The false consensus effect: believing everyone thinks as you do
We regularly overestimate the proportion of people who share our positions. This error of estimation has concrete effects on the way we debate.
A bias of estimation, not of conviction
The work of Lee Ross and his colleagues, at the end of the 1970s, brought to light a robust tendency: asked about the proportion of people sharing their choice, people estimate it higher than it is, and this whatever choice they made.
It is therefore neither arrogance nor a defect of character. The simplest explanation is available: we estimate a population from the sample we know, and that sample is made up of people who resemble us more than the average.
The bias is moreover stronger on questions where we hold a strong position, and weaker on those that leave us indifferent. It therefore grows precisely where it does the most damage.
What the bias produces in debate
If one believes one's position widely shared, one concludes that it is obvious, and that a particular reason is therefore needed not to hold it. That particular reason quickly takes the form of a doubtful intention or a lack of information in the other, which closes the discussion before it begins.
The bias also produces surprise, sometimes massive, at election results or opinion surveys. The surprise is proportional to the homogeneity of one's circle, which makes it a useful indicator: the more a result astonishes you, the more it tells you about the composition of your personal sample.
It finally feeds a frequent strategic error: believing that it is enough to mobilise in order to win, since the majority would already be secured. Campaigns built on that belief neglect persuasion, which is precisely what they would need.
The reverse side, less well known
It also happens that one underestimates support for one's own position, generally when one knows it is poorly regarded in one's immediate circle. This underestimation is another way of confusing one's circle with the population.
Its effects are different and sometimes heavier: it leads one to keep quiet about an opinion that is nonetheless shared, which makes it even less visible, and thus feeds its own apparent disappearance. Work on the silence of supposedly minority majorities describes this circle.
A simple corrective
Before looking at a survey result, write down your estimate. The gap between your figure and the observed figure is a direct measure of the distortion of your sample, and that measure is more instructive than the result itself.
The exercise has an additional merit: it works in both directions and can be repeated. After a few attempts, one learns on which subjects one is systematically wrong, which is rare and personal information.
What this changes for reading a poll
A well-read poll is the best known remedy for this bias: it is one of the few ways of knowing what a population thinks rather than what one's circle says of it. That is also why results contradicting our intuition deserve more attention than those confirming it.
The practical rule is simple. A figure that surprises you should first be checked in its wording, then, if it holds, taken seriously as information about your own position within the population.
Key points
We overestimate the share of people who share our positions, whatever they are, and all the more when they matter to us.
The explanation lies in the composition of our circle, not in a defect of character.
The reverse side exists: underestimating real support leads one to keep quiet about an opinion, and so to make it invisible.
Estimating in writing before looking at the figures directly measures that distortion.
This article describes documented mechanisms and disagreements. It recommends no position, no party and no vote, and commits to no reading other than that of the work it cites.
And you, where do you stand?
The Civix test situates your convictions on six dimensions and matches you with one of the twelve profiles, without ever telling you what to think.
Start my test