A judge gives a verdict and leaves, never meeting the person who has to live with the outcome. Someone quickly reviews a resume or ignores a request, and their day goes on as usual. But the person on the receiving end feels the impact, whether it’s at the table, in their inbox, or in a meeting room.

Willis and Todorov found that just 100 milliseconds of seeing a face is enough for people to form an impression. (Source: NIH) In another example, when insurers gave underwriters the same information, their quotes still varied by a median of 55%. (Source: Kahneman, Sibony, and Sunstein, Noise: A Flaw in Human Judgment, 2021) No system links these mistakes back to the people who made them.

This pattern is even stronger in the 2026 job market. Now, nine out of ten U.S. employers use AI software to rank job applicants. (Source: Stanford). In April 2026, an Enhancv survey of 1,066 job seekers showed that 50.5% got rejections without any explanation from a person. (Source: Enhancev)

A court case can come back to the judge who made the call. A diner gets a silent shake of the head, an applicant gets a form letter, an employee gets a number in a spreadsheet. The person who decided moves on. No one checks if the call matched the facts. The fastest answer wins by showing up first.

What Does Judging Others Cost the Judge?

Quick Summary: When you make your own decisions, you feel the results personally. (Source: Nassim Nicholas Taleb, Skin in the Game: Hidden Asymmetries in Daily Life, Random House, 2018.) But if you judge from afar, nothing comes back to you. A careless decision just fades away. Someone who gives approval or blame from a distance never feels the impact in their own life. Whether they say something harsh or kind, it doesn’t affect them.

The question is: why does judgment come so easily? Every judge we’ve talked about walks away with no cost for their decision. I often hear that it costs nothing to be kind. The other side of the truth, which seldom gets said, is that it also costs nothing to be unkind.

The fact that there is no cost to being wrong when judging someone we don’t know is the reason why it continues. At least the person handing out the verdict faces no consequences.

How Fast Do We Judge a Stranger?

Quick Summary: People form first impressions from a stranger’s photo in just a tenth of a second, even before speaking to them. Looking longer only makes people more confident, but it does not change their initial ratings. There is no way to check if these first impressions match the person’s real character.

When someone sees a face, they often judge it before they even realize it. Willis and Todorov wanted to see how quickly people form first impressions. In five experiments, participants looked at unfamiliar faces and rated them for different traits. Sometimes, they showed the face for just 100 milliseconds.

The ratings given after just 100 milliseconds were similar to those from people who had as much time as they wanted. (Source: PubMed) More time made people feel more confident in their impressions, but it did not make their ratings match any better. (Source: BPS)

The experiment compared quick glances with longer looks, but it never checked if the first impressions were actually accurate. The process ended once participants gave their ratings. After just a tenth of a second, people judged, but no one checked whether they were right.

What Are the Consequences of Judging Others Unfairly?

Quick Summary: A server returns to the kitchen, still thinking about a diner’s sudden warning about a fire he never saw. A job candidate leaves the building, replaying the interviewer’s shallow questions. An employee reads a performance review and notices the manager’s own mistakes hidden in the feedback. The person on the receiving end keeps holding onto what the other person passed along.

The person waiting for the water faces no consequences when he calls his server careless, even though a kitchen fire actually happened that nobody mentioned.

The hiring manager who rejects what might be the perfect hire because they fail to ask the right questions feels no consequence when they eventually hire someone who may not fill what’s needed.

Fifteen years in paid search taught me that every campaign faces an audit. Hiring calls face none.

The manager who tells an employee they are performing poorly, when the performance actually reflects the manager, faces no consequences for giving the employee a poor rating on their performance review.

Kahneman, Sibony, and Sunstein document this variation in performance reviews and personnel selection. (Source: Harvard)

In every case, the cost of the verdict falls on the person receiving it, not the person giving it. The verdict doesn’t affect the judge at all. They move on to the next table, the next application, or the next review without really caring how the judgment affects anyone.

Is Costless Judgment a Moral Hazard?

Quick Summary: One person chooses how cautious to be, but someone else might pay the price if that caution is not enough. In fields like insurance and banking, contracts and oversight help manage these risks. However, in everyday situations where one person takes a risk for another, no contract is usually involved. As a result, the cost often falls on someone who did not make the decision.

