The assumption that the individual effects aren’t correlated with the independent variable(s) doesn’t imply that the individual effects are identical for every observation. Rather, it implies that their values are random (some negative and some positive) with no association with the observed values of the independent variable(s).