[Response by Sophie Trémolet and Diane Binder, June 2009]
Targeting subsidies is essential for improving affordability for the poor and promoting universal access. For those customers who are already connected, it has often been recommended to introduce lifeline rates. Lifeline rates are targeted subsidies based on the consumption level of households, i.e. subsidised rates for a first block of consumption, which is enough to cover basic needs (for water, for example, 25 litres / capita/ day). This consists of using consumption volume as a targeting mechanism and provides an easy quantitative target as to what and how much to subsidize.
Lifeline rates are a way of improving the design of increasing block tariffs, since only the first block, covering basic needs, is subsidized. Anything above would be charged at a commercial rate, i.e. based on the marginal cost of service provision. This mechanism appears more accurate than increasing block tariffs, since in this case, only the lower block is subsidized. However, that means that the definition of the lower or the “lifeline” block is very critical.
Strengths of lifeline rates approaches in providing affordable access to basic services for the poor include:
- Lifeline rates are seen as fair and necessary to provide basic levels of service to the poor. They are considered as an instrument of social policy and as a way to increase the purchasing power of the poor.
- Lifeline rates are seen as one of many instruments that the government can use to mitigate the burden of increased tariffs on the poor.
- Restricting the price subsidy to the initial block of consumption offers a less costly alternative to across-the-board price subsidies while preserving their politically attractive universal protection feature.
- Lifeline rates are easy to implement at minimal administrative costs, and provide highly predictable support to the poor.
- Lifeline rates also provide an incentive for large consumers to economize on use and thereby fit a sector efficiency objective.
However, evidence is mixed whether lifeline rates actually reach their objectives:
- In practice, quantity-based consumption subsidies do a poor job of targeting benefits to the poor, and risk of exclusion is frequent: poor consumers are not necessarily small consumers.
- Lifeline rates may be regressive if poor (and more numerous) families consume more than the lower block. This may also be the case when several families are renting a property together or when homeowners resell the service to their neighbours who are not connected.
- The implementation of lifeline rates suggests that households are metered and consumption is easily estimated; however, poor households often do not have access to the service or are not metered and they would therefore not receive the subsidy.
- They can create significant price distortions in the behaviour of water utilities and their consumers as the price paid would not reflect the marginal cost of the service.
- Lifeline rates and increasing blocks tariff systems raise the question of financial sustainability. When, like in most cases in Latin America and Asia, only the last block is commensurate to O&M costs (and even more rarely to capital costs) at a level never reached by residential consumption, cross-subsidies that finance lifeline rates are not sustainable.
- Lifeline blocks are detrimental to coping strategies such as secondary retailing, as the lifeline tariff would not be transferred to poor customers when SSIPs buy in bulk and sell the service on a retail basis.
A few ways to improve tariff design when metering is in place are:
- To offer a separate social tariff (or vouchers) determined on the basis of means tests of various kinds;
- To link subsidies to service levels so as to use “self-selection” as a targeting mechanism;
- To implement connection rather than consumption subsidies, as they are a better way of targeting subsidies especially when coverage rates are low.
- To consider pre-paid meters (possibly at subsidized tariffs) which allow households to control their total expense on the service.