A problem with observational equivalence: Disentangling the Renter Illusion hypothesis