I see there is a kerfuffle about not including land in the Eggertson and Mehrotra’s “Models of Secular Stagnation“, which should raise the equilibrium rate to be negative. I want to make two comments here as a sticky:
1. Land is not a risk free asset. Just think of a Lucas Tree model but replace the tree with land. The tree gives variable fruit, so the land gives a variable return not a risk-free return. The result is that land must pay a risk premium and can have a positive real return every period even if the risk-free return is negative in every period (c.f. Summers et. al in”Assessing Dynamic Efficiency“.). The risk free return is the relevant return for MSS. That means with a high risk premium, you can add land or capital to the model and it still works. Moreover, I think it’s no accident that falling land values and are associated with this type of secular stagnation. Finally, this is why the MSS world need not be one of dynamic inefficiency even if it has negative risk free rates.
2. The issue isn’t really whether the risk free rate is negative forever. It’s enough to be negative for a many periods to get stagnation for many periods. So that means “low” returns from land/capital for a secular period of time are sufficient.