Geographic Access to Physician Services

Abstract

Prior to the last decade, it was widely accepted that physicians were geo­ graphically maldistributed; indeed, a major federal initiative, the National Health Service Corps, was based on that premise (24). The maldistribution most commonly referred to was between urban and rural areas, although there was some reference to maldistribution within urban areas as well. The usual evidence cited to show the existence of mal distribution was unequal physician/population ratios between urban and rural areas, examples of which are in Table 1 . (Although the data that I and others cite in fact distinguish metropolitan and nonmetropolitan communities, I use the less exact, but more familiar terms, urban and rural.) In addition, especially in the late 1960s and early 1970s, there were examples of towns that werc losing physicians; that is, physicians were retiring and not being replaced. The conventional view of maldistribution was not undergirded by a well­ articulated theory of how physicians chose to locate their practices; numbers analogous to those in Table 1 were believed to speak for themselves. To the degree a theory of physician location existed, it seemed to be of the following sort: Physicians prefer to locate in cities. Because of virtually unlimited demand for their services, or their ability to create demand, physicians can afford to locate in cities and thus can continue to indulge their preferences. (See Ref. 19, Appendix A, for examples of quotations along these lines.) Moreover, it was probably the case that during the 1960s, maldistribution, as conventionally defined, was worsening; that is, the inequalities in physician/population ratios were growing. This, however, is difficult to