Anchoring bias is the tendency to rely too heavily on the first piece of information you receive on a topic, using it as a reference point for later judgments regardless of whether that information is accurate. The first number becomes the anchor, and your final estimate tends to sit close to it because you adjust away from the starting point only part of the way. This happens even when the starting number is obviously arbitrary or irrelevant to the question at hand.
What Anchoring Bias Is
An anchor is any initial value or impression that sets a reference point for a judgment you make afterwards. Anchors can be external, such as a suggested retail price on a shelf, or internal, such as a reference point drawn from your own beliefs, experiences, or contextual clues.
The important feature is that the anchor shapes the answer even when it carries no real information about the correct value. Initially presented numerical values systematically influence later judgments of uncertain quantities, and they do so even when the numbers are clearly arbitrary.
Why the First Number Sticks
Adjustment That Stops Too Early
The classic explanation is the anchoring-and-adjustment heuristic. You start from the anchor, adjust step by step toward what seems right, and stop as soon as you reach an answer that feels acceptable, so the adjustment is often insufficient. The mental walk away from the anchor simply ends before it has gone far enough.
A Competing Explanation
A second account holds that the anchor automatically triggers a positive test strategy: an intuitive search for evidence consistent with the anchor as a possible answer. On this view, you are not adjusting at all but gathering reasons the starting number might be right. There is no consensus on which mechanism is correct, and both are used to explain the phenomenon.
A Classic Demonstration
In the well-known experiment on the percentage of African countries in the United Nations, one group was shown the number 10 and another the number 65 before estimating. The first group estimated about 25 percent, and the second about 45 percent.
Nothing about 10 or 65 revealed the answer, yet the two starting points pulled the final estimates roughly 20 percentage points apart. This is the core pattern: an arbitrary number shifts a judgment that should be independent of it.
Where Anchoring Shows Up in Everyday Decisions
Shopping and Pricing
An initial price can set the standard for everything that follows. In a used car negotiation, the first price mentioned frames the rest of the discussion, so later lower prices feel like good deals even when they still exceed the car's actual value. A "was" price or a manufacturer's suggestion works the same way as an external anchor.
Salary and Negotiation
The first salary figure mentioned in a job offer can anchor the entire negotiation, encouraging a job seeker to evaluate later offers against that figure rather than against their own target. A low opening number can quietly redefine what feels like a reasonable outcome.
The same pattern appears in group settings. When a proposal or figure is raised early in a discussion, groups tend to refer back to that anchor when assigning sentences, setting policies, hiring, or negotiating with contractors. This bias can shape decisions made by individuals and groups alike.
Estimates and Time Planning
Anchoring also affects how long we expect tasks to take. People often anchor to an initial time estimate and resist adjusting it even when the task clearly runs longer. The first guess becomes a plan, and later evidence struggles to dislodge it.
Reported Findings in Professional Settings
Explainers on this bias report that anchoring can sway prison sentences assigned by jurors and judges, and that clinicians may stick with initial impressions even as new information becomes available. These are reported findings in secondary accounts rather than fully documented specifics, so treat them as indicators of scope rather than measured effect sizes.
When Anchors Matter More or Less
Not all anchors work equally, but the range of influence is wider than most people expect.
- Implausible anchors still shift judgments. The effect is pervasive enough that highly implausible anchors, such as asking whether Gandhi lived to be 140, can affect later estimates.
- Moderate anchors can pull harder than extreme ones. The relationship between anchor distance and effect size appears curvilinear, with moderate anchors producing stronger effects than extreme ones.
- Direction matters. Larger anchoring effects have been observed for anchors set above the correct answer than for anchors set below it.
- Knowledge is protective but not absolute. The more you know about a topic, the less likely you are to fall for anchoring; when you lack the information to value something properly, anchors exert more influence.
- Individual differences exist but are modest. In a group of more reflective subjects, a substantial negative correlation between intelligence and anchoring was detected. Yet measures of susceptibility to anchoring shared only a small portion of variability with intelligence, cognitive reflection, and basic personality traits, so being thoughtful or intelligent is no guarantee of immunity.
Anchoring can also cause people to overlook new, relevant information and rely instead on an initial anchor that may have little to do with the decision.
What the Research Supports and Where It Is Contested
A Robust Core Effect
The anchoring effect is described as one of the most reliable results of experimental psychology. It has been demonstrated across many domains, in both laboratory and real-world settings, and among both novice and expert subjects.
The foundational work belongs to Tversky and Kahneman's 1974 paper on judgment under uncertainty, which identified anchoring as one of three heuristics that systematically distort judgment alongside representativeness and availability.
Where the Evidence Gets Thinner
That foundational research was conducted mainly through laboratory experiments using abstract probability tasks, word problems, and hypothetical scenarios with no domain expertise at stake. Transferring those demonstrations to naturalistic professional judgment in high-stakes settings is contested, because domain expertise, feedback richness, and deliberate calibration can reduce or amplify bias effects in ways the original controlled studies did not model.
Research on why adjustments are insufficient, including work associated with Epley and Gilovich, is often cited in this area, but the practical takeaway is narrower than the headlines suggest: the mechanisms are still debated, and the size of anchoring in any given real-world decision is not established by the evidence discussed here.
How to Reduce Anchoring, and Its Limits
No single technique reliably removes anchoring in every situation. That said, a few practical habits address the mechanism directly.
- Form your own number first. Before you see a price, an offer, or a suggested figure, commit to an estimate or a target based on your own information. The heuristic model predicts that adjustment stops early, so starting from your own value removes the anchor rather than asking you to correct for it.
- Ask whether the anchor is relevant. The starting number is often arbitrary or unrelated to the real question. Naming that irrelevance out loud makes it harder for the number to feel like evidence.
- Set a specific target before negotiating. One explainer recommends setting specific, ambitious goals beforehand to improve negotiation outcomes. Treat this as sensible framing rather than a guaranteed outcome, since the supporting detail is limited.
- Seek disconfirming information. Because one explanation is a search for evidence consistent with the anchor, deliberately looking for reasons the anchor is wrong works against that pull.
- Build knowledge where it matters. Anchors hold less sway when you genuinely know the topic, so research before high-value decisions rather than after.
These steps reduce exposure and give your judgment a fairer starting point, but they do not make you immune. Even knowing about anchoring does not reliably eliminate it.
Anchoring Versus Related Shortcuts
Anchoring is often confused with other judgment shortcuts, and the distinctions help you name what is happening.
| Shortcut | What Sets the Reference | Where the Bias Enters |
|---|---|---|
| Anchoring bias | A first number or impression | Adjusting away from it insufficiently |
| Confirmation-style search | The anchor as a hypothesis | Seeking evidence consistent with it |
| Framing | The wording or context around a choice | The frame, not the facts, drives the evaluation |
These can work together in a single decision. In a salary negotiation, for example, the opening figure anchors you, you then look for reasons it is reasonable, and the language of "the going rate" frames it as normal. Anchoring is the specific pull of the first number, while the others describe how the pull is sustained or packaged.
What to Watch for Next Time
The practical rule is simple: notice the first number you see, and decide whether it deserves to be your reference point. Before a purchase, a negotiation, or an estimate, write down your own figure first and treat any incoming number as a proposal rather than a fact. Anchoring is one of the most reliably demonstrated effects in this area of psychology, so the habit worth building is not eliminating the first number but refusing to let it make the decision.
