The setting of a transportation network is the outcome of decisions about connecting locations. Using a simple example involving four locations, three general network connectivity options can be considered:
- Least cost to use. This network seeks maximum connectivity by linking all locations directly. Assuming a distance of 10 km between each node, the network will have a total length of 60 km. This is the most convenient network for users, as each location can be reached in the shortest possible manner. However, this is the most expensive network for infrastructure providers to build and operate. Such networks usually cannot be built because of high costs and redundancy.
- Least cost to build. This network seeks the minimum construction costs by providing minimal connectivity. Each location is connected, but with a high level of detour for some connections. In the above example, the least cost to build a network is a length of 20 km, three times less than the least cost to use. This network configuration comes at a high cost for the users that spend, on average more time and energy traveling between locations.
- Hybrid. Most existing networks result from a compromise between the least cost to use and the least cost to build. In the above case, the compromise has a length of 30 km with moderate connectivity. For instance, a least-cost air transportation network would be a pure hub and spoke network. However, such networks are also complemented by direct connections between major destinations.
Economic development has an important impact on network connectivity, as the higher the density of economic activities, the better the incentives to increase connectivity. A regional network could begin as a least-cost-to-build network and evolve to become a least-cost-to-use network because the demand has increased.