I have blogged earlier that instead of granting home ownership rights, urban housing programs should focus on providing for an adequate supply of housing stock that can be rented out.
Apart from their poor quality, the commonplace problems of urban housing include ineligible beneficiaries, sale by originally allotted owners and their reversion to squatting elsewhere, and capture by local musclemen who in turn rents them out. Most unfortunately, since ownership rights gets allotted, the allotment process itself becomes high stakes and liable to be captured by the local leaders. There are numerous housing units completed but lying unoccupied in many cities for lack of agreement about the beneficiary list (and all this despite the presence of an original beneficiaries list!).
In view of the general migratory nature of urban poor, an approach that relies on allotment of ownership rights, upfront or deferred, cannot effectively address these problems. Instead, a policy of constructing and adding to the urban housing stock and then renting them out to migrant workers will help mitigate many of the aforementioned issues.
The fundamental attraction (and resultant incentive distortion) of ownership right on a housing unit is the opportunity to exit with a huge payout by selling it. This in turn generates a series of incentive distortions among all stakeholders - original allottees, prospective buyers with greater need, and local muscle-men - that results in inefficient outcomes. Rental housing eliminates this incentive straight away.
Apart from not having to make any upfront payment, the obvious advantage with rental housing is that though assured of inhabitation, the beneficiary is not inflexibly yoked to his house. In other words, it does not hinder labor mobility, both in search of better livelihood opportunities to other places and to larger houses at other locations in same city as incomes improve.
The monthly rent payment reduces the likelihood that a local leader can extort an additional ransom. Any additional payment by the beneficiary would then be, atleast partially, a reflection of their willingness to pay (and this premium on the rent will in any case be a much smaller amount, if the rent is reasonably close to market clearing value). Without any ownership rights, the allottees can only transfer habitation rights, which in any case is legitimate and the new occupant can in turn formally get himself registered and pay rents. Even an informal transfer, wherein the original allottee makes the rental payment while collecting a higher rent from the new buyer, is efficient in so far it allocates housing to those most in need.
Either way, if the rents fixed are market clearing, then the magnitude of such incentive distortions can be minimized. However, even if the rents are not high enough and the supply not adequate, such informal dynamics would only serve to allocate houses to those with the highest willingness to pay.
The only administrative challenge with this approach would be the collection of rents. Here too, I am inclined to the opinion that once the house is allotted upfront clearly on a rental basis (through rental vouchers or some other mechanism), unlike ownership conferred on them, the households are psychologically primed to make their monthly rental payments.
Critics would argue that these outcomes, while economically efficient are not fair, in so far as it would deprive the poorest and those with the lowest purchasing power. Here are three observations on that. One, if the quality of housing is with the most basic specifications, the actual occupant will always be one of the poorest. Second, there is no sharp socio-economic distinction between the poorest and the remaining poor, who are the most likely occupants of such houses. Third, given the population of large Indian cities, those belonging to this category are too large for any urban housing program to saturate demand. Therefore, even after building a massive stock of housing, atleast for the foreseeable future, large numbers of urban poor will always be in search of housing (and therefore the aforementioned possible incentive distortion effect due to them cannot be eliminated).
The other criticism about the original allottees moving elsewhere to squat does not hold much ground. Given their quality, only the poorest (and thereby eligible) will generally occupy these houses. If Ram, who sold the house to Ravi, and reverted to squatting, had not sold it, then Ravi would have been squatting. The critical issue here is only the net addition to the housing stock, the rate of which has to be higher than the rate of immigration of urban poor.
In any case, the public policy challenge was to design a policy that would be both more efficient and fair, while being more transparent and easier to administer, than the present one. It cannot be denied that even with all its residual problems, a rental housing scheme will have less likelihood of those aforementioned incentive distortions.
See also this article which explores how a rental housing program can be administered.