Here’s what we found. First, the correct information did change people’s estimates of the size of the foreign-born population. In one experiment, we asked people their estimates before and after seeing the information. Those estimates dropped 6 points among those who saw the information and 11 points among those whose estimates were explicitly corrected.

But this information did not consistently change attitudes about immigration — either people’s perceptions about the consequences of immigration or their views of whether immigration should be increased, decreased, or kept at its current level. Most notably, correcting estimates — that is, making them less exaggerated — did not make people more favorable to immigration.

This was also true in one experiment that focused on the local immigrant population, and provided correct information about the fraction of the population in the surrounding community.