The Geography of Resentment

 The Sequential Theory of Political Animosity provides a place-based framework to explain the global rise of political hostility by adjudicating between "bottom-up" socio-economic explanations and "top-down" political ones. This conceptual framework bridges fragmented literatures—such as those on affective polarization, populism, and social group prejudice—by offering a broader, encompassing definition of political animosity.

Definition of Political Animosity

Within this framework, political animosity is defined as animosity regarding others' behaviors, attitudes, goals, and identities in the realm of politics. Key features of this definition include:

• Broad Emotional Scope: It encompasses a "family of hostile emotions" including anger, resentment, indignation, contempt, and disgust.
• Encompassing Targets: Unlike "partisan animosity," which focuses on specific outgroups like parties or ideologies, this concept includes animosity toward the elite establishment, place-based groups (e.g., urban vs. rural), and social groups such as immigrants or the "educated elite".
• Political Relevance: It specifically concerns how resources, rights, esteem, and power are distributed in society, and the actors influencing government actions.

The Three-Part Sequential Theory

The theory proposes that the rise of this animosity occurs in a specific causal sequence rooted in geographic location:

1. Bottom-Up Emergence in Declining Places: Widening geographic inequalities generate political animosity in relatively declining areas even before the salient rise of populists. Residents in these areas perceive economic and symbolic threats to their place-based identity, which translates into resentment toward governments and urban centers.
2. Facilitation of Populist Rise: This pre-existing, accumulated animosity creates an "opportunity structure" that right-wing populists exploit. Consequently, declining places with higher pre-existing animosity show larger electoral swings toward populist candidates.
3. Top-Down Reaction in Prospering Places: The subsequent electoral success and Manichean rhetoric of populists incite threat perceptions in prospering areas. This produces a reactive political animosity in these "winning" areas, which may previously have had low levels of hostility because the status quo served their interests.

Theoretical Implications

The framework argues that current high "emotional temperatures" in politics are not driven solely by the "left-behind". Instead, it posits a cycle where socio-economic processes first ignite animosity in declining places, while political processes (populist mobilization and victory) later spread and amplify that animosity to prospering places, eventually consuming the entire political landscape. This allows researchers to view animosity as a reaction to context and threat perceptions rather than an inherent trait of specific social groups.

The Sequential Theory of Political Animosity is a place-based framework designed to explain the global rise of political hostility by adjudicating between "bottom-up" socio-economic explanations and "top-down" political ones. This framework proposes that the high "emotional temperature" of current politics is the result of a specific three-part causal sequence that spreads animosity across different geographic areas over time.

Part 1: Bottom-Up Emergence in Declining Places

The theory begins with widening geographic inequalities, where the emergence of the knowledge economy concentrates prosperity in urban hubs while other areas decline.

• Socio-economic Threats: Residents in relatively declining places experience direct economic threats (e.g., falling property values, loss of services) and symbolic threats (e.g., territorial stigma, sociocultural marginalization).
• Pre-existing Animosity: These threats generate high levels of political animosity—resentment toward governments, urban centers, and social minorities—before the salient rise of populist politicians. This suggests that economic discontent can translate into political hostility even without elite mobilization.

Part 2: Facilitation of the Rise of Populists

The pre-existing animosity accumulated in declining places creates an "opportunity structure" that right-wing populists can exploit.

• Electoral Swings: The theory posits that right-wing populists see larger electoral gains specifically in declining places that already exhibited high levels of pre-existing animosity.
• Direct Link: Findings confirm that once economic decline produces political animosity, that animosity itself has a direct effect on populist support beyond what economic decline alone can explain.

Part 3: Top-Down Reactive Animosity in Prospering Places

The final stage of the sequence occurs when the electoral success and rhetoric of populists incite threat perceptions in prospering areas.

• Reactive Animosity: Unlike declining places, prospering areas typically exhibit low political animosity before populists rise because the status quo serves their interests. However, once populists win or use salient rhetoric, residents in these "winning" areas perceive a threat to their economic status and political control.
• The "Gap" Closes: This generates a reactive political animosity toward populist elites, their supporters, and associated social groups (like rural or less-educated voters). This process effectively spreads political animosity to the entire landscape, closing the "animosity gap" between declining and prospering places.

