A basic rural road network, or core network, should be set out so that most of the population of an area are at a reasonable walking distance from it. This core network should link to higher level networks where rural bus services can operate to towns where services are concentrated. Normally existing rural networks must extended and improved to provide basic access. Once in place they must be maintained. The coverage of this core network must be defined with reference to a required service quality, generally expressed in terms of an objective measure of the maximum time or effort that users must put into getting to where they want to go.
Resources available dictate the stringency of the measure of service quality. The richer the society the more it can pay to avoid wasting time and effort in travel. Again, the index of service quality could require that certain key facilities be accessible to all within a given maximal travel time (for example, to medical emergencies). To satisfy it would involve not only improving roads, but also improving the transport services using them Such an access criterion is good in that it takes account of the entire journey, including travel time, vehicle availability, and walking distance but it can be complicated to apply. The core network must not be too extensive as it will become unsustainable due to lack of resources to maintain it. In densely populated areas the existing network may need to be severely pruned.
A simple measure of service quality is preferable for the rural roads which interest us, since simplicity and transparence are essential to involve people in the planning process. Service quality is normally defined in terms of a maximum walking distance to a year-round motorable road, linked to the higher level road network that interconnects the towns and cities. This index, developed by the World Bank is called the Rural Access Index (RAI). Assuming motor transport is available (and the money to pay the fare), services, generally grouped in the market towns, are made accessible to the villages. In general, markets are the primary destination for rural people. The modest income earned from sale of their produce can be used to buy basic necessities. At the same time, market days break the isolation of rural life.
This walking distance will ideally be based on a nationally defined maximum distance. In the absence of a national benchmark, future transport users should be involved in setting it. For example, it could be agreed upon that all rural villages be within 2km (25 to 40 minutes walking time) of a motorable road. The motorable road network must be improved or extended (or perhaps even reduced) to meet this. The walking distance can vary, depending essentially on resources available to extend or repair the network. Remember that the network required to bring roads closer to people grows as walking distance is shortened or populations are more sparsely distributed. Levels of income and population density must enter into negotiating the distance. Topography and climate must also be considered, for obvious reasons. A 2km stroll over flat and dry terrain to get the next bus is not the same as a scramble over steep mountain paths in torrential rain carrying a 40kg load.
In Sub-Saharan Africa, poverty and lack of funds to construct and maintain, allied to low population densities, may oblige planners to apply quite a long walking distance to limit the size of the core network. In Asia, on the other hand, where incomes are higher and topography often difficult, a shorter distance may be used. In all cases stakeholders should be consulted, especially when planners know little or nothing about the transport conditions on the tracks and paths which the people use to get to a motorable road. Consultation is also essential in dense networks, where there are many choices, so that the core network be selected that best fits local travel patterns.
Since there can be a number of alternative routes, depending on the density of the network, a simple ranking criterion for choosing among road improvements or extensions must be used, such as the lifetime cost per km for each link (direct path to the core motorable network) relative to population served. When reliable data exists it can usefully be weighted to to favour specific problem areas by incorporating, for example, indicators of the spatial distribution of poverty.