Excellent points from Robert Lerman:
Bad terminology can create bad policy. Nowhere is this more evident than in housing policy.
This approach is problematic. First, subsidies tied to specific “low-income” homes substantially restrict where recipients can live, and what’s available may be a poor choice for their families. Second, the cost of subsidizing construction programs is higher than the cost of boosting people’s purchasing power to rent or buy their own dwellings, even assuming the construction units last at least 30 years. For both reasons, the government’s cost is often far higher than the recipient’s benefit.
Even the much-vaunted low-income housing tax credit, endorsed strongly by the New York Times editorial board, is costly and does little to expand housing supply. The tax credit aims to encourage developers to invest in affordable housing. They sell the credits to investors, lowering the amount they need to borrow to build or fix up property. But developers generally sell their tax credits at a discount, leaving them with only about 70-75 percent of the government subsidy. To advocates of these programs, the subsidies add to the stock of “affordable” housing. But, as research has shown, the added housing financed by government is largely or completely offset by less private-financed housing.
By shifting from construction incentives to rent vouchers, the government can save 20 percent or more on its current housing outlays, meaning it could offer vouchers to many more low-income families at the same costs. Moreover, as shown elsewhere, if the new vouchers emphasized homeownership instead of renting, the government’s costs would be even lower—which could mean even more available vouchers and more families covered by subsidized housing. That’s quite a benefit from thinking more clearly about the housing of low-income families!
We could scrap the mortgage interest deduction and convert it to a fixed price housing voucher for everyone.