Understanding Political Polarization: The Psychology of Morality and Harm

According to Kurt Gray's dyadic theory of morality, all moral judgments are rooted in a single cognitive template of harm, which consists of an agent causing damage to a vulnerable patient. This challenges the popular moral foundations theory that proposes distinct moral modules (harm, fairness, loyalty, authority, purity). Gray argues that what appears to be 'harmless wrongs' (like incest or purity violations) are actually perceived as harmful by most people, and that perceptions of harm—rather than separate moral foundations—are the fundamental driver of moral judgment. This unified framework explains moral diversity across liberals and conservatives as differences in how people perceive harm, rather than having fundamentally different moral minds.

Dyadic morality proposes that all moral judgments ultimately stem from perceptions of harm, where an agent intentionally causes damage to a vulnerable patient; unlike Moral Foundations Theory which posits multiple dedicated circuits for different moral concerns, dyadic morality suggests that the apparent diversity in moral judgments can be explained by combining a basic template of agent-patient relationships with varying degrees of perceived harm across different domains.

The diadic theory proposes that moral judgments require three elements: an agent (someone who causes harm), a patient (someone who suffers), and intentional harm. This simple template explains many moral intuitions. When we perceive harm, we tend to complete the diadic template by assuming there is an intentional agent responsible. Moral encasement is the tendency to permanently assign moral roles to people - once someone is seen as an aggressor or victim, that role tends to remain fixed.

This section establishes the theoretical foundation for understanding gender bias in moral judgment. Kurt Gray's dyadic template theory proposes that humans naturally categorize individuals in moral situations as either agents/perpetrators (deserving blame and punishment) or patients/victims (evoking sympathy and aid). These categories are mutually exclusive - seeing someone as more agentic makes it harder to see them as a victim. The research argues that gender shapes how this template is applied: men are more readily associated with agentic roles due to historical warfare participation, greater musculature linked to lower pity perceptions, and higher rates of physical aggression. Conversely, women are more likely to be typecast as victims due to evolutionary pressures favoring protection of pregnant women (who set reproductive limits) and physiological differences showing lower average pain thresholds. This creates a cognitive framework where women fit the victim prototype and men fit the perpetrator prototype.

The diadic theory of morality proposes that our mental template for judging moral acts is very simple: there is an aggressor or perpetrator who intentionally causes harm, and there is a victim who suffers from that action. The agent (perpetrator) has agency and intention, while the victim has suffering and emotion. The agent has the capacity for guilt and moral responsibility, while the victim has only the capacity to feel pain and has rights but no responsibilities. This simple framework explains why we tend to view victims as morally superior and why moral judgments often focus on identifying who is the 'bad guy' and who is the 'victim.'

Moral outrage itself is good—it signals when something is wrong and needs fixing—but the natural pleasure of being morally right is easily hacked by online content and algorithms. The attention economy transforms genuine concern into 'outrage porn,' where users seek out moral clarity over moral outcomes and prioritize the feeling of being right over actually doing what's right. This manifests as writing off disagreeing people as holistically bad (preserving one's sense of clarity) and spending time online posting righteous content without ever creating the civility or action that such outrage should produce. The algorithm feeds users infuriating takes because it measures engagement, creating a cycle where outrage feels productive but doesn't actually address underlying issues.

Social media platforms use algorithms that prioritize content with moral and emotional qualities, particularly outrage, because such content generates the highest engagement and advertising revenue; research shows that immoral events learned online trigger more outrage than those encountered through traditional media, potentially leading to 'outrage fatigue' where constant exposure depletes our capacity for meaningful moral outrage over important issues.

Social media algorithms drive outrage by showing users the worst of the opposing side to trigger dopamine releases. Users aren't just hearing their own side—they're seeing cherry-picked worst-case scenarios that justify their feelings. This creates addiction to outrage and reinforces negative perceptions of the other side, making productive dialogue increasingly difficult.

The introduction of smartphones and social media around 2013 fundamentally changed how information spreads. Unlike traditional journalism which provided context, bystander videos uploaded to platforms like Facebook are filtered through algorithms designed to maximize engagement by showing content to people most likely to be angered by it. This creates a feedback loop where emotionally charged narratives about race—such as the idea that cops are uniquely targeting black people—spread rapidly without journalistic scrutiny. Critical race theory then provides explanations for these observations, creating a self-reinforcing cycle that perpetuates racial polarization.

Pernicious polarization represents an advanced stage where political division leads to harmful consequences for democracy, characterized by hostile camps sharing identical issue positions and identities. The United States is the only advanced liberal democracy experiencing this level of polarization. Potential structural reforms include ranked choice voting, which allows voters to rank multiple candidates and potentially produces winners with broader appeal, and multi-member districts, which would give smaller parties representation and reduce binary choice structures. These reforms aim to move away from single-member districts and winner-take-all systems that currently favor extreme candidates in primaries, thereby encouraging more moderate political behavior and facilitating cross-party cooperation.

Multiple approaches can reduce polarization: Structural reforms include ranked choice voting, proportional representation, and making Election Day a holiday to increase participation. Bipartisan congressional committees and joint trips to facilities humanize opponents and build relationships. Psychological interventions that facilitate meaningful contact between people from different backgrounds correct misperceptions about opposing groups. Listening without preconceptions allows recognition of shared struggles across political divides. While some problems are intractable, these combined approaches offer pathways toward healing polarization and restoring functional democratic discourse.

Three institutional reforms can address polarization: (1) Teaching integrative negotiation to legislators using Mary Parker Follett's techniques focusing on underlying interests rather than positions; (2) Implementing civic lotteries where randomly selected citizens deliberate on issues, often producing moderate compromises; (3) Establishing recursive representation with continuous feedback between representatives and constituents. Land reform that increases housing supply can simultaneously boost economic growth and reduce polarization by fostering social mixing. PPE education helps politicians think more rigorously about morality while understanding institutional realities. Ranked choice voting reduces polarization by changing electoral incentives—candidates must appeal to broader coalitions. These approaches recognize that polarization results from structural incentives and cognitive frameworks, not merely individual attitudes.

Multiple institutional reforms can address polarization: redistricting reform using tripartite citizen commissions (as in California) to draw competitive districts; open non-partisan primaries replacing closed party primaries; ranked choice voting (adopted in Alaska, Maine, and multiple cities) making politicians less beholden to extremists; restoring the Voting Rights Act; and vote-at-home systems. These reforms are advancing in Nevada, Missouri, Hawaii, Washington DC, Portland, Oregon, and other jurisdictions. Comparative analysis shows the U.S. faces more severe polarization than most Western democracies, partly due to winner-take-all electoral systems and lack of multi-party coalition-building requirements found in countries like Germany and the UK.

Structural political reforms could combat polarization by removing incentives for divisive rhetoric. Gerrymandering creates electoral districts requiring extreme ideological positioning, analogous to recommendation algorithms creating echo chambers. Two-tier primary systems, like California's top-two primary, require candidates to appeal to broader constituencies rather than just primary voters. These reforms deprive demagogues of the oxygen they need to thrive. Combined with media accountability measures, such structural changes could reduce the incentive structures rewarding misinformation and polarization.