Chris Blattman (who everyone who reads development/aid blogs must surely follow by now) has an article on giving money to the poor. You must read it: here.
These are very important points. I regularly find myself struggling with others (and myself) around whether the work we do in development actually adds more value than just giving the money directly to the people we're trying to help. In all projects, in all sectors of development. Is it better to give someone money and let them use it learn what they want or to provide organised training sessions that may or may not achieve any learning? Is it better to buy someone's bead-work or just to give them the cash? Is it better to bring in volunteers to build a classroom in a poor community or just to hand over the funds? Okay, that last one is pretty obvious but a lot of the time it is really tricky to figure out what is better. I think there is value added through some of the other activities that the money is used for. I think there is pretty clear evidence that training, etc. may sometimes play a strong role in making the money/assets provided to poor families more useful and therefore increasing the impact. Some of this is in terms of tangible benefits, some in terms of improved knowledge, social capital and all the other fluffy things that are so hard to measure accurately.
But I think measuring the value-add that the organisation brings is equally important. Even if it's tough. And who and how the value of the training and other activities is determined is important. Ideally, the beneficiaries should be given the chance to say whether the additional activities that were paid for from that money are more valuable to them than just getting the cash. They're the ones who lose out on that money in exchange for the supposed value added by other activities. But that is tough to do. Or at least, it's tough for any aid agency/NGO monitoring and evaluation department to get away with doing. Plus there is the issue that some of the benefits may be at a community rather than an individual or household level. Chris Blattman is not a fundamentalist - he calls for more experiments. Which is great. It would be fantastic to see more of those. The aid sector needs more research to push it (us?) to reconsider whether our activities are worth the extra cost and the opportunity cost of not being able to put that money directly into the hands of the poor.
Full disclosure: Am somewhat affiliated with the international organisation mentioned in the article and do support their work but would love to see some counter-examples. Also note to self to encouraging more sharing of data.