Spatial Polarization of Presidential Voting in the United States, 1992–2012: The “Big Sort” Revisited

Abstract

Much has been written in recent years about the claimed polarization of the U.S. electorate, with substantial differences as to whether there has been greater spatial polarization, at several geographical scales, over recent decades. To assess the veracity of those alternative views, a bespoke data set showing percentage support for the Democratic Party's presidential candidates at the county, state, and divisional scales has been analyzed using a robust, statistically based measure of polarization and segregation. The ecological results provide clear and compelling evidence of a trend toward greater polarization across the nine census divisions, across the forty-nine states within those divisions, and across the 3,077 counties within the states—with strong evidence that the differences over time at the last of those scales are highly statistically significant. Within those general trends, polarization has been greater in some states than others and also within some states more than others—identifying additional geographies calling for further research.

近年来，已有许多有关美国选区极化的着作，但这些着作在论及近数十年来在若干地理尺度上是否有加剧的空间极化现象时，则有着显着的差异。为了评估上述替代观点的准确度，本文运用根据统计测量极化与隔离现象的强健方法，分析一个显示出民主党候总统候选人在郡县、州以及选区尺度的支持率之定製数据。生态结果提供了清楚且具有可信度的证据，証实在九个人口普查选区中，以及在这些选区的四十九个州当中，还有在这些州的三千零七十七个郡县之中，皆有朝向更为极化的发展趋势——且有显着的证据证实，在最后的尺度上随着时间推移的差异，在统计上是高度显着的。在这些普遍的趋势中，在若干州内的两极化较其它州更显着，并且在若干州之间的两极化，亦较其他更为明显——指认出需进行未来研究的其他地理学。

En años recientes se ha escrito mucho acerca de la presunta polarización del electorado en los Estados Unidos, con diferencias sustanciales sobre si, en décadas recientes, se ha dado una polarización espacial mayor, a varias escalas geográficas. Para evaluar la veracidad de esos puntos de vista alternativos se analizó un conjunto de datos generados a la medida que muestra el porcentaje de apoyo a los candidatos presidenciales del Partido Demócrata, a nivel de condado, estado y escalas divisorias, utilizando una sólida medida estadística de polarización y segregación. Los resultados ecológicos proveen evidencia clara y convincente de una tendencia hacia una mayor polarización en nueve divisiones censales, a lo largo de cuarenta y nueve estados situados dentro de esas divisiones, y a través de 3.077 condados dentro de los estados—con fuerte evidencia de que las diferencias a través del tiempo a la última de aquellas escalas son altamente significativas estadísticamente. Dentro de esas tendencias generales, la polarización ha sido mayor en algunos estados que en otros, y también dentro de algunos estados más que otros—identificando geografías adicionales que claman por más investigación.

Key Words:

• polarization
• presidential voting
• spatial scale
• United States

关键词：:

• 两极化，总统选举，空间尺度，美国。

Palabras clave:

• polarización
• elecciones presidenciales
• escala espacial
• Estados Unidos

Acknowledgments

We are very grateful to Clark Archer, Fred Shelley, and Bob Watrell for allowing us to use their data set (see Appendix A) in this study.

Notes

1. Of course, some movers have more choice than others because of their income and wealth and ability to compete in housing markets. Those with constrained choices are necessarily restricted to certain housing market segments only, and increasingly these will be populated by people like themselves.

2. Ross Perot won 18.9 and 8.0 percent of the popular vote in the 1992 and 1996 presidential elections, respectively.

3. Other studies (e.g., Reardon, Yun, and Eitle 2000; Fischer et al. 2004) have looked at nested patterns of segregation using Thiel's entropy measure of segregation, but those applications include no measure of the statistical significance of differences over space or time, which is an important and original feature of the MOR measure deployed here. One frequently criticized aspect of many segregation studies is that they are, in one sense, aspatial because they take no account of the relative location of the spatial units analyzed—hence work introducing spatial autocorrelation measures into segregation studies. The method deployed here does take spatial clustering into account, however, as the ratios for smaller units are calculated within the next-level units where they are nested (as set out in more detail in Manley et al. 2015a).

4. On the nature of the data set, see Appendix A.

5. There are major differences between states in the number of smaller divisions—South Dakota has just three, for example, and Massachusetts has 254—raising potential MAUP issues that will be addressed in further research.

6. This is not to claim that those are the only scales at which polarization occurred. There might well have been within-county polarization in at least some counties—probably mainly those in the country's metropolitan areas. Analysis of these requires finer grained data than we have available, and they could not be obtained for all counties, requiring separate analyses of different states (as in Myers 2013).