Volokh posts on water privatization, and a reader sends in a link to this study on water privatization in Argentina. I had heard about the success of that effort, but hadn't seen this particular paper yet.

It may be counter-intuitive, but is it generally true that privatization produces the greatest successes in the poorest regions? "we find that child mortality fell 5 to 7 percent in areas that privatized their water services overall; and that the effect was largest in the poorest areas."

That's not actually too surprising, because in the richer areas there is likely to be surplus (money, water, other resources or services). In California, those whose water is "taken away" from them (we could consider this to be every organization except farms) are still abundantly supplied with water. But in a poorer area, inefficient water distribution may mean that those who most need a jug of water for a feverish child may not have it, and cannot pay any price for it. Making the water distribution more efficient means at least it's reliably there in need.

My personal preference would be for water privatization to be accompanied with a system of "water coupons" to allow poor people to buy a family's weekly/monthly supply of water. Like food coupons in some cities, and similar to the Heifer Project, these could be supported by private donation from richer regions. This allows even small donations to have an immediate, direct and measurable effect, whereas foreign aid to public water suppliers in poor countries is frequently wasted and has a delayed and uncertain effect. A charitable system of water coupons requires privatization in order to work at all.