Nice little natural experiment, recounted by Aid Watch:
Two NGOs working in Zambia, Oxfam GB and Concern Worldwide, tried a different approach: they handed out envelopes stuffed with cash—from $25 to $50 per month per affected family, with no strings attached. An evaluation found that common fears about cash transfers—that the cash infusion will cause inflation in the market, that the money will be squandered, or that men will take control of the money—were unrealized.
What did people buy with the money? The list includes maize, beans, salt, cooking oil, meat, vegetables, clothes and blankets, paraffin, transport, soap and body lotion, and lots of other mundane household items. They also loaned it to friends, used it to pay back debts, purchased health care, education and transport, and rebuilt their homes. Only a very small fraction of the money (less than .5%) was spent on “unproductive” items, like liquor for the men.
This approach is different from traditional aid programs that distribute items like seeds, fertilizer, medical supplies, or other relief goods. Instead of doing this, simply giving cash can be quicker, more cost-effective, and actually allow the recipients to spend the money on things that have the most value to them. This seems like a pretty obvious strategy, yet it is fairly uncommon in foreign aid programs, or even domestic social programs (e.g. we give food stamps rather than cash). I think there are two primary reasons for this.
1. Paternalism, a.k.a. the "I'm afraid you'll buy booze with it" syndrome: When we help people, we want to see results. Unfortunately, what we consider to be good results are not necessarily what the aid recipients would consider to be good results. So we give seed instead of the money to buy it, or we give an apartment in a housing project instead of an equivalent amount of cash. This is fine and good if the aid giver's preferences are perfectly aligned with the recipient's preferences. Of course, that nearly never happens.
2. Politics: The majority of foreign "aid" programs are really direct subsidies to local producers, because we force the recipients to use the aid money to buy from certain firms. This "tied" aid often leads to inefficiencies and poorer outcomes, but that only matters if you care about the aid recipients over the domestic firms. Often, that isn't the case. This also happens domestically, as conditionality on food stamp programs are considered by some to be little more than welfare for American corn farmers.
What's the take away? If you want to help people who don't have much money, then give them money and let them spend it how they please. They'll do a better job getting the goods and supplies that they want and need than we could do, and there is basically no overhead costs of shipping, storing, and distributing goods. Therefore, more of the aid money can go directly to the needy rather than funding a supply chain.
And if they decide to spend it on booze... so what? If you're a Zambian subsistence farmer and your livelihood has just been wiped out in a flood, you might need a stiff drink more than a bag of seeds. Ditto if you're an American autoworker whose factory has just been closed.
P.S. Aid Watch has also hosted a nice little spat between Dani Rodrik and William Easterly over the utility of industrial policy in fostering economic development. It starts here, then here, and finishes here.