[vasilipupkin]

r > g is definitely a force for divergence. my intuition was that it is empirically unrealistic to expect r > g and Rognile's paper seems to provide quantitative evidence for this intuition