I agree with piq; partial information can be harmful.
Imagine a hypothetical situation (I call it hypothetical because I don't actually know the correct numbers for the data).
Datum A: One person per year is killed by an airbag.
Datum B: One thousand people per year are saved by an airbag.
Person X does not know A or B.
Person Y knows A and does not know B.
Person Z knows A and B.
Person Y's best information is that airbags are dangerous. This may lead person Y to disable the airbags.
Two things to note: First, person Y is in a worse situation than either person X or person Z. Partial information has led person Y to do something which he/she would have been better off not doing. So X, with no information, is better off than Y, with partial information. The second thing to note is that person Z is in the best situation: X has no reason to turn off the airbag, and Z knows not to turn it off, so they're both in the same situation as far as that goes, but Z has made a conscious decision about not turning it off. Z cannot be misled by Y; X could be.