Omission Bias

Omission Bias is judging harm caused by inaction as less bad, or less blameworthy, than equivalent harm caused by action.

People often rate actions that cause harm as worse than omissions that allow the same harm. "I didn't do it" feels morally cleaner than "I did something and it went wrong," even when the outcomes match. Withholding a treatment, declining a vaccine, or leaving a risky process unchanged can feel safer than an active choice with a visible downside. The harm still happens - it just arrives through the path of letting things run.

Parents and patients may prefer a disease risk from "nature" over a smaller risk from an intervention because the intervention would be something they chose. The causal story matters more than the expected harm in the moment of decision. Institutions reinforce the pattern. A visible change that fails gets a postmortem. A quiet default that fails for years may never get a name. That asymmetry trains people to protect themselves by not acting, even when acting would reduce total harm.

This differs from status quo bias (preferring the current state because it is familiar) and from zero-risk bias (wanting to eliminate one risk completely). Omission bias is specifically about the moral and emotional asymmetry between acting and refraining when expected outcomes are comparable.

Sometimes not intervening is correct - uncertainty, consent, or expected harm can favor waiting. The bug is rejecting a better expected outcome mainly because acting would make you the author of a possible bad result, while inaction's authorship stays invisible.

• "If something goes wrong from the vaccine, it's on me - better skip it."
• "I didn't cause the outage; I just didn't change the system."
• "Selling would lock in the loss - holding isn't really a decision."
• "Nature can take its course; intervening feels like playing God."
• "If I act and fail, they'll blame me. If I wait, it's just how things went."
• "Leaving the default settings is safer than flipping the switch."
• "I won't recommend the surgery - what if complications are my fault?"
• "Doing nothing means I didn't hurt anyone."

Vaccine authorship fear: a parent rejects a vaccine with a tiny risk because a side effect would feel chosen, while a larger disease risk feels like fate. Spranca and colleagues' omission-commission gap is doing the moral math.

Silent technical debt: an engineer leaves a brittle system untouched. A rewrite might cause a short outage under their name. Years of quiet failures under the old default feel less personally owned.

Holding as non-choice: an investor refuses to sell a loser because selling would "make" the loss. Holding is treated as neutral, even though it is an active allocation of capital.

Policy inertia: a manager skips a safety change that could prevent rare harm because a visible change that backfires would be career-ending. The status quo's victims stay statistically anonymous.

Medical hesitation: a clinician delays a beneficial intervention to avoid being the proximate cause of a side effect. Expected harm rises while moral comfort stays high.

Omission bias leaves harmful defaults in place. Preventable damage accumulates under the cover of "I didn't do anything."

Medicine, parenting, and public health suffer when smaller intervention risks lose to larger natural risks because of authorship feelings.

Organizations under-invest in upgrades. People optimize for personal blame avoidance rather than total harm reduction.

Accountability becomes asymmetric: actors get punished; omitters get excused, so culture selects for paralysis.

Over time, the safest career move becomes never being the person who touched the lever - even when the lever would have helped.

Causal and moral intuition treats actions as stronger causes than omissions. Blame attaches more cleanly to a deed than to a failure to intervene.

Anticipated regret is louder for commissions: "I caused this" hurts more in prospect than "I let this happen," even when outcomes match.

Institutions and law often punish visible acts more than silent neglect. Norms about "do no harm" get misread as "do nothing" when harm is already on the path.

Spranca, Minsk, and Baron's 1991 Journal of Experimental Social Psychology paper, Omission and Commission in Judgment and Choice, documented that people often evaluate harmful omissions as less bad than harmful actions with similar outcomes.

Related work on vaccination and medical decisions shows the same authorship asymmetry: people may accept higher natural risk to avoid lower intervention risk they would feel responsible for.

The practical lesson is to compare expected harms on both paths. Once you know the risks, inaction is also a choice - and it can be the more harmful one.

• You feel cleaner declining an intervention than accepting a smaller natural risk.
• "I didn't do anything" is your main defense after harm occurs.
• You treat holding or waiting as non-decisions.
• Fear of being blamed for acting outweighs total harm math.
• Defaults survive for years without an expected-value review.
• You would judge someone else harshly for the same omission you are making.

Score action and inaction with the same expected-harm rules.

• Write both paths' risks side by side before you decide.
• Ask who is harmed if you do nothing - name them.
• Treat holding, waiting, and defaults as explicit choices.
• Separate "blame if visible" from "harm if likely."
• In reviews, audit omissions with the same seriousness as commissions.
• Pre-commit criteria for when intervention is required so fear of authorship cannot veto.

When is doing nothing the better choice?

When action has irreversible downside and waiting gathers cheap information - watchful waiting in medicine, delay on a pushy sale. The bug is treating harm from inaction as morally lighter than the same harm from action.

If I am unwilling to intervene, is that always omission bias?

Not if you weighed both harms and chose. It is the bias when "I didn't touch it" feels cleaner than an equally damaging act you would refuse to own.

Do vaccination-style dilemmas prove omission bias is everywhere?

They show the pattern vividly: people often prefer risk from nature to risk from choice. Use that as a check, not as a universal smear on caution.

Can institutions exploit omission bias?

Yes - defaults that harm through inaction, or blame cultures that punish errors of commission more than neglect. Design accountability for both.

What if action would make me the visible villain?

Visibility is a real political cost. Still separate "who gets blamed" from "which option harms less." Sometimes leadership is owning the commission.

Against Omission Bias, own both paths - then pick the one with lower expected harm, not the one with quieter authorship.

Intervention risk

Holding the loser

Leave the system

Practice this pattern in the Reframing App - capture the trigger, label it (like Omission Bias), check evidence, and write a more balanced thought.