Economists have a name for this condition. Paul Krugman describes moral hazard as a split: one person sets the level of risk, and another person covers the loss. (Source: Ethics Unwrapped)

Insurance shows it best. A homeowner with fire insurance may take less care against fire because the insurer absorbs the loss. (Source: Ethics Unwrapped)

Swap the parties, and the split holds. The restaurant patron decides how much care to take, and the server absorbs the loss. The hiring manager decides how many questions to ask, and the candidate absorbs the miss.

Contracts, regulations, and oversight exist to keep risk and consequence from drifting apart. (Source: The Behavioral Scientist) Everyday assessment has none of the three.

The fit is structural. Strictly, moral hazard describes risk taken after a transaction. (Source: Economics Online) A verdict carries no transaction, yet it carries the same split: one party decides, another pays.

Why Does Costless Judgment Work Like a Trial in Absentia?

Quick Summary: If a rater makes a mistake, the system charges them. If they get it right, the system pays them. When a trial happens without the defendant present, no one corrects the record after the verdict. If two raters review the same case and reach different conclusions, the difference goes unnoticed. No one tracks it, pays for it, or tries to fix it.

Judgment rarely self-corrects because it holds no significance for the judge.

Sometimes misinformation gets called out and corrected, but those aren’t the same scenarios. In these scenarios, no feedback loop charges the judge for getting it wrong.

Kahneman, Sibony, and Sunstein call the variation between judgments of identical cases noise. (Source: Harvard) Two judges in one courthouse can sentence the same crime differently. One judge can rule differently on Monday than on Wednesday. Nobody bills either judge for the gap.

This is all built into the system and works perfectly. The structure of a trial in absentia guarantees that the judge never finds out if they were wrong. This is structurally built into a trial in absentia. Another reason why courts forbid it as much as possible.

Not only does the defendant never get the chance to appear and correct the record, but the judge also isn’t exposed to correction. That’s even if the judge wants to correct it. The system isn’t built for it. The cost of being wrong and being right is the same: $free.99.

There’s no incentive to push toward more care in these evaluations because, from the judge’s perspective, care has no distinguishable payoff compared with carelessness.

How Do Courts Make Judges Pay for Bad Rulings?

Quick Summary: If someone loses a trial, they can take their case to federal court and ask for review. If an appellate judge changes the decision, the case goes back to the original judge. Each year, tens of thousands of these appeals move through the federal courts.

Compared to a real court case, with all its flaws, a judge who rules badly enough and often enough can be reversed on appeal. In an ideal situation, the cost of a bad verdict is designed to eventually come back to the judge who issued it.

Why Do Everyday Verdicts Get No Appeal?

Quick Summary: Everyday assessors of strangers answer to no appeal, no oversight, and no public reputation. Nobody tracks a rater’s hit rate, so the feedback loop never closes. The person on the receiving end holds all the risk while the assessor moves on.

Verdicts outside of the courtroom don’t work the same way. The trials we go through daily are never public, which means there is no appeal, no oversight, and no reputation at stake for the judge. No one tracks the judge’s accuracy at all. The only thing that gets tracked is the outcome the defendant has to endure because of the verdict.

Nassim Nicholas Taleb calls this asymmetry skin in the game. Decision makers who carry no downside from a decision produce errors that others absorb. (Source: Nassim Nicholas Taleb, Skin in the Game: Hidden Asymmetries in Daily Life, Random House, 2018.) Every unappealed verdict runs on this asymmetry.

In none of the examples given does the person rendering the verdict pay a cost for giving it. I can multiply these examples by many more.

It’s not because people are unusually cruel or careless with their judgments. It’s because that judgment costs nothing, and the system has no mechanism to correct this flaw.

A verdict that costs the judge nothing means verdicts will continue to be rendered as fast and as thin as they always have been. Nothing about these judgments gives the judge any reason to slow down.

How to Track and Correct Your Own Judgments of Other People

Date	Call	Evidence seen	Confidence	Outcome	Result	Rule Change
Sept 3	The server is careless	Water arrived late	4 of 5	A kitchen fire slowed all tables	Wrong	Ask staff what happened before rating anyone

Check your record once a month

After a month, go down the page and count each Right, Wrong, and Unknown. Begin with the calls you circled five times, the ones you felt confident about. Notice when those guesses miss just as often as the ones you were unsure about. Your sense of certainty fades. The page lets you see what you believed and what really happened, side by side.

Questions and Responses