Significance in the Larger Framework

The three-part sequential theory challenges the narrative that rising animosity is exclusively driven by "left-behind" communities or populist supporters. Instead, it offers a cycle where socio-economic processes first ignite animosity in declining places, and political processes (populist victory) later spread it to prospering places. This allows researchers to view animosity as a reaction to context and threat perceptions rather than an inherent trait of specific social groups.

To test the Sequential Theory of Political Animosity, the source utilizes a multi-methodological approach that combines massive social media datasets with deep learning and econometrics to track emotional shifts across time and space. This methodology allows the researcher to move beyond traditional survey-based "feeling thermometers" and capture organic, real-time political discourse.

Data: 3 Billion Geo-located Tweets

The core of the analysis is based on 3 billion tweets collected between 2013 and 2022 by the Harvard Center for Geographic Analysis.

• Geographic Granularity: The data includes precise GPS coordinates or place tags, allowing the researcher to link individual expressions of animosity to fine-grained local indicators of decline and prosperity, such as U.S. counties, French employment zones, and Korean sigungus.
• Temporal Reach: Spanning a decade, the data covers the periods before, during, and after the salient rise of right-wing populists (Donald Trump, Marine Le Pen, and Yoon Suk Yeol).
• Focus on Mass Opinion: To ensure the data reflects the views of ordinary citizens, the researcher excludes tweets from identified bots, job advertisements, and accounts belonging to institutional entities or professional politicians.

Measuring Political Animosity

Measuring animosity in natural language required a two-step deep learning process:

1. Classification: Using BERT (Bidirectional Encoder Representation Transformer) models, the researcher fine-tuned country-specific language models on hand-coded data to classify whether a tweet was "political". This resulted in a dataset of over 186 million political tweets in the U.S. alone.
2. Scoring Animosity: The Google Perspective API was used to assign a continuous score (0–100) of animosity to each political tweet, identifying language that is "hateful, aggressive, disrespectful, [or] rude".
3. Relative vs. Absolute Measures: For decade-long trends, the researcher uses relative percentiles of animosity to account for changes in platform-wide data characteristics over time. For short-term event studies, absolute levels are used.

Measuring Economic Decline and Prosperity

To define the status of a "place," the researcher used specific socioeconomic indicators:

• Relative Decline: This is primarily measured as the percentile of a place’s economic growth over time.
• Specific Metrics: Median household income growth was used for the U.S. and France, while per capita Gross Regional Domestic Product (GRDP) was used for South Korea.
• Community Types: In the U.S. analysis, the data was further broken down into a 15-type county typology (e.g., "Big Cities" vs. "Working Class Country") to see if different types of declining places behaved similarly.

Analytical Framework for Testing the Theory

The theory's three stages were tested using different statistical techniques:

• H1 & H4 (Pre-existing Animosity): Bivariate linear regressions were used to correlate community characteristics (unemployment, commute times, industry decline) with pre-populist animosity levels.
• H2 (Populist Swings): The researcher analyzed electoral swings by mapping the size of the vote shift toward populists against both economic decline and pre-existing animosity.
• H3 & H5 (Reactive Animosity): Interrupted Time Series (ITS) analysis was employed to test for "discontinuous changes" in animosity levels on the exact days of populist campaign launches or election victories.

The geographic findings within the Sequential Political Animosity Theory highlight how spatial inequalities and place-based identities drive the evolution of political hostility across different types of localities. By analyzing geo-located data, the sources uncover specific patterns of animosity that vary according to a location's economic trajectory and its relationship to national power centers.

Spatial Distribution of Pre-existing Animosity

Before the prominent rise of populists, political animosity was geographically concentrated in relatively declining places.

• Declining Regions: In the U.S., high levels of animosity were found in the American Rust Belt and the economically insecure Deep South. In France, animosity was highest in traditional manufacturing areas in the Northeast, while in South Korea, it was concentrated in primary industry areas in the Southwest, such as the Jeolla province.
• Community Typologies: In the U.S., "relatively declining" community types—including Working Class Country, Middle Suburbs, and Native American Lands—exhibited high levels of animosity as early as 2013. Conversely, culturally conservative areas like Evangelical Hubs and Graying America initially showed lower levels of outward animosity but were later brought into antagonistic politics through populist mobilization.

Urban-Rural vs. Center-Periphery Inequality

The geographic findings demonstrate that different national structures influence where animosity emerges.

