Child at public water station (Bolivia) — Photo: Stephan Bechenheimer/World Bank

How using an old water tariff design could help improve water service in developing countries

It could increase cost recovery, encourage water conservation, make poor household subsidies much more efficient, and possibly help to reduce open defecation.

The Problem

In most developing countries income disparities are high, with the majority of the population poor. Tariffs are chronically set low, almost always below the costs of providing sustainable service, to accommodate the poor.

Below-cost tariffs undermine the utility’s ability to maintain the infrastructure needed to provide service. As the infrastructure deteriorates, service quality decreases, eroding customer willingness to pay, which further weakens the utility financially.

Breaking this cycle is impossible without an infusion of funds. But, project-based financing is inaccessible without increased, reliable, revenue to improve the financial viability of the utility.

Revenue gains from one-off investments, such as infrastructure improvements to reduce non-revenue water (e.g., leaks) can help significantly. But, the gains usually dissipate rapidly after project completion unless internal funds are available to sustain them.

Increasing domestic tariffs are particularly hard to do. Politicians fear the impact they will have on the poor. It is easy for regulators to rule that current poor service does not justify them.

The wealthy in the population, who should be paying the full cost, benefit the most from low tariffs. They are more often connected to the network, live in areas that receive better service (e.g., less rationing, higher pressures), and use more water per capita than the poor.

Attempts to increase revenues, and encourage conservation by, charging increasing block tariffs, which increase with consumption, hurt the poor when large extended families live together as single households.

Government attempts to subsidize the poor directly based on household income are usually difficult to implement.

What is needed is for everyone but the poor or low-income households to pay the full cost.

How an old tariff design could help

In the early years before water meters were introduced, all tariffs were flat, with water bills often determined based on a cost allocation method, like the number of water fixtures (e.g., sinks, bathtubs, toilets, outside facets) in or associated the household (e.g. in the yard).

While flat water tariffs are not helpful, the cost allocation method is. The number of household fixtures can be used as an indicator of household income and as an indicator of water demand.

The very poor usually don’t have household connections. They get their water from shared taps or kiosks. Others, who are a little better off, have a single household faucet. Those further up the income ladder might have two fixtures (two household faucets, maybe a toilet), or three (two faucets and a toilet). The wealthiest usually have many fixtures in their large houses — multiple sink faucets , multiple toilets, multiple showers or bathtubs, multiple yard facets, possibly a pool.

One would not expect households that only have one or two fixtures do typically use as much water or have as high a peak demand for water as households with four, five for more fixtures, where three or more could be in use at the same time.

In Africa, “…62% of urbanites live and work in slums”. Many slum dwellers don’t have in-house water fixtures, those that do, have no more than one or two (a kitchen faucet, a toilet, or a second faucet).

How Fixture Counts Could be Used to Design Effective Tariffs

Water tariffs should be designed to cover the total annual costs of sustainably providing the service. Ideally, this includes the costs of operating and sustaining the infrastructure (i.e., maintenance and repairs to damaged or deteriorated infrastructure); the payment of debt service; and generate sufficient “net revenues” — revenues in excess of expenses — to enable the utility to financially withstand periodic reductions in demand (e.g. from economic, weather-related fluctuations) and to be able attract capital and to partially self-fund larger projects needed to rehabilitate deteriorated, or build new, infrastructure.

All customers that can afford to pay the full cost of service should be asked to do so. Those who can’t pay the full cost should be asked to pay what they can. Using the same tariff for both has effectively caps the tariff for those who can afford to pay the full cost.

Considering that households with few fixtures are typically poor and have a lower peak demand, it makes sense to classify them as a separate customer class for tariff purposes — households with only one or two fixtures could be identified, grouped into a “limited fixtures” class and charged a separate tariff from the households with a more household fixtures.

Benefits of creating a separate “limited fixtures” class

Tariff designs typically address seven objectives: economic efficiency, fairness, equity, revenue sufficiency, net revenue stability, simplicity and understandability, and resource conservation.

Economic efficiency refers to the optimal use of resources. The tariff should be designed to help the utilities provide service to each household or entity in a way that meets their needs while minimizing waste and inefficiency. By creating a separate “limited-fixtures” class, the utility could alter their service characteristics (e.g. pressure) reducing the cost of service without having a negative impact on household use.

The tariffs charged by the utility should be perceived as “fair” by water users and the public generally. Having lower tariffs for poor or low-income households based on their having fewer fixtures, and corresponding a lower peak demand, would be perceived as more fair than existing tariffs that charge every household the same tariff.

Tariff equity means that equals are treated equally. Creating a separate “limited fixtures” class would treat all limited-fixture households the same. Currently, tariff designs typically charge poor, limited-fixture households for pressure and peak capacity that they shouldn’t be responsible for.

Increasing the perception of fairness and equity regarding the tariff might also increase poorer customers’ willingness to pay.

Tariffs should provide sufficient revenues to cover the utilities costs of service, including “net revenues” discuss above. Separating the poor, limited-fixture households from the rest would allow for increases in the tariffs charged to wealthier households, increasing utilities’ revenues and providing them with additional resources to improve and sustain their services.

“Limited fixtures” tariffs would be simple and understandable. They would not be difficult to implement. Tariffs for capacity and consumption by limited-fixture households would designed the same way as for other customer classes. Fixture counts to identify limited-fixture customers could be verified when connecting the household. Revenue audits could help ensure compliance.

Creating a separate “limited fixtures” tariff would allow tariffs to be designed for both “limited fixtures” households and other wealthier households to encourage resource conservation. The impact of incentives are relative to income. Those with higher incomes are less responsive to water conservation pricing than those lower down on the income scale. Thus, separating the poor with limited fixtures from the more wealthy, with more water fixtures, would enable conservation tariffs to be designed that are effective for each class, with higher block tariffs design to influence wealthier customers.

A separate “limited fixtures” class would make customer class subsidies or government transfers in support of the poor more efficient, as the amount of the subsidy or transfer would be specific to the class, meaning that it would the difference between the estimated cost of serving the “limited fixtures” class and estimated tariff revenue from the class.


Obviously, depending on where the line is drawn, not every limited-fixture household identified would be poor or low-income. It may make sense to grant the utility some discretion to exclude wealthy households, ensuring consistent implementation could be difficult. It might also make sense stratify middle-income households by fixture-count blocks to be fair to them when income inequality is significant, creating a number of fixture-count customer classes with separate tariffs.

How limited fixture tariffs could help reduce open defecation

Developing “limited fixtures” classes might also help reduce open defecation in slum areas by designing tariffs that incentivize toilet use. For example, by allowing toilet fixtures to be excluded from the fixture count or by charging higher water tariffs in areas with high open-defecation rates.