We have produced a high-resolution global admittance map of Venus using a spatiospectral localization technique, calculating 64,800 admittance spectra evenly spaced in latitude and longitude. We grouped the admittance spectra into 35 classes that we inverted for elastic, crustal, and lithospheric thickness using bottom-loading, top-loading, and short-wavelength top-loading models with long-wavelength bottom-loading ("hot spot") models. Our best fit models identify 17 classes consistent with top loading (40% of the surface), 7 with bottom loading (39%), and 4 with a long-wavelength bottom-loading model (15%). The remaining seven classes (6% of the surface) generally had large-amplitude top-loading signatures that could not be fit with our models. For the classes that were successfully modeled the estimated range of crustal (0--90 km) and elastic thickness (0--100 km) is larger than results for most previous studies, primarily because of the incorporation of bottom loading. Nearly half of the planet has elastic thickness <20 km, which we cannot distinguish from isostasy, suggesting that these regions are tectonically inactive. A key finding is that the lithospheric properties typically vary over scales of 1000 km or less except in some plains and crustal plateau regions. Both the scale and range of crustal and elastic thickness variations indicate that Venus remains an active planet and that geologic processes may be more complex than previously recognized. In many regions the correlation with surface geology is not obvious, and this suggests that deformation of the lower lithosphere and crust may occur with little surface manifestation.