The compact, nonthermal radio source Sgr A* at the Galactic Center appears to be coincident with a ~ 2.6* 106 Mo point-like object. Its energy source may be the release of gravitational energy as gas from the interstellar medium descends into its deep potential well. However, simple attempts at calculating the radiative spectrum and flux based on this picture have come tantalizingly close to the observations, yet have had difficulty in accounting for the unusually low efficiency in this source. Regardless of whether the radiating particles in the accretion flow are thermal or nonthermal, there now appear to be two principal reasons for this low conversion rate of dissipated energy into radiation: (1) the plasma separates into two temperatures, with the protons attaining a significantly higher temperature than that of the radiating electrons, and (2) the magnetic field B is sub-equipartition, which reduces the magnetic bremsstrahlung emissivity, and therefore the overall power of Sgr A*. In this paper, we investigate the latter with a considerable improvement over what has been attempted before. In particular, rather than calculating B based on some presumed model (e.g., equipartition with the thermal energy of the gas), we instead infer its distribution with radius empirically with the requirement that the resulting spectrum matches the observations. Our assumed ansatz for B(r) is motivated in part by earlier calculations of the expected magnetic dissipation rate due to reconnection in a compressed flow. We find reasonable agreement with the observed spectrum of Sgr A* as long as its distribution consists of 3 primary components: an outer equipartition field, a roughly constant field at intermediate radii (~ 103 Schwarzschild radii), and an inner dynamo (more or less within the last stable orbit for a non-rotating black hole) which increases B to about 100 Gauss. The latter component accounts very well for the observed sub-millimiter hump in this source.