There is one additional lesson of a methodological nature that has been learned. This concerns city size. Size implies both numbers of people and geographic area or territory. Typically, we use population numbers as the measure of size. Numbers grow either through natural increase, or net migration. All of this is well understood and usually included in assessing and reporting city size. But cities also grow through expanding their boundaries, often incorporating more people into the administrative boundaries of the city. This is less commonly included in reports of urban size and growth and constitutes an important weakness in our measures.

We saw that Chinese cities underwent major changes in the 1980s, often greatly expanding their administrative boundaries, and incorporating both rural areas and rural populations (Xie Wenhui, 1987). We also saw this most recently in the study of Changchun, where the city administrative area included 18,000 square kilometers, while the actual urban area itself was a mere 1,100 square kilometers. In Indonesia, we also found that cities on Java are typically "underbounded," meaning that the actual built-up urban areas extends beyond the administrative boundaries. In the less densely settled outer islands, however, cities are typically "overbounded," including extensive rural areas and populations within their administrative boundaries (Kartomo, 1992).*4

Note as well that these different sources of growth represent different demands on urban resources and different problems for urban administrators. Growth by natural increase implies demands for maternal and child care and for early education. Growth by inmigration usually implies the demand for employment and for housing and services, usually for a population with large numbers of young males. Growth by aerial expansion, on the other hand, often implies no increased demand for services, but may imply an increase in revenues and in capacities for more rational planning.

In effect, our measures of city size should systematically include territory as well as population, and should show the relative proportion of these three sources of growth.


*4. The Xie Wenhui paper appeared in the proceedings of the 1987 Kobe Conference on Medium-Sized Cities. The Kartomo paper appeared in Population Dynamics and Port City Development, AUICK, 1992.