It should be no surprise that people tend to think they are better than average, or at least that they think they are better than they truly are. The only possible exception I can think to this optimism are Southern European who seem to constantly self-deprecate themselves. In some ways this is disturbing news, for example when we think about rational expectations: our theories say people should not be making systematic mistakes.
Eugenio Proto and Daniel Sgroi go even further. They find that people perceive the distribution of the others to depend on their own place in the true distribution. In probably better words: If you are in the tails of the true distribution, you perceive to be closer to the center than reality. I think the most likely explanation for this has to do with the fact that we do not interact with the complete distribution of people, but most likely those that are most like oneself. I tend to interact mostly with highly educated people, because of my profession, and thus I tend to think most people are well educated, notwithstanding my disappointments when teaching undergraduates.
With respect to rational expectations, the next questions is whether these personal biases cancel out in the aggregate. Possibly, but this misses the true concern. While most aggregates may not be affected by this kind of bias, the distribution of outcomes may, and this could even affect some aggregates (say, savings, human capital). How much? I really do not know.