Mapping Election Polarization and Competitiveness using Election Results

Overview

• The paper argues that the simplified hypothesis that elections are polarized does not fully explain recent electoral outcomes worldwide.
• It proposes two complementary concepts, Election Polarization (EP) and Election Competitiveness (EC), to understand voting patterns on election day.
• The paper presents an agnostic approach that relies solely on election data and validates it using synthetic and real-world election data across 13 countries.

Plain English Explanation

The paper suggests that the common idea of elections being polarized may not fully explain the real-world outcomes we've seen in recent elections around the world. Instead, the researchers propose looking at two new concepts: Election Polarization and Election Competitiveness.

The researchers argue that by looking at actual voting patterns on election day, rather than just ideological surveys, we can get a better understanding of how voters are dividing on election issues. They use an approach that doesn't rely on ideological data, which can be limited, and instead focuses only on the election results themselves.

The researchers test their approach using both simulated election data and real-world data from 13 countries in Europe, North America, Latin America, and New Zealand. Overall, they find they can identify and distinguish between polarized and competitive elections in these places. Interestingly, they also find that their Election Polarization metric correlates with a measure of political polarization in the US, suggesting this new approach could unlock ways to study polarization at regional levels and in countries where detailed surveys are scarce.

Technical Explanation

The paper presents a framework for understanding election dynamics based on two key concepts: Election Polarization (EP) and Election Competitiveness (EC).

The researchers argue that the common narrative of elections being "polarized" is often based more on perceptions of voting patterns rather than actual ideological data from voters. To address this, they develop an agnostic approach that relies solely on election data, without requiring ideological surveys which can be limited in scope.

The paper outlines mathematical definitions for EP and EC, which capture different aspects of how voters divide on election day. EP measures the degree to which voters sort into distinct camps, while EC looks at how close the race is between the top candidates.

The researchers validate their approach using both synthetic election data and real-world results from 13 countries. They are able to successfully identify and distinguish between polarized and competitive elections in these diverse settings.

Interestingly, the paper also finds that their EP metric positively correlates with a measure of political polarization in the US, suggesting this new approach could enable studying polarization at regional levels and in lower/middle-income countries where in-depth voter surveys are scarce.

Critical Analysis

The paper presents a promising new framework for analyzing election dynamics that goes beyond the simplistic "polarization" narrative. By focusing on actual voting patterns rather than just ideological data, the researchers are able to capture more nuanced aspects of how electorates divide.

However, the validation of the approach is limited to a relatively small set of 13 countries, all of which are economically developed nations. Further research would be needed to assess how well the EP and EC metrics generalize to a wider range of electoral contexts, including less developed democracies.

Additionally, the paper acknowledges that its agnostic, data-driven approach does not provide direct insight into the

of polarization or competitiveness. Complementary research exploring the underlying social, economic, and political factors would be valuable to fully understand election dynamics.

Lastly, the researchers note that their work focuses on national-level elections, but polarization and competitiveness can also manifest at subnational scales. Extending the framework to analyze regional and local election data could yield important additional insights.

Conclusion

This paper offers a fresh perspective on understanding election outcomes by moving beyond simplistic notions of "polarization." The proposed Election Polarization and Election Competitiveness metrics provide a more nuanced way to analyze voting patterns and could unlock new avenues for studying electoral dynamics, especially in contexts where detailed voter surveys are lacking.

While the current validation is limited, the core ideas in this paper represent a promising step towards a data-driven, empirically-grounded framework for making sense of the complex realities of modern elections.