Brilliant, Intuitive, and (Possibly) Wrong
August 3rd, 2011
August 3rd, 2011
Economists have some funny ideas. Often these ideas are controversial. Sometimes these ideas are brilliant and intuitive. Other times these ideas are brilliant, intuitive, and just plain wrong. It’s often difficult to know the difference between the two. It usually takes another economist to figure it out.
Economists have long held the (brilliant, intuitive, and possibly wrong) belief that people make the best decisions for themselves. This is called rational choice theory (or, simply, rationality) and it underlies almost all economic theory. This confuses a lot of non-economists who understand “rationality” to mean “sane” or “making good choices.” Rationality to economists just means that individuals balance costs and benefits of each decision to arrive at the choice that will give them the most utility or happiness (if you’re interested in the graphical representation of this model, check out this explanation). In this framework, someone who chooses to give up work for play is rational because they value the fun activity more than the money. Someone might also give up a vacation inItaly to work over the weekend because they value the money more than the trip. A drug user is rational. So is a spendthrift. That’s not to say any of these are “good” decisions. Just “rational” ones.
So from this theory, it follows that the best way to help someone is not to give them food or clothes or medicine, but rather cold, hard cash. In a perfect world, this would lead to the most efficient outcome as individuals are able to decide for themselves what to buy, instead of allowing an external agency or government or organization decide for them. A poor person might really need a pair of winter boots, but receive a cow instead. A woman in need might truly want a new dress for her job interview, but receive a vaccine instead. An impoverished man might want to buy a new roof for his house, but receive a bag of rice instead. Giving a person cash bypasses these problems.
It should come as no surprise, then, that is was a group of economists who developed the nascent charity GiveDirectly. The (brilliant, intuitive, and possibly wrong) idea is simple: give poor people money so they can buy exactly what they want instead of what the charity decides they should want.
Obviously the main concern (and the oft cited reason for why aid is not done this way) is that recipients will go and spend their awards on choices that are self-harmful, like alcohol. The charity (and I, too) have a bit more faith in humanity than that, but there is compelling evidence to support our convictions. For example, two studies of cash transfers, one in Nicaragua and one in Colombia, found no effect on the purchase of alcohol, tobacco, and entertainment. Another study of cash transfers in Mexico actually found they have a negative effect on the share of the budget a household spends on alcohol and tobacco. If you’re interested in reading more about the evidence behind the theory, this excellent paper reviews the recent economic literature on the topic.
There are other drawbacks, though. GiveWell, a charity evaluator, points out that “giving out cash, transparently, to some people and not others within a community” could incite “jealousy and conflict…and if power dynamics are imbalanced enough,” could permit “more powerful village members [to] take the benefits away from others.” This is quite a valid criticism.
As an economist, the idea appeals to me. As a campaigner for transparency, I would note the charity reduces the risk of corruption by giving money directly to recipients, instead of funneling it through public or private agencies. But I am cautious to endorse it. I do not see this charity as a substitute for traditional aid. Beyond the concerns listed above, it does not effectively provide people with all of the political and personal necessities for solving endemic poverty like education, empowerment of women, and better governance. But it is a compelling notion. And it’s so intuitive, it just might work.