Heifer international may be more impactful than GiveDirectly

Bill Gates has recently indicated that he would like to support the raising of chickens through heifer international.

Gates is known to be interested in the effectiveness of his donations, however some would doubt whether donations to Heifer International would be more effective than the Givewell recommended charity Givedirectly, which gives cash directly to the poor.

In the case of giving chickens to the poor, it is reasonable to assume that this is worse than giving cash, since poor people can fairly easily buy a chicken for themselves, therefore providing chickens takes choice and power away from beneficiaries.

I applaud those who raise such questions, and in general I would consider this a good question to ask, but I was prompted to write this post because I'm concerned that in some quarters, this line of reasoning may have gone too far this time.

In particular, this argument as it stands misrepresents the work that Heifer International does. In fairness, this is an understandable misunderstanding, given that the language Gates used in the article (e.g. he says that "I will donate—on your behalf via Heifer—a flock of chickens to a family in poverty")

A quick look at Heifer International's annual report and accounts suggests that their work does not involve just giving chickens (or other livestock) to the poor. It also involves
- providing training on planning projects, evaluations, and livestock management
- upskilling of local community animal health workers
- creating a culture of sharing and caring, and asking beneficiaries to "pass on the gift" to other people in their communities.

Contrasting this with GiveDirectly's work, it is fair to say that on some measures such as consumption, there is decent evidence that these improve. More comments on this below*.

However later research indicates that giving cash to the poor causes jealousy - in particular that the "decrease in life satisfaction induced by transfers to neighbors more than offsets the direct positive effect of transfers". So it seems on soft measures, it may do more harm than good.

This is in stark contrast to Heifer's aim to create a culture of sharing, caring and generosity (although conversely you could also argue that we have no evidence of their success)

The arguments that I have presented here are not yet enough to prove that Heifer International achieves more good than giving money directly to the poor. We don't yet know, for example, whether the extra money invested in training people to make the most of their chickens will indeed result in greater returns to those families (and, in particular, returns that are higher than taking the money that would have been spent on that training and just giving it to the poor).

I have also ignored the impact on animal welfare and the environment, although on balance I would suggest that many of the negative consequences that are common with regard to western consumers eating meat are less likely to apply here, or at least not to the same extent, since the chickens are unlikely to be battery farmed, and land is unlikely to be re-appropriated to house the chickens.

If you felt that you did not have enough information to conclude that Heifer International's work is worth supporting, this is reasonable - indeed Givewell's argument is broadly along these lines, and they note that there are some really proven interventions that they can say with confidence do good. Bill Gates, clearly, feels differently. Independently of Bill Gates's opinion, however, I would hope that this blog post convinces you that - while not proven - it may well be true that Heifer International's impact could be better than that of GiveDirectly.



------------------------------------------------------------------

* More comments on the impact of GiveDirectly's work
The impact on stress is less clear based on the article mentioned above, with the paper noting that "the average effect of treatment on cortisol levels is small and not significant" (cortisol levels being a biomarker for stress), although it appears that this confusion arises from the paper not being ideally suited to assessing the effectiveness of GiveDirectly's work, because this paper refers to giving cash to the poor in lots of different ways (differing amounts and frequencies) whereas GiveDirectly only gives large amounts (c $1000) on a one-off basis (prior to the universal basic income experiment that they are currently conducting), which seems likely to improve stress levels. It also seems that life satisfaction does improve.