Casey Mulligan complains that, in my preceding post, I misread his article because the first and last paragraphs mention the nontaxation of imputed rent point. The references are perhaps a bit oblique, but I take his point and apologize for the snark.
That said, I still find his analysis inadequate because I really don't think one can assess the home mortgage interest deduction independently of the imputed rent issue. Denying the interest deduction would be an indirect partial repeal of the exclusion, partly reducing its scope (good) but also potentially creating fresh distortions that would not follow from a more direct, even partial, repeal (bad).
I think the good from denying the deduction would outweigh the bad. I would merely have disagreed with Mulligan, rather than finding his analysis as snark-worthy as I did, had he addressed this issue. But, in my view, treating the under-taxation of home ownership in other respects as effectively irrelevant to analysis of the home mortgage interest deduction was a disservice to the readers of his NYT blog entry.
UPDATE: To put it another way, Mulligan says that the Treasury doesn't lose revenue from the mortgage interest deduction as such, assuming deduction and inclusion by the borrower and lender at the same marginal tax rate. OK, fine. But the Treasury does lose an opportunity to recoup some of its losses from the imputed rental exclusion, and that's the whole point of the argument for mortgage interest disallowance.