Source: Adapted from O'Connor, K. (1995) "Airport Development in Southeast Asia", Journal of Transport Geography, Vol. 3, No. 4, pp. 269-279.
- Stage 1 (Initial development; connecting effect). During the 1930s basic linear services were established. Since the technical capabilities of aircrafts were limited, especially in terms of range and capacity, intermediate stops had to be made even if many of these stops represented limited commercial opportunities. For instance, a flight across the Pacific required stops at Hawaii, Wake, Midway, Guam and the Philippines islands.
- Stage 2 (By-passing effect). During the 1940s and 1950s, the technical capabilities of aircrafts improved substantially, which enabled to by-pass several intermediate stops. Routes between major destinations still had a linear structure.
- Stage 3 (Proximity effect). In the 1960s and 1970s, aircraft technology permitted long distance routes, which supported the development of a network structure reflecting better the size and function of markets. As such, the airports of large cities started to see the development of feeder services. The structure of the network remained highly influence by the existing urban hierarchy.
- Stage 4 (Hubbing effect). Through the accumulated effect of hubbing, the 1980s and 1990s saw the creation of large hubs handling the great majority of air traffic, especially at the international level. Hub and spoke networks have the advantage of offering a larger market coverage (number of airports) with a smaller number of services. A hub also enables to reconcile more effectively long distance and regional air services. However, the selection of a hub was the outcome of commercial decisions made by air carriers. It does not necessarily reflects the existing urban hierarchy. For instance, although Atlanta and Dubai are important cities, their air traffic level are more an outcome of their hubbing function than of their rank in the global urban system.