Should GiveDirectly spread its grants more widely?

The new charity GiveDirectly has made quite a splash with its program of giving money to poor families in Africa through mobile phone transfers.  GiveDirectly has been featured on National Public Radio and in The Guardian, among other prominent news sites.  In 2013, GiveDirectly was rated as one of the top three charities by the charity evaluator GiveWell. 

I have happily added GiveDirectly to the list of charities I support.  I am troubled, however, that GiveDirectly makes large grants to a relatively small number of families, rather than smaller grants to a larger number of families.  GiveDirectly’s standard grant of $1,000 per family is on average more than 100% of the annual consumption of the poorest families it helps, according to GiveWell’s analysis.  The $1,000 grant is also more than 100% of the annual consumption of the poorest families GiveDirectly does not help.

The large grants distributed by GiveDirectly trouble me because I believe that those who are poorer are likely to benefit more from additional money than those who are richer.  This is an intuitive and widely-held view, even if it is not always enunciated.  If charitable donors did not believe that those who are poorer are likely to benefit more from additional money than those who are richer, they probably would not give charity at all

GiveDirectly’s grants are so large, in relation to the income and wealth of potential recipients, that they open a wide gulf between families that receive the grants and similarly-situated families that do not receive the grants.  To me, this suggests that GiveDirectly could do more good by splitting up its grants into smaller amounts – say, $250 per family – and distributing them more widely.  Distributing smaller grants to a larger number of families could also help to mitigate feelings of envy, disappointment, and perceived unfairness in families that are not selected to receive the grants. 

Despite this criticism, I will continue to contribute to GiveDirectly, and I hope I do not discourage others from contributing.

Updated February 9, 2014 (minor edit)