty, and unemployment, the analysis assesses both linear and nonlinear relationships between local socioeconomic conditions and Conservative vote share. A multi-model framework is employed, including linear and logistic regression, generalized additive models (GAM), and tree-based methods, to balance interpretability and predictive performance. The results show that median household income and educational attainment are the most influential predictors of Conservative support, while poverty and unemployment play more limited roles. Importantly, the effects of income and education are strongly nonlinear, with the largest marginal impacts concentrated among counties with lower socioeconomic levels and diminishing at higher levels. Model comparisons indicate that GAM offers the most informative and interpretable representation of these relationships, whereas random forest models achieve higher predictive accuracy. Overall, the findings suggest that regional political polarization in the United States is closely associated with structural socioeconomic inequality at the county level.
References
[1] Pew Research Center (2018). The Partisan Divide on Political Values Grows Even Wider. Washington D.C.
[2] Rodden, J. (2019). Why Cities Lose: The Deep Roots of the Urban-Rural Political Divide. Basic Books.
[3] Gelman, A., Kenworthy, L., and Su, Y. (2010). Income Inequality and Partisan Voting in the United States. Social Science Quarterly, 91(5), 1203–1223.
[4] Lipset, S. M. (1959). Some Social Requisites of Democracy: Economic Development and Political Legitimacy. American Political Science Review, 53(1), 69–105.
[5] Autor, D. H., Dorn, D., Hanson, G. H., and Majlesi, K. (2020). Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure. American Economic Review, 110(10), 3139–3183.
[6] Bishop, B. (2008). The Big Sort: Why the Clustering of Like-Minded America Is Tearing Us Apart. Mariner Books.
[7] Glaeser, E. L., and Ward, B. A. (2006). Myths and Realities of American Political Geography. Journal of Economic Perspectives, 20(2), 119–144.
[8] McCarty, N., Poole, K. T., and Rosenthal, H. (2016). Polarized America: The Dance of Ideology and UnequalRiches. MIT Press.
[9] MIT Election Data and Science Lab (2020). County-Level Presidential Election Returns (2000–2020). Harvard Dataverse.
[10] U.S. Census Bureau (2016). County Facts Database. Washington D.C.
[11] Hastie, T., Tibshirani, R., and Friedman, J. (2009). The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Springer.
[12] Field, A., Miles, J., and Field, Z. (2012). Discovering Statistics Using R. SAGE Publications.
[13] Fox, J., and Weisberg, S. (2019). An R Companion to Applied Regression (3rd ed.). Sage Publications.
[14] O'Brien, R. M. (2007). A Caution Regarding Rules of Thumb for Variance Inflation Factors. Quality & Quantity, 41(5), 673–690.
[15] Kuhn, M., and Johnson, K. (2013). Applied Predictive Modeling. Springer.
[16] Breiman, L., Friedman, J. H., Olshen, R. A., and Stone, C. J. (1984). Classification and Regression Trees. Wadsworth International Group.
[17] Kuhn, M., and Silge, J. (2022). Tidy Modeling with R. O’Reilly Media.