Conservatism bias

She had taken the same highway route to the coast for two summers running and it had always been fast. This weekend, the traffic app reported major construction delays on that road. She acknowledged the warning, glanced at the alternative route, and then chose the highway anyway. The old route had been fast before; surely it would clear up.

It did not clear up. The journey took twice as long.

The information had been available. She had seen it. She had simply weighted it at a fraction of what it deserved - not enough to change the decision, just enough to briefly consider it. Her prior belief in the route's speed had been updated, but only a little. Much less than the evidence warranted.

What It Is

Conservatism bias is the tendency to insufficiently revise one's beliefs when presented with new evidence. When new data arrives, people adjust - but the adjustment is smaller than a rational, proportional update would require.

Conservatism bias results in insufficient adjustment of beliefs toward new evidence. This is not stubbornness or closed-mindedness in a general sense. It is a specific, quantitative failure: the magnitude of belief change is too small relative to the evidential weight of what has arrived.

The Mechanism

The process of belief updating requires holding a prior belief, weighing incoming evidence against it, and revising accordingly. The revision step is where conservatism operates.

Several factors constrain the update. Anchoring to the initial belief creates a pull toward the starting point. Processing new evidence thoroughly requires cognitive effort - especially if the evidence is complex or arrives in an unfamiliar format. There is also a desire for consistency: a large belief update can feel like admitting the prior was badly wrong, which carries its own psychological cost.

The result is a conservative shift - real but insufficient. Beliefs change, but they move a shorter distance toward the evidence than they should.

Where It Shows Up

Medical diagnosis. A clinician with an initial working diagnosis receives a test result that substantially supports a different diagnosis. Rather than revising strongly toward the new diagnosis, the adjustment may be modest - the new result influencing rather than overriding the prior judgment. Conservative updating in this context means patients can remain attached to less accurate diagnoses longer than the evidence warrants.

Financial analysis. An analyst holding an earnings estimate receives quarterly results showing significant deviation. Rather than revising the forecast substantially, the adjustment is minor - the prior estimate retaining most of its weight. Conservatism bias in financial contexts contributes to systematic underreaction to earnings surprises.

Forecasting. When base rates or new trend data suggest a clear shift, forecasts often move less than the evidence warrants. Prior forecasts carry more weight than rational updating would assign them.

Everyday planning. Old routes, familiar methods, and established assessments resist revision even when incoming information about delays, changes, or new conditions clearly warrants a larger update.

What It Is Not

Conservatism bias is not the same as confirmation bias. Both involve insufficient responsiveness to evidence, but through different mechanisms. Confirmation bias concerns which evidence a person seeks and how they interpret it - filtering toward what confirms existing beliefs. Conservatism bias concerns the magnitude of update - how far the belief moves when evidence is processed. A person can process disconfirming evidence without bias and still update too conservatively. The problems are distinct and can operate simultaneously.

It is also not simply a general resistance to change. Conservatism bias is a specific quantitative under-updating - it operates even in people who are fully aware that new evidence has arrived and who genuinely intend to update.