Many developing countries face severe challenges with the reliability of water supplies.
These supplies are often characterised by intermittence, low pressure and poor water
quality. Despite its contribution towards water-related illness and the significant coping
burden it imposes on households, water supply reliability remains a difficult attribute to
measure. A key challenge is the lack of a universal definition of water supply reliability.
The issue of unreliability in water supply and the financial cost it imposes on
households is of profound relevance in South Africa – a country whose social policies
include a Free Basic Water policy which entitles all households to a free lifeline supply
of 6,000 litres per month. This thesis examines household experiences of unreliable
water supplies and in particular, explores the question as to what constitutes a reliable
water supply, and household responses to unreliable water supplies.
The analysis draws on literature reviews and a household survey conducted in periurban
communities in the Limpopo Province of South Africa in 2012. A systematic
review of definitions and assessment criteria used in studies of water supply reliability
demonstrates that there is no consensus on what constitutes a reliable water supply.
Assessment criteria also vary greatly, with the most common criterion in urban settings
being the duration and/or continuity of supply in hours per day. In rural settings, the
proportion of functional water systems is commonly assessed. A discrete choice
experiment was conducted to elicit households’ preferences for a reliable water supply.
Results indicate that overall, households value notification of interruptions and having
water available for longer durations during the day, and would be willing to pay for
these improvements. However, there is some heterogeneity in these preferences as
wealthier households, who have drilled their own wells and are no longer dependant on
the public supply are less willing to pay for improvements in the water supply.
Findings from a systematic review of household strategies to cope with unreliability
reveal that relatively wealthy households incur significant direct costs from strategies
such as drilling wells and installing water storage tanks, poor households expend time
and energy in collecting water from other sources. Income, level of education, land
tenure and extent of unreliability are the main determinants of which strategies are
adopted. Results from the survey in Limpopo highlight that Free Basic Water is not
actually free; households spend significant proportions of their income on buying water,
ii
drilling wells and treating the water prior to consumption. Coping costs increase with
wealth status and are higher in communities without alternative water sources such as
springs. Notably, for many households the lifeline supply of 6,000 litres per month is
unmet.
The findings from this thesis highlight the need for consensus on the definition, and
assessment approach for water supply reliability. Further, the analysis of households’
responses to unreliable water supplies in South Africa draws attention to how poor
reliability negates the Free Basic Water policy. Without reliable water supply services,
the objectives of improving public health and promoting equity cannot be met.