Moral Foundations Theory — Why Good People Disagree

You see a trolley heading toward five people. You can pull a lever and divert it — but that kills one person on the other track. You pull the lever. Your friend, equally thoughtful and equally compassionate, says they would never intervene: you'd be using someone as a means to an end. You're both moral people reasoning carefully. So why the complete disagreement?

The answer, according to psychologist Jonathan Haidt, is that you're not using the same moral language. His Moral Foundations Theory proposes that human morality is built on several distinct psychological systems — and different people weight them differently. Understanding which foundations drive your reasoning changes how you interpret moral conflict.

The six moral foundations

Haidt and his colleagues identify six foundational moral intuitions that appear across cultures, though they are expressed differently in different societies:

• –Care / Harm — sensitivity to suffering and the impulse to protect the vulnerable. This is the foundation that makes cruelty feel viscerally wrong.
• –Fairness / Cheating — concern for equal treatment, justice, and the punishment of those who take more than their share.
• –Loyalty / Betrayal — the moral weight of group membership, solidarity, and the duty not to betray your tribe.
• –Authority / Subversion — respect for legitimate hierarchy, tradition, and the institutions that coordinate social life.
• –Sanctity / Degradation — the sense that some things are sacred and that certain acts are morally polluting regardless of their consequences.
• –Liberty / Oppression — resistance to domination and the right to self-determination, even against authorities claiming to act for the common good.

These six foundations are not equally weighted by everyone. Research suggests that people who self-identify as politically liberal tend to weight Care and Fairness most heavily — and are often sceptical of Loyalty, Authority, and Sanctity as valid moral concerns. People who self-identify as conservative tend to value all six more evenly, treating Loyalty, Authority, and Sanctity as genuine moral considerations, not mere social conventions.

The trolley problem through a foundations lens

Return to the trolley problem. Someone who weights Care heavily sees five lives that can be saved and acts. Someone who weights Sanctity and Loyalty more heavily may feel that actively causing a death — even to save more — crosses a line that passive inaction does not. Neither position is irrational. They are reasoning from different foundations.

Fairness and the problem of innocent conviction

Fairness-driven reasoning is particularly sensitive to procedural justice — not just whether outcomes are good, but whether the process that produces them is legitimate. This distinction shows up sharply in legal dilemmas: should you follow rules even when following them produces an outcome you know is unjust? Someone who weights Authority alongside Fairness may say yes — the system's legitimacy depends on consistent rule-following. Someone weighting only Fairness may say no — the point of the rules was never to produce injustice.

Loyalty and the hardest kind of honesty

Loyalty is the foundation that makes betraying a friend feel almost physically wrong — even when the friend has done something objectively harmful. This is not irrationality. It reflects a genuine moral concern: relationships of trust are fragile and valuable, and the social fabric depends on people keeping them. The moral tension arises when Loyalty conflicts with Care (your friend's partner is being hurt) or Fairness (the truth matters). How you resolve that tension often reveals which foundation is running the deeper script in your moral reasoning.

Liberty, Authority, and the surveillance dilemma

The Liberty foundation creates particular tension with Authority: what does a city owe its citizens when safety and freedom pull in opposite directions? People who weight Liberty highly experience surveillance as a direct violation — an imposition of control that undermines the autonomy the state exists to protect. People who weight Authority and Care more evenly may regard pervasive monitoring as an acceptable trade: a managed constraint in exchange for genuine security. Neither is simply wrong. They are calibrating different foundations.

What the research says — and where it gets complicated

Moral Foundations Theory has generated substantial research and also substantial criticism. Some critics argue that the theory conflates descriptive findings (people do weight these foundations) with normative claims (they should). Others note that the six foundations are not culturally universal in the way early formulations implied — they appear across societies, but their content and ranking vary considerably. The Liberty foundation was added to the original five partly in response to critics who felt the theory did not adequately capture libertarian moral intuitions.

The practical insight that survives the theoretical debates is durable: most serious moral disagreements are not disagreements about facts or even about which outcomes matter. They are disagreements about which moral dimensions are relevant in the first place — and which should take priority when they conflict. Understanding that your moral opponent is weighting a different foundation, not ignoring the foundations you care about, can change how a difficult conversation goes.

Your moral profile on SplitVote

Every dilemma on SplitVote implicitly activates one or more moral foundations. The scenarios involving harm and protection activate Care. Dilemmas about following unjust rules activate Fairness and Authority in tension. Questions about loyalty, secrecy, and betrayal activate Loyalty. Scenarios involving surveillance, bodily autonomy, or drug policy activate Liberty and Sanctity. Voting across enough dilemmas builds a picture — not of your "moral type", but of which considerations you consistently treat as decisive when they come into conflict.

Moral Foundations Theory is a theoretical framework in moral psychology, not a validated psychological test. References to Haidt and colleagues' work are for contextual background only. SplitVote dilemmas are for reflection and discussion — not clinical assessment. Results represent our community's votes, not scientific conclusions about individual personalities.