Absolutely. The problem with these kind of studies is how difficult it is to generalise from them. One of the cited examples concludes making people accountable for the money was less efficient even though accountable people spent the money better because it cost "up to $1800" per $150 loan for followup studies. That hires a well-educated person for a year in a developing country, which strongly suggests that you'd reach a wholly different conclusion if the programme was designed and administered 100% locally. Of course, there's also the issue that if a particular group doesn't have any followup studies associated with the research, then observations of how they spent the money are inevitably less accurate.
Then there's the simple observation that repayable grants can be passed on to others (less admin costs) and are sustainable on a larger scale in an economy. Teach a man to fish and his interest repayments can teach his neighbours to fish, provided those neighbours also reasonably expect to get a positive return on the rod they invest in. That latter bit is important in assuming the net effect doesn't involve very expensive fishing rods and depleted fish stocks...
Whilst loan based aid is not unproblematic (there's some hideously expensive administration and profiteering in some microfinance problems, and poor people many people do use them as expensive consumer loans) the same technology that earns Kenyan grants such rave reviews for their effectiveness can also be applied to strings-attached microfinance: M-PESA can potentially cut the cost of receiving microloan repayments even more effectively than it cuts corruption in doling out the aid. Just because the desperately poor don't spend all their grant money on prostitutes and alcohol (which isn't that surprising if you've ever visited a developing country) doesn't necessarily mean they can't be efficiency helped in other ways
Then there's the simple observation that repayable grants can be passed on to others (less admin costs) and are sustainable on a larger scale in an economy. Teach a man to fish and his interest repayments can teach his neighbours to fish, provided those neighbours also reasonably expect to get a positive return on the rod they invest in. That latter bit is important in assuming the net effect doesn't involve very expensive fishing rods and depleted fish stocks...
Whilst loan based aid is not unproblematic (there's some hideously expensive administration and profiteering in some microfinance problems, and poor people many people do use them as expensive consumer loans) the same technology that earns Kenyan grants such rave reviews for their effectiveness can also be applied to strings-attached microfinance: M-PESA can potentially cut the cost of receiving microloan repayments even more effectively than it cuts corruption in doling out the aid. Just because the desperately poor don't spend all their grant money on prostitutes and alcohol (which isn't that surprising if you've ever visited a developing country) doesn't necessarily mean they can't be efficiency helped in other ways