This is something I have observed but do not fully understand myself. This is what I've pieced together. Chemical shifts are based on relative shielding/deshielding effects of the neighboring atoms. Inductively withdrawing groups such as halides pull electron density away from nearby atoms thereby having a deshielding effect. However, sometimes I have found that when a lot of strongly electron withdrawing groups with a lot of electrons are attached the reverse effect is seen. I attribute this to the electrons on those withdrawing groups effectively shielding the atoms in question despite the polarization of the bonds. Case in point, in 31P NMR, phosphites, P(OR)3, are found around ~150-170 ppm and phosphates, O=P(OR)3 are close to 0 (relatively shielded). Again, this is a rationalization based on an observation rather than understanding. Hope this helps.