Decomposing geographic accessibility into component parts: Methods and an application to hospitals

While many studies define and measure the geographic accessibility of facilities, research has failed to explain why the accessibility is high or low, except to conjecture that it has to do with (1) the number of facilities or (2) the locations of these facilities. We demonstrate that accessibility may also be low in a region because (3) the transportation network is inefficient or (4) the population distribution is difficult to serve with few facilities. This paper also develops measures for the degree each of these four factors affect accessibility using p-median and GIS techniques. An example is provided using hospital locations in four southern US states.