• Multi-centric vs. Mono-centric: In the multi-centric U.S., animosity followed a broad urban-rural divide. In the mono-centric systems of France and Korea, however, center-periphery inequality was more salient.
• The "Bedroom Community" Effect: In France and Korea, political animosity was relatively low inside the capital cities (Paris and Seoul) but significantly higher in their surrounding bedroom communities. Findings show that longer commutes were strongly associated with higher pre-existing animosity in the Seoul metropolitan area, reflecting the frustration of residents living in the periphery of prospering centers.

The Geographic Shift: Prospering Places React

A key finding of the sequential theory is the dramatic shift in animosity within prospering areas following populist successes.

• Reactive Surge: Areas like Big Cities, Urban Burbs, and College Towns initially expressed the lowest levels of political animosity. However, following the election of figures like Donald Trump or Yoon Suk Yeol, these "winning" areas experienced the sharpest increases in animosity, effectively closing the geographic "animosity gap".
• Within-City Dynamics: In Paris, a distinct shift occurred where pre-existing animosity was originally higher in deprived outer arrondissements; after Marine Le Pen's 2017 campaign, animosity surged in the wealthier central arrondissements, suggesting a reactive response from the urban elite to the populist threat.

Indicators of Geographic Resentment

The sources identify several local factors that correlate with high pre-existing political animosity in a given area.

• Economic Decline: Manifestations of decline—such as falling housing values, population loss, and industrial employment decline—are consistently associated with higher animosity.
• Service Deprivation: In France and Korea, weaker provision of public services and cultural facilities at the local level was a significant predictor of regional resentment.
• Place Attachment: The findings suggest that when economic threats overlap with geographic segregation, people develop exaggerated perceptions of group differences, which intensifies animosity toward distant, prospering outgroups.

The Sequential Theory of Political Animosity offers a place-based framework that integrates "bottom-up" socioeconomic factors with "top-down" political processes to explain the global rise of political hostility. The core takeaways from the sources center on the timing, location, and reactive nature of this animosity across the political landscape.

1. Animosity Emerges Organically in Declining Places

A primary takeaway is that political animosity often develops before the rise of salient populist mobilization. In relatively declining areas—such as the American Rust Belt, French manufacturing zones, or Korean primary industry regions—widening geographic inequalities generate economic and symbolic threats. These threats translate into organic resentment toward governments and distant urban centers even without elite cues, suggesting that economic discontent can become politicized through internal psychological mechanisms.

2. Pre-existing Animosity as an "Opportunity Structure"

The sources highlight that this accumulated hostility in declining places is a prerequisite for populist breakthroughs. Data from the U.S. and France confirm that declining regions with higher pre-existing animosity exhibited larger electoral swings toward right-wing populists like Donald Trump and Marine Le Pen. In this sense, the animosity generated by geographic inequality serves as an opportunity structure that populists exploit to gain power.

3. The "Animosity Gap" Closes Through Reactive Hostility

One of the most significant takeaways is the phenomenon of top-down reactive animosity in prospering areas.

• Initial Security: Prospering places (e.g., big cities and college towns) typically begin with low animosity because the status quo serves their interests.
• The Reaction: When populists achieve salient victories or use aggressive rhetoric, residents in these "winning" areas perceive a threat to their social status and political control.
• The Result: This triggers a sharp increase in hostility toward populists and their supporters, effectively closing the geographic animosity gap and spreading political hostility to the entire nation.

4. Broadened Conceptualization of Animosity

The theory moves beyond traditional "partisan animosity"—which focuses on dislike for opposing parties—to a broader concept of political animosity. This definition encompasses a "family of hostile emotions" (anger, contempt, disgust) directed at a wide range of targets, including the elite establishment, urban or rural residents, and specific social groups like immigrants or the "uneducated". This allows for a comparison of animosity across different segments of society that may not have strong traditional partisan identities.

5. Normative Threats to Liberal Democracy

The final takeaway concerns the long-term health of democratic systems. As animosity consumes the political sphere, it fosters anti-pluralistic governance desires:

• Populism in Declining Places: Residents may desire the unchecked governance of the "will of the people," potentially ignoring liberal principles.
• Elitism in Prospering Places: Reactive hostility in prospering hubs can transform into technocratic elitism, where residents may feel that "ordinary people" lack the knowledge to make sensible judgments, potentially contradicting democratic principles.

Ultimately, the sources argue that high political temperatures are not just a trait of certain groups but are reactions to context and threat perceptions that cycle through different geographic regions.