Proper exposure would be simply a multiplication in linear space, so the shape would have to change as values get redistributed. What you are thinking of would happen if they did an addition to all the pixels. That would also change the exposure but would not correspond in any way closely to what happens in case you had changed the exposure in-camera.
In practice, Lightroom seems to have its own secret recipe for changing exposures and other params. It probably performs multiplication on the luminance channel of mid-tone values while compressing shadow areas and highlights to maintain a smooth distribution of tonalities. There are plenty of mysterious things that happen when looking at the Lightroom histogram and some of them still puzzle me too!