Depreciation of rental buildings (the land is excluded) is just a tax timing issue and reduces the basis in the property, so it's eventually recovered via recapture.

The paper you cite takes as an assumption a monopoly of landlords. It does not prove that's the case. "It should be clarified that our model here is one where...landlords have monopolistic power. ... We do not claim that this case is in any way closer to the reality of the housing market than the perfect competition assumption."