GiveDirectly has been recognized recently for devising a new model of giving in which a person donates money and almost the entire sum sent via cell phone to a recipient that needs it. I had the opportunity to see this model up close while in Kenya, and have been very impressed with its use of technology to deliver instant results with very low overhead. Compared to [name your favorite bloated high-overhead aid organization], GiveDirectly delivers significantly higher impact and ensures that the money is channeled to the intended recipients. For this reason, GiveWell and Google Impact Awards have both recently recognized GiveDirectly for its innovative approach.
Most of the conversation around GiveDirectly seems to imply that providing money directly to people will bypass the paternalism of organizations hoping to channel funds to what they think poor people need. GiveDirectly says that it hopes to have better outcomes metrics than any other aid model, as described by Sasha Dichter’s (as always) excellent blog post.
However, this argument seems to fly in the face of most of our understanding of international development for three reasons:
(1) People are rationally irrational
Research shows that people don’t invest in new (better) seed and crops, buy or use preventative health products, put enough money in savings, and do other things that are thought to be good for the same reason many people in America spend their tax refunds on electronic gadgets instead of a retirement fund. These actions aren’t rational, but people have really good reasons for doing them.
People who receive a large unanticipated sum of money will likely spend it on that big-ticket item that they’ve been lusting after. This could mean replacing the thatched roof with tin, buying a motorcycle, investing in a few cows. If this is what the recipient wants, maybe it’s best to let her choose. But there is ample evidence that certain investments lead to better long-term outcomes but less short-run pleasure. People often make the rationally irrational decision to choose the item that will benefit them most tomorrow, not 10 years in the future. In those cases, the donor loses the opportunity to invest in something with a proven impact and channel the funding where it might be better allocated.
(2) GiveDirectly’s model doesn’t address barriers to access
Melinda Gates gave a great Ted Talk about how aid organizations could learn from Coca Cola’s distribution method – Coca Cola can be found in the most remote villages, but many life-improving and high-impact products are only distributed in larger cities and towns. GiveDirectly recipients will spend their money on products that are readily available, which means that to improve demand and access to life-improving products, distribution networks should be expanded in tandem with mobile giving.
Microfinance banks do this well – they often couple distribution of loans at bank meetings with special deals on products like cookstoves and solar lamps to channel money towards these high-value investments. Until we can guarantee that deworming tablets, clean cookstoves, and safe water technologies will be in a fair fight against brown sugar water and satellite television, a little paternalism might be helpful.
(3) The best donations are investments
As the two reasons above indicate, poverty is the result of intractable problems on both the demand and supply side, and the best solutions to these problems often don’t yet exist. New ideas will generally arise through experimentation, creativity, and a commitment to innovation. But no organization that engages in innovation and R&D can guarantee that its investment will pay off, making the process inherently risky.
Giving money to a person rather than an entity limits possible impact to the existing suite of goods and services for that community. People will spend the money on existing things (or savings products) not on novel solutions to old problems. A better long-term use of your money might be to send it to someone that will use it for creative need-finding and problem-solving.
GiveDirectly is a far better option than most aid models because it ensures that beneficiaries receive the money, but I don’t think that it is an ideal solution. Though other organizations might not be able to prove that an equivalent amount is channeled directly to recipients, there is something to be said for problem-solving, need-finding, and embracing risk as part of the process of development. All of this is to say that my money this holiday season is going to entities that combine high-impact ways to spark demand for necessary but unpopular products, address distribution barriers, and experiment with new and more cost-effective ways of addressing age-old problems.