We use two separate quasi-natural experiments to explore the relative importance of information and administrative costs in explaining non-take-up of water subsidy. The first "experiment" shows that the take-up rate of a household with lower administrative costs is not significantly different from otherwise identical households. In contrast, using the same program, the second "experiment" reveals that the take-up rate of a household that is more likely to be informed is substantially higher

more »

... ared to otherwise identical households. These findings support the idea that information plays a major role in explaining non-take-up of water subsidy.