FIP measures a limited set of stats that correlate to what is thought to be in control of the pitcher (e.g, walks, strikeouts, homers.) It ignores balls put into play (BABIP), because those are assumed to be outside of the pitcher's control.
https://www.baseball-reference.com/b...ndent_Pitching
ERA measures all outcome and assigns it all to the pitcher.
By inference, (excluding random luck), a pitcher whose ERA is higher (lower) than his FIP is being hurt (helped) by his defense. E.g, an average pitcher pitching in neutral parks with zero luck effects, would be expected to have a higher ERA than his FIP if he played for a poor defensive team.
Of course, one might argue that a pitcher who is thought to induce batters to hit ground balls more than average (aka, a ground ball pitcher), should also show the same pattern, (his ERA is higher than his FIP, luck excluded and playing for a neutral defense), but that seems to be attributed to the defense, not the pitcher. The theory (based upon regression models), seems to imply that there aren't any true groundball pitchers and that over a long term, any such "groundballer" are in fact just pitchers that have been pitching for teams with above average infielders grabbing more ground balls than average. So, if your infield was composed of Brooksie, Ozzie, Maz and Vic Power, you'd get many more ground ball outs than average, regardless.