The results of the analysis show that the contextual variance of the level of access to drinking water falls from 1.034 in the empty model to 0.454 in the model taking into account contextual variables (EPE, population served and type of village), i.e. a fall of 56%. This justifies the existence of contextual effects that play a part in determining household levels of access to drinking water. We can deduce from this that the variability in the level of access to drinking water could well be explained by the village context. It seems clear that households in villages with boreholes are more likely to have easy access to water than those in villages connected to a water supply network. The purpose of the logistic model is to measure this evidence.
The results show that villages with boreholes are five times more likely to have easy access to drinking water than villages without water supply. All other things being equal.
For households in villages without water supply, access to drinking water is a daily chore (Fig. 3). All the households interviewed in villages without water supply said they had difficulty accessing drinking water. Some women in these villages travel between 1 and 3 kilometers with their basins to use the water points in villages with a water supply system.
In addition, an analysis of the nature of the relationships between users of the same service point can give an indication of their level of access to water. Naturally, conflictual relations between users of the same water point lead to difficult access for them.
In many villages where there is no water supply, wells cannot be used continuously. They sometimes dry up temporarily during the day, particularly at peak times. This leads to quarrels, and sometimes even conflicts, between users. In these villages, 59% of households surveyed felt that they had strained relations with their "co-users". These conflicts are also seen in some villages with water supply, but to a lesser extent. Conflicts between users are even less persistent in villages with boreholes, where access to drinking water through private connections is more widespread.
The provision of domestic water from private connections is a solution to conflicts between users. The public nature of wells and standpipes means that users have equal access to and use of the equipment. This could lead to conflicts over access to and sharing of the resource. Figure 4 shows the nature of relations between users of the same water point in villages without a water supply system.
In addition, households' levels of access to drinking water vary according to the means used to access the resource. The results of the multi-level analysis of access to drinking water show that households with private connections are seven times more likely to have easy access to drinking water than those who walk. Households that walk to water points are more exposed to the physical difficulties associated with the distance and transport of drinking water. Households that have a tap at home are not exposed to the constraints associated with jostling at public drinking water points. They use their drinking water equipment in good time. This is permanent access with no constraints (distance, transport) or congestion caused by other users.
The cart is used as a locomotive to transport drinking water from service points to people's homes. In this way, the physical effort required by households to walk to water points is borne by the animal (usually a donkey or horse) pulling the cart. This is a relief and saves time for the households that use the cart. As a result, they have easier access to drinking water than households that walk. Given that the price of water at the standpipe and the difficulty of drawing it from the well are the same for all households, the use of a cart to access service points can reduce the physical burden of transporting drinking water.
Levels of access to drinking water deteriorate progressively if the means of access used changes from a private connection to walking. The means of access is therefore an indicator for assessing a household's level of access to drinking water. Indeed, 77% of households that walk to access drinking water consider that they have difficult access, compared with 75% of households that use carts and 43% of those that have taps at home.
The use of carts, or the need to walk, can also give an indication of the distances separating households' places of residence and drinking water supply points. The field surveys revealed that the use of a cart highlights two essential parameters that need to be taken into account when assessing a household's access to drinking water: (i) distance (transport of water) and (ii) the volume of water required by the household (the household's water requirements).
In this study, the perception of distance is analysed according to three categories (close, far and very far), which are left to the discretion of the household surveyed:
-
Close: when the distance between the place of residence and the place of supply is less than 200 meters;
-
Far: when the distance between the place of residence and the place of supply is between 200 and 500 meters;
-
Very far: when the distance between the place of residence and the place of supply is more than 500 meters.
Depending on their geographical location, rural households consider themselves to be near, far or very far from drinking water supply points, according to the indicators proposed.
Normally, households that are close to water points should be more likely to have easy access to drinking water than those located far or very far from service facilities. The use of multi-level analysis confirmed this probability. Households located less than 200 meters from a water point are 2.6 times more likely to have easier access to drinking water than households located more than 500 meters from the same water point. All other things being equal.
Another factor contributing to the disparity in levels of access to drinking water is the price at which water is sold, particularly in villages with a water supply system. "Water drawn from nature is free. It then undergoes a number of costly operations (the water is collected, stored, treated, transported, purified and purified). Users cover part of the costs incurred in collecting water and treating it, and the public authorities top up the price paid by users when they pay subsidies" [SMETS, 2004].
In Senegal, the selling price of water includes the costs of production, maintenance and renewal of equipment and other works. "Determining the cost is now based on the notion of a balance between costs and revenues" [DIOP, 2011].
For the purposes of this study, rural populations have incomes classified under three headings: low, medium and medium-high. As a result, low-income households have more difficulty accessing drinking water than middle-income or medium-high-income households. The price at which water is sold is assessed by households according to whether it is low, affordable or high. The results of the multi-level analysis show that households that consider the price of water to be low are 1.9 times more likely to have easy access to drinking water than those that consider the price of water to be high. For the latter, the price of water is a constraint on access to the resource. The purchase price of water is an element that differentiates the level of access to drinking water. The lower the price of water for a household, the easier its access to drinking water. All other things being equal.
The last individual variable included in the model for analysing access to drinking water is the source most used by the rural household. However, the list of individual characteristics (inherent to households) that can help explain disparities in levels of access to drinking water between households is far from exhaustive. There are a number of other individual effects that can be taken into account to give further thought to the determinants that explain the disparate levels of access to drinking water in rural households.