The observation that superconductors have predominantly positive Hall coefficients was made already in 1962 by Chapnik. The discovery of superconductivity in hole-doped oxides in 1986, and the measurement of a positive Hall coefficient in electron-doped oxides in 1989, further underscore the existence of a distinct asymmetry between electrons and holes with respect to superconductivity, which calls for an explanation. No such explanation is provided by most proposed superconductivity mechanisms including the electron-phonon interaction. Here we discuss the simplest possible local interaction that breaks electron-hole symmetry in solids. This interaction causes holes to be heavier than electrons, and is repulsive for electrons and attractive for holes. Thus, it provides an explanation for the above observations. It leads to superconductivity in the absence of any other attraction mechanism, and the resulting theory exhibits many characteristic features observed in the high Tc oxides. We review the predictions of the theory and their relationship with existing and future experiments in high Tc oxides. A brief discussion of the implications of this mechanism for the understanding of superconductivity in "conventional materials" is also given.