March 2004

Report No 1296/1/04


E.1 Background to and Motivation for the Project

In 2001/2002 Marketing Surveys and Statistical Analysis (MSSA), a consulting market research company based in Pretoria, undertook a pilot study into payment strategies for low-income water consumers in the then Pretoria, Durban, Johannesburg and Cape Town Metropolitan areas1. Likewise between 1998 and 2000, Economic Project Evaluation (Pty) Ltd., (EPE), a Johannesburg based economic consultancy, undertook a comprehensive study to determine the price elasticity of demand for water amongst low, medium and high-income groups in Alberton and Thokoza2 . Both these were undertaken for and funded by the Water Research Commission of South Africa (WRC). As a consequence of the interest engendered by these studies from local government departments and private companies such as consulting engineers, in April 2002 the WRC commissioned MSSA and EPE, working jointly, to undertake a more in-depth study along the lines of the earlier projects.

1Van Vuuren DS. Researching, Developing and Testing of Payment Strategies for the Lower Income Groups at four selected communities in order to manage changes for water use, WRC Report No 1202/01/03.
2Veck A. and Bill M. Estimation of the Residential Price Elasticity of Demand for Water by means of a Contingent Valuation Approach, WRC Report No. 790/1/00, 2000

E.2 Aims of the Study

The aims of the present study are as follows:

E.3 Hypotheses to be Tested

Two interrelated hypotheses were tested, these being:

  1. Price does have an effect on the amount of water demanded by all classes of water consumers, and
  2. Consumer's attitudes to water usage and their perception about water consumption may be changed by appropriate water payment strategies.

The results of the study showed that these hypotheses were correct.

E.4 Methodological Approach to the Study

Both the PPST (Participative Payment Strategy Testing) and CV3 (Contingent valuation) methodologies required a systematic process of interaction between researchers and water consumers, which culminates in data being gathered by means of two surveys (called Survey No. 1 and Survey No. 2) for each methodology. For the CV experiment, these surveys record the different ways consumers use water, i.e., for indoor/outdoor usage, cooking, washing, etc., and changes to the amount of water used for these different purposes as the price charged for water changes. For the PPST experiment, these surveys record data about household water consumption, their water bills, their perceptions about the quantity of the water used and other related perceptions regarding their payment behaviour.

Surveys were conducted by face-to-face interviews among low, medium and high-income population groups of residential water users in three newly created metropoles, Tshwane, Ethekwini and the city of Cape Town.4

3It is important to note that the CV methodology yields the short-term price elasticity for water. To determine the long-term price elasticity of demand, a macro-economic model needs to be developed and exercised using historic data based on water usage. This study estimated the short-term price elasticity of demand for water; the attempt to estimate the long-term price elasticity of demand for water was unsuccessful due to the lack of historic information; this will be commented upon further later.
4The Tshwane metropole includes the former Pretoria metropolitan area and the Ethekwini metropole includes the former Durban metropolitan area.

E.5 Summary of Results of the Study

E.5.1 The PPST Experiment

  1. Respondent and Demographic Information
    In each metropole 225 households comprising of the low, middle, and high-income groups took part in the study. Information gathered included the respondent's name, address, residency period on their property, size of property etc. Demographic information included household income, gender, age of occupants, number of adults and number of children living in the household.
  2. Perceptions about the Water Account
    It was found that most of the lower income households in the three metropoles find their accounts difficult to understand and pay their accounts by cash. The higher income groups make use of a variety of payment options such as cash, credit card, bank transfers, etc.
  3. Estimated Water Usage and Sources of Variation
    In all three metropoles, the degree of correlation between estimated and actual household water usage was obtained including sources of variation between these two variables. The sources of variation included the number of people living on the stand, the size of the stand, income and education levels.

E.5.2 The CV Experiment

a) Perceived Water Consumption - Survey No. 1
The results of this survey produced a water usage profile for consumers in each of the three metropoles in each of the three income groups, i.e., low, mid and high. Comparisons between the profiles for the low income group in each of the three metropoles are shown in Fig E.1 below.

Fig E1 Comparison of Perceived Detaile Monthly Water Usage Profile for the Low Income Groups - Survey 1

b) Determination of the Price Elasticity of Water - Survey No. 2

Using the water usage profiles obtained in Survey No. 1, the CV experiment was undertaken (Survey No. 2) which yielded the price elasticities of demand for indoor and outdoor water usage amongst the three income groups in the three metropoles. The results obtained are shown in Table E.1 below.

Income Groups Total price Elasticity of Demand Range of Values Falling within 95% Confidence Level
Low Income Groups:
Cape Town
-0.227, -0.502
-0.048, -0.290
-0.038, -0.195
Mid Income Groups:
Cape Town
-0.022, -0.191
-0.025, -0.177
-0.012, -0.161
High Income groups:
Cape Town
-0.039, -0.220
-0.048, -0.223
-0.050, -0.225

Note: * = Statistically significant difference (on a 5% level) in price elasticity between low, middle, and high income.

Table E.1 Price Elasticities of Demand for Water in the Three Metropoles

From Table E.1, it can be seen that in all the three metropoles and for all income groups the price elasticity of demand for water is less than -1; this means that the elasticities are inelastic.

E.6 Conclusions Drawn from the Study
Conclusions drawn form the study centred around how the results of the study could influence water policy formulation, these conclusions are summarized below for each experiment.

E.6.1 Policy Implications

E.7 Future Work
This study has concentrated on residential water only, so far as future work is concerned it is recommended that similar studies be undertaken in the industrial and agricultural sectors of the economy.

E.8 Final Note
Water policy in South Africa has historically addressed the supply-demand nexus very narrowly. It has focussed almost exclusively in developing new supplies for specific needs such as agricultural development. It is suggested that policy should change to focus on regional and national needs and must also address the demand side of the equation as well as the supply-side.