Biological molecules are easily damaged by high-energy electrons, thus limiting the exposures that can be used to image such specimens by electron microscopy. It is argued here that many-electron, volume-plasmon excitations, which promptly transition into multiple types of single-electron ionization and excitation events, seem to be the predominant cause of damage in such materials. Although reducing the rate at which primary radiolysis occurs would allow one to record images that were much less noisy, many novel proposals for achieving this are unlikely to be realized in the near future, while others are manifestly ill-founded. As a result, the most realistic option currently is to more effectively use the available "budget" of electron exposure, i.e. to further improve the "dose efficiency" by which images are recorded. While progress in that direction is currently under way for both "conventional" (i.e. fixed-beam) and scanning EM, the former is expected to set a high standard for the latter to surpass.