From an interesting paper by Alan G. Gilbert (Habitat International, 2014):
The argument is not about whether or not the government should devote money to helping the poor in a region where social spending is generally very low. The issue is whether the money could be better spent in other ways. Social spending could be used to subsidise poor families’ consumption of water or reducing the cost of their using the public transport system. Alternatively, some would consider larger income supplements to be preferable to offering the poor free housing. The former provide a current income flow rather than a potential capital gain and can be used by the family to match their specific needs - education, health, better housing or even to eat. Of course, income supplements can be misused - they can be gambled away or used to get drunk. But some evidence exists that such irresponsible behaviour can be reduced by distributing the funds to only to mothers of families and by making payments conditional on certain kinds of behavior (Hanlon, Barrientos, & Hulme, 2010; Rawlings & Rubio, 2005). (p. 260).
From the abstract:
The article is concerned particularly with the question of whether providing free housing is the most effective way of helping the desperately poor. It is concerned with evidence that previous programmes to subsidise the cost of formal housing have not helped the poor. In particular, they have provided poor quality accommodation, failed to provide adequate services or a decent living environment, and have sometimes contrived to create the problem neighbourhoods of the future. Offering families a home for nothing does not solve the fundamental problem facing the poor - their very low incomes.