Water Demand Management and Household Level Metering in Low-Income Areas of Urban Africa

Introduction

Increasing the level of water metering is an objective of most initiatives for the operational restructuring of African water utilities promoted by donors and development agencies from the ‘90s (Tyman and Kingdom, 2002; WSP 2008). Water metering penetration is among the most common benchmarking indicators which are adopted to measure the performances of water utilities throughout the world. Two indicators are used for this purpose: proportion of metered connection over the total number of connections and proportion of metered water over the total volume of water sold. Household level water metering is usually considered a good practice and recommended (IBNET, 2017).¹ Under an economic perspective, measuring water consumption is expected to improve efficiency in the allocation of the water resource, by allowing the adoption of volumetric fees instead of fixed fees and the transmission of price signals to water users. For this reason, metering is considered a main tool for Water Demand Management (WDM) (Arlosoroff, 1998; Macy, 1999; Gumbo and Van der Zaag, 2002; Mulwafu et al., 2003), which aims at keeping the demand for water compatible with environmental scarcity. The idea is that consumption can be limited by confronting people with their actual consumption and charging them accordingly. Under a more technical perspective, metering is used to reduce Non Revenue Water (NRW), caused by leakages and illegal connections, as it makes it possible for technicians to identify the area where leakages occur and to fix the issue. Metering is not only supported by water conservation arguments, but sometimes also by an equity argument (Staddon 2010), because, through metering, everyone is charged what she actually consumes. This reflects a liberal view of equity, while the actual social outcomes of a metered system in terms of social equity might be more complex and depend on the tariff system which is used.

In contrast with other benchmarks and targets set for the African water sector, which remain largely unmet - for example access to safely managed water source in urban Sub Saharan Africa was only 50% in 2017 (JMP 2019) -, water metering at household and at water point level is quite successful. “In Africa, water metering is surprisingly widespread, with many utilities reporting 100% metering in their service areas. (…) the average metering ratio is 75%” (Banerjee, 2010, p. 6). The idea that metering level should approximate universal coverage is widely accepted and recommended for developing countries (Kingdom et al., 2006; WSP, 2008).

The study discusses the arguments behind the widespread acceptance of the target of 100% metering, focusing on the suitability of household level metering for low-income settlements of urban Africa. An alternative to universal metering is selective metering, which can be based on the selection, by the utility of the customers to be metered, or can be based on their voluntary enrolment in a metering program. Voluntary enrolment is prone to an adverse selection bias, as only the customers whose consumption is low are likely to save money by measuring the actual volumes, while customers with high consumption will probably prefer the flat rate (Lago and Möller-Gulland, 2012). Adverse selection is related to information asymmetries and it occurs when a product or service is selected by only a certain group of people who offer the worst return for the company and, in this case, also for the society which fails to achieve the water conservation outcomes envisaged. Selective metering can also be part of a strategy by the water utility to discriminate between customers’ categories, instead of allowing self-selection. The chapter argues that poor, informal neighbourhoods of African cities are not a priority target in this sense.