“…The data indicate that a better way to urxJerstand variables relevant to statelocal relations may be to consider pditical variables and administrative variables as elements of a larger set of "governmental" variables (Morgan and England, 1984). Making this distinction, one can consider the states as being more or iess centralized on either or both a pditicai dimension and an administrative dimension (Stonecash, 1981a(Stonecash, , 1981b. Conceming the latter dimension, the firxlings may suggest that the administrative "exchanges" t)etween bureaucratic officials are more important than previously recognized